Document LKX2o1rJ1w9rj8087RB8LLdj7

DRAFT DO NOT CITE OR QUOTE May 2010 External Review Draft APPENDIX A Dioxin Workshop Report NOTICE THIS DOCUMENT IS AN EXTERNAL REVIEW DRAFT. It has not been formally released by the U.S. Environmental Protection Agency and should not at this stage be construed to represent Agency policy. It is being circulated for comment on its technical accuracy and policy implications. National Center for Environmental Assessment Office of Research and Development U.S. Environmental Protection Agency Cincinnati, OH EPA/600/R-09/027 May 2009 Summary of U.S. EPA Dioxin Workshop February 18-20, 2009 Cincinnati, Ohio National Center for Environmental Assessment Office of Research and Development U.S. Environmental Protection Agency Cincinnati, OH 45268 DISCLAIM ER This document summarizes the discussions presented at the Dioxin Workshop in February 2009, in Cincinnati, OH, as documented by the Session Co-Chairs. This document is not all inclusive or binding. Conclusions and recommendations to the U.S. EPA may not represent full consensus. The views expressed in this document are those of the Dioxin Workshop Panelists and do not necessarily reflect the views and policies of the U.S. Environmental Protection Agency. Mention of trade names or commercial products does not constitute endorsement or recommendation for use. Preferred Citation: U.S. Environmental Protection Agency (U.S. EPA). (2009) Summary o f U.S. EPA Dioxin Workshop: February 18-20, 2009. U.S. Environmental Protection Agency, National Center for Environmental Assessment, Cincinnati, OH. EPA/600/R-09/027. This document is a draftfor review purposes only and does not constitute Agency policy. A-ii DRAFT--DO NOT CITE OR QUOTE TABLE OF CONTENTS DIOXIN WORKSHOP TEAM................................................................................................. A-iv ACKNOWLEDGMENTS......................................................................................................... A-iv INTRODUCTION...................................................................................................................... A-1 REFERENCES............................................................................................................... A-2 SCIENTIFIC WORKSHOP TO INFORM THE TECHNICAL WORK PLAN FOR US. EPA'S RESPONSE TO NAS COMMENTS ON THE HEALTH EFFECTS OF DIOXIN PRESENTED IN US. EPA'S DIOXIN REASSESSMENT................................ A-3 SESSION 1: QUANTITATIVE DOSE-RESPONSE MODELING ISSUES................ A-3 SESSION 2: IMMUNOTOXICITY............................................................................... A-6 SESSION 3A: DOSE-RESPONSE FOR NEUROTOXICITY AND NONREPRODUCTIVE ENDOCRINE EFFECTS............................................... A-8 SESSION 3B: DOSE-RESPONSE FOR CARDIOVASCULAR TOXICITY AND HEPATOTOXICITY................................................................................. A-11 SESSION 4A: DOSE-RESPONSE FOR CANCER.................................................... A-13 SESSION 4B: DOSE-RESPONSE FOR REPRODUCTIVE/DEVELOPMENTAL TOXICITY....................................... A-16 SESSION 5: QUANTITATIVE UNCERTAINTY ANALYSIS OF DOSE- RESPONSE.......................................................................................................... A-20 APPENDIX A: 2009 US. EPA DIOXIN WORKSHOP AGENDA........................................ A-24 APPENDIX B: 2009 US. EPA DIOXIN WORKSHOP QUESTIONS TO GUIDE PANEL DISCUSSIONS..................................................................................................... A-31 APPENDIX C: 2009 US. EPA DIOXIN WORKSHOP DRAFT SELECTION CRITERIA TO IDENTIFY KEY IN VIVO MAMMALIAN STUDIES THAT INFORM DOSE-RESPONSE MODELING FOR 2,3,7,8-TETRACHLORODIBENZO-p-DIOXIN (TCDD)................................................ A-34 This document is a draftfor review purposes only and does not constitute Agency policy. A-iii DRAFT--DO NOT CITE OR QUOTE DIOXIN W ORKSHOP TEAM The Dioxin Workshop Team, under the leadership of Peter W. Preuss, Director, NCEA, comprised the following members: National Center for Environmental Assessment, Office of Research and Development, U.S. Environmental Protection Agency, Cincinnati, OH 45268 Belinda S. Hawkins Janet Hess-Wilson Glenn Rice Jeff Swartout Linda K. Teuschler Bette Zwayer Argonne National Laboratory, Argonne, IL 60439 Maryka H. Bhattacharyya Andrew Davidson Mary E. Finster Margaret M. MacDonell David P. Peterson ACKNOWLEDGMENTS The Track Group, Alexandria, VA 22312 Kara Hennigan Alan Minton Brandy Quinn ECFlex, Inc., Fairborn, OH 45324 Dan Heing Heidi Glick Amy Prues Lana Wood IntelliTech Systems, Inc., Fairborn, OH 45324 Cris Broyles Luella Kessler Stacey Lewis Linda Tackett This document is a draftfor review purposes only and does not constitute Agency policy. A-iv DRAFT--DO NOT CITE OR QUOTE IN T R O D U C T IO N This document provides a summary of the Scientific Workshop to Inform EPA's Response to National Academy of Science Comments on the Health Effects of Dioxin in EPA's 2003 Dioxin Reassessment. The U.S. Environmental Protection Agency (U.S. EPA) and Argonne National Laboratories (ANL), through an inter-Agency agreement with the U.S. Department of Energy, convened this scientific workshop ("Dioxin Workshop") on February 18-20, 2009, in Cincinnati, Ohio. The goals of the Dioxin Workshop were to identify and address issues related to the dose-response assessment of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). This report summarizes the discussions and conclusions from this workshop. Previously, at the request of the U.S. EPA, the National Academy of Sciences (NAS) prepared a report, H ea lth R isk s fr o m D io xin a n d R e la te d C om pounds: E va lu a tio n o f the E P A R ea ssessm en t (NAS, 2006), which made a number of recommendations to improve the U.S. EPA's risk assessment for TCDD (U.S. EPA, 2003). The 3-day Dioxin Workshop was convened specifically to ensure that the U.S. EPA's response to the NAS recommendations focuses on the key issues and reflects the most meaningful science. The Dioxin Workshop included seven scientific sessions: (1) Session 1: (2) Session 2: (3 ) Session 3A (4) Session 3B (5) Session 4A (6) Session 4B (7 ) Session 5: Quantitative Dose-Response Modeling Issues Immunotoxicity Dose-Response for Neurotoxicity and Nonreproductive Endocrine Effects Dose-Response for Cardiovascular Toxicity and Hepatotoxicity Dose-Response for Cancer Dose-Response for Reproductive/Developmental Toxicity Quantitative Uncertainty Analysis of Dose-Response During each session, the U.S. EPA asked a panel of expert scientists to: identify and discuss the technical challenges involved in addressing the key NAS comments on the TCDD dose-response assessment in the U.S. EPA Reassessment (U.S. EPA, 2003); discuss approaches for addressing the key NAS comments; and identify important published, independently peer-reviewed literature, particularly studies describing epidemiologic and in vivo mammalian bioassays, which are expected to be most useful for informing the U.S. EPA's response. The sessions were followed by open comment periods during which members of the audience were invited to address the Panels. At the conclusion of the open comment periods, the Panel Co-Chairs were asked to summarize and present the results of the panel discussions. The summaries could include minority opinions stated by panelists. The main points derived from the session summaries were used to prepare this document. Additionally, this document includes a list of the session panelists and their affiliations and three appendices. Appendix A presents the Dioxin Workshop Agenda. Appendix B identifies the charge questions presented to the Panel. Appendix C describes draft study selection criteria proposed by the Dioxin Workshop Team for consideration by the workshop panelists. This document is a draftfor review purposes only and does not constitute Agency policy. A-1 DRAFT--DO NOT CITE OR QUOTE REFERENCES NAS (National Academy of Sciences). 2006. Health Risks from Dioxin and Related Compounds: Evaluation of the EPA Reassessment. National Academies Press, Washington, DC (July). Available at http://www.nap.edu/catalog.php?record id=11688. U.S. EPA (U.S. Environmental Protection Agency). 2003. Exposure and Human Health Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. NAS review draft, Volumes 1-3 (EPA/600/P-00/001Cb, Volume 1). U.S. Environmental Protection Agency, National Center for Environmental Assessment, Washington, DC (December). Available at http://www.epa.gov/nceawww1/pdfs/dioxin/nas-review/. This document is a draftfor review purposes only and does not constitute Agency policy. A-2 DRAFT--DO NOT CITE OR QUOTE SC IENTIFIC W O R K SH O P TO IN FO R M TH E TECH NICA L W O R K PLA N FO R U.S. EPA'S RESPONSE TO NAS COM M ENTS ON THE HEALTH EFFECTS OF DIOXIN PR ESENTED IN U.S. E P A 'S DIO XIN REASSESSM ENT Dioxin Workshop Co-Chairs: Peter W. Preuss and Glenn Rice The Dioxin Workshop session summaries were prepared by the session panel Co-Chairs with input from the panelists, as requested by the U.S. EPA prior to the workshop. The Co Chairs subsequently presented these summaries to all of the workshop participants during designated periods at the workshop. In these summaries, the U.S. EPA asked that the Co-Chairs summarize the key issues from the panel discussions. Because the sessions were not designed to achieve consensus among the panelists, the summaries do not necessarily represent consensus opinions; rather, they reflect the essence of the panel discussions. Some of the specific points may represent the views of multiple panelists, while others only the views of a single panelist. Prior to the summarizations, there were opportunities for public comments on the discussion topics. Some Co-Chairs met with their sessions' panelists after their sessions ended to develop these summaries, while others developed reports based on their personal notes. Because Session 5 was the last session of the workshop--with little time provided to develop the summary--the Co-Chairs circulated a draft for comment by the Session 5 panelists after the workshop, prior to finalizing the session summary. The U.S. EPA collected the session summaries and then prepared this document. A draft of this document was distributed to all of the session Co-Chairs to provide them with a final opportunity to comment and make revisions. Finally, it should be noted that U.S. EPA was not prescriptive to the session Co-Chairs with respect to the format of the presentation materials and provided no specific instructions, resulting in unique formats among the session summaries. S E SSIO N 1: Q U A N T IT A T IV E D O S E -R E S P O N S E M O D E L IN G ISSU E S This session discussed the general dose-response modeling issues related to TCDD. Many of these issues were highlighted by NAS (2006). There was a general introductory presentation on TCDD kinetics, including information and uncertainties pertaining to the conversion of administered doses in animals to human body burden (BB) and additivity to background issues. This presentation was followed by a Panel discussion on the state of the science regarding dioxin dose-response modeling issues. Session 1 Panelists (Session Co-Chairs are identified by asterisk) Bruce Allen, Bruce Allen Consulting Lesa Aylward, Summit Toxicology Roger Cooke, Resources for the Future Kenny Crump, Louisiana Tech University Mike DeVito, U.S. EPA Dale Hattis, Clark University Rick Hertzberg, Biomath Consulting Rob McDowell, U.S. Department of Agriculture Jim Olson, State University of New York, University at Buffalo This document is a draftfor review purposes only and does not constitute Agency policy. A-3 DRAFT--DO NOT CITE OR QUOTE *Lorenz Rhomberg, Gradient Woody Setzer, U.S. EPA *Jeff Swartout, U.S. EPA Please note that the use o f the term "concluded" or "recommended" in this summary does not mean that a consensus was reached. Session Summaries were written from the material prepared by the non-EPA/ANL Co-Chair and represent a synopsis o f the panel discussions. Key Study Selection Criteria The Panel discussed the advantages and disadvantages of using key study criteria (Appendix C). They concluded that a p rio ri criteria foster transparency and consistency, and could deflect a posterio ri criticism. However, the Panel also acknowledged that having a p rio ri criteria could introduce the potential for excluding useful data. Although the key study criteria provided by the U.S. EPA listed studies using TCDD only as a criterion, the Panel posed the possibility of using closely related dioxin-like compounds (DLCs) as surrogates for TCDD. The criterion for use of data from mammalian studies only was one criterion that received generalized support due to the lack of extrapolation protocols for nonmammalian species. The Panel also discussed the specific exposure-duration criterion and asked if there should be a preference for longer-term rather than acute studies. The Panel made three suggestions to modify U.S. EPA's key study selection criteria: (1) Define more relevant exposure-level (i.e., dose) cut points using tissue concentrations. (2) Reword statistical criteria to include do-it-yourself analysis. (3) Reword the response criteria to clarify "outside of normal range." Dose M etrics The Panel discussed the relative merits of various measures of dose for modeling TCDD dose response. One general conclusion was that tissue concentration (TC) is the preferred metric, especially lipid-adjusted TC, because this measure more closely approximates exposures close to the target tissue when compared to administered doses. However, the Panel acknowledged that these data are often unavailable. They further noted that BB, which is defined as the concentration of TCDD in the body (ng/kg body weight) (U.S. EPA, 2003), might be useful as a surrogate for TC provided the two measures were proportional. The Panel suggested that a linear approach to BB estimation, which was utilized by U.S. EPA (2003), is too simplistic because this approach does not take into account toxicokinetic issues related to TCDD--e.g., sequestration in the liver and fat, age-dependent elimination, and changing elimination rates over time. The Panel recommended the use of kinetic/mechanistic modeling to the extent possible to quantify tissue-based metrics. The Panel raised the issue of whether the preferred dose metric would be different for different endpoints and exposure durations. This led to the Panel's comment that the peak exposure might be a more important metric than average BB for variable exposure scenarios. Given this discussion about different exposure durations being relevant to a specific endpoint, the Panel suggested that the U.S. EPA also consider peak measures in dose-response modeling. This document is a draftfor review purposes only and does not constitute Agency policy. A-4 DRAFT--DO NOT CITE OR QUOTE The last point raised in this part of the discussion centered on the possibility of dose errors in experimental studies. The Panel highlighted the need for the U.S. EPA to consider dose error (i.e., uncertainty in the x-axis of the dose-response curve) when using dose surrogates. Dose-Response M odeling of M am m alian Bioassays The Panel considered several issues related to dose-response modeling of mammalian bioassay data for TCDD: supralinearity and incomplete response data ("anchoring"), defining the benchmark response (BMR) level with respect to establishing the point of departure (POD), and the use of threshold modeling--as further explained below. The Panel discussed the specific issues of supralinearity and anchoring raised by the U.S. EPA with respect to modeling noncancer endpoints. The panel recognized that, for many of the most sensitive endpoints, the response at the lowest dose is high (e.g., quantal responses above 25% and continuous endpoints differ substantially from the mean, often implying 100% incidence in the treated animals). This lack of response anchoring at the low end of the doseresponse curve (near the BMR) results in the higher responses determining the shape of the curve. The Panel asked whether new tools might be needed or whether the current tools could be applied differently. In the context of developing new tools, the Panel emphasized the need for collaboration between biologists and mathematicians. When discussing application, the Panel suggested that the problem with supralinearity might be overcome by simply dropping the requirement for using the lower bound on the Benchmark Dose. In addition, the Panel posed several more approaches for further consideration in dose-response modeling by the U.S. EPA: (1) Combine similar data sets to fill in data gaps. (2) Use mechanistic approaches to model the data gaps. (3) Dichotomize continuous data. Finally, the Panel acknowledged that, in certain situations, there simply may not be enough information to provide meaningful answers. The Panel discussed the BMR level for establishing a POD in the context of deriving a Reference Dose (RfD). The Panel generally agreed that, while the effective dose level (ED01) used in the 2003 Reassessment may be useful for comparative analysis across endpoints, the ED01 estimates developed for all endpoints considered in the Reassessment were not appropriate for deriving an RfD because they were not based on the effect's adversity. The panel noted that ED01 also is much lower than typical EPA BMR levels. The Panel recommended that the U.S. EPA work to define endpoint-specific BMRs based on the consideration of adversity. Given that the same uncertainty factor framework is applied to all PODs, the Panel emphasized the need for consistency in BMRs; numerical consistency is needed for quantal BMRs and consistency in the choice of biological relevance should be applied for continuous BMRs. The Panel generally discouraged threshold modeling by stating that thresholds are very difficult to pin down and suggested that the lower bound may always be zero. This document is a draftfor review purposes only and does not constitute Agency policy. A-5 DRAFT--DO NOT CITE OR QUOTE Dose-Response M odeling of Epidem iological Studies The Panel noted that many studies have been published with measured concentrations of TCDD that could be used for dose reconstruction. In this discussion, the Panel acknowledged that use of these data would entail dealing with toxicity equivalence (TEQ) issues and pharmacokinetic (PK) modeling. Pertaining to the use of these data for quantitative risk assessment by the U.S. EPA, the Panel posed the question, "At what point does indirect or confounded human data supersede controlled animal bioassay data?", or alternatively, "How much human data uncertainty can we tolerate?" The Panel suggested, at the least, that the epidemiologic data could be used to "ground-truth" the animal bioassay modeling results. Supporting Inform ation The Panel acknowledged that Ah receptor (AhR) binding affinities are not necessarily tied to endpoint sensitivity, but they reiterated the need to consider mechanistic modeling to aid in developing appropriate dose metrics or filling in data gaps in the existing dose-response data. References NAS (National Academy of Sciences). 2006. Health Risks from Dioxin and Related Compounds: Evaluation of the EPA Reassessment. National Academies Press, Washington, DC (July). Available at http://www.nap.edu/catalog.php?record id=11688. U.S. EPA (U.S. Environmental Protection Agency). 2003. Exposure and Human Health Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. NAS Review Draft (EPA/600/P-00/001Cb). U.S. Environmental Protection Agency, National Center for Environmental Assessment, Washington, DC. Available at http://www.epa.gov/nceawww1/pdfs/dioxin/nas-review/. SESSIO N 2: IM M U N O TO X IC ITY The U.S. EPA plans to consider development of a quantitative dose-response assessment for the immunologic effects associated with TCDD exposure. Such an assessment would be based on information in U.S. EPA (2003), NAS (2006) and key studies identified in this workshop. The purpose of this session was to identify and discuss key issues pertaining to doseresponse assessment for dioxin-induced immunologic effects. Session 2 Panelists (Session Co-Chairs are identified by asterisk) Roger Cooke, Resources for the Future Rob Goble, Clark University *Belinda Hawkins, U.S. EPA Nancy Kerkvliet, Oregon State University Manolis Kogevinas, Centre for Research in Environmental Epidemiology Robert Luebke, U.S. EPA Paolo Mocarelli, University of Milan *Allen Silverstone, State University of New York, Upstate Medical University This document is a draftfor review purposes only and does not constitute Agency policy. A-6 DRAFT--DO NOT CITE OR QUOTE  Courtney Sulentic, Wright State University Nigel Walker, National Institute of Environmental Health Sciences Please note that the use o f the term "concluded" or "recommended" in this summary does not mean that a consensus was reached. Session Summaries were written from the material prepared by the non-EPA/ANL Co-Chair and represent a synopsis o f the panel discussions. Key Study Selection Criteria The Panel first addressed the Key Study Selection Criteria proposed by the U.S. EPA (Appendix C). The Panel raised the issue that the key study criteria do not apply to most studies designed to investigate immunotoxicity, including those used to calculate ED0is (U.S. EPA, 2003). The Panel observed that most dioxin immunotoxicity studies are relatively high dose (>200 ng/kg-d) acute studies and/or use parenteral rather than oral administration. The Panel discussed several studies often considered important for assessing the immunotoxic effects of TCDD exposure. The Oughton et al. (1995) mouse bioassay was discussed and, although the study does meet the proposed criteria, it could not be considered a key study; specifically, the Panel contended that since there were no functional alterations observed or measured in this bioassay, the changes in cellular phenotypes are only "suggestive" of immune alterations and cannot be regarded as having immunopathologic significance. The Panel discussed two additional studies for further consideration by the U.S. EPA: Baccarelli et al. (2002). The Panel discussed this as a potentially key human epidemiological study that should be reviewed and considered further by the U.S. EPA. It measured the level of IgG, demonstrating a significant decline relative to dioxin body burdens. Smialowicz et al. (2008). The Panel noted that this study identified the antibody response to sheep red blood cells (SRBCs) as the critical effect, labeling this protocol as a functional assay. The Panel stated that if modeled, the U.S. EPA could calculate the BMR for this endpoint as 1 standard deviation from the control mean. References Baccarelli, A., P. Mocarelli, D.G. Patterson et al. 2002. Immunologic effects of dioxin: New results from Seveso and comparison with other studies. Environ. Health Perspect. 110(12):1169-1173. NAS (National Academy of Sciences). 2006. Health Risks from Dioxin and Related Compounds: Evaluation of the EPA Reassessment. National Academies Press, Washington, DC (July). Available at http://www.nap.edu/catalog.php7record id=11688. Oughton, J.A., C.B. Pereira, G.K. Dekrey, J.M. Collier, A.A. Frank and N.I. Kerkvliet. 1995. Phenotypic analysis of spleen, thymus, and peripheral blood cells in aged C57BI/6 mice following long-term exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol. Sci. 25(1):60-69. This document is a draftfor review purposes only and does not constitute Agency policy. A-7 DRAFT--DO NOT CITE OR QUOTE Smialowicz, R.J., M.J. DeVito, W.C. Williams and L.S. Birnbaum. 2008. Relative potency based on hepatic enzyme induction predicts immunosuppressive effects of a mixture of PCDDS/PCDFS and PCBS. Toxicol. Appl. Pharmacol. 227(3):477-484. U.S. EPA (U.S. Environmental Protection Agency). 2003. Exposure and Human Health Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. NAS Review Draft (EPA/600/P-00/001Cb). U.S. Environmental Protection Agency, National Center for Environmental Assessment, Washington, DC. Available at http://www.epa.gov/nceawww1/pdfs/dioxin/nas-review/. SESSION 3A: DOSE-RESPONSE FOR NEUROTOXICITY AND NONREPRODUCTIVE ENDOCRINE EFFECTS The U.S. EPA plans to consider development of a quantitative dose-response assessment for neurological and/or nonreproductive endocrine effects associated with TCDD exposure. Such an assessment would be based on information in U.S. EPA (2003), NAS (2006) and key studies identified in this workshop. The purpose of this session was to identify and discuss key issues pertaining to dose-response assessment for dioxin-induced neurological and/or nonreproductive endocrine effects. Session 3A Panelists (Session Co-Chairs are identified by asterisk) *Maryka Bhattacharyya, Argonne National Laboratory Mike DeVito, U.S. EPA Mary Gilbert, U.S. EPA Rob Goble, Clark University Nancy Kerkvliet, Oregon State University Fumio Matsumura, University of California-Davis Paolo Mocarelli, University of Milan Chris Portier, National Institute of Environmental Health Sciences Lorenz Rhomberg, Gradient Allen Silverstone, State University of New York, Upstate Medical University Marie Sweeney, National Institute of Occupational Safety and Health *Bernie Weiss, University of Rochester Please note that the use o f the term "concluded" or "recommended" in this summary does not mean that a consensus was reached. Session Summaries were written from the material prepared by the non-EPA/ANL Co-Chair and represent a synopsis o f the panel discussions. W hat Are the Key Questions Regarding These Endpoints? The Panel used the following question to initiate discussion: "A re there identifiable indices o f neurotoxicity and nonreproductive endocrine effects in anim al studies and human p o p u la tio n s? " Under this discussion topic, the Panel discussed three endpoints: neurotoxicity (with focus on developmental exposures), thyroid dysfunction (e.g., thyroid hormone deficits), and diabetes. The Panel also addressed the relevance of windows of vulnerability to each This document is a draftfor review purposes only and does not constitute Agency policy. A-8 DRAFT--DO NOT CITE OR QUOTE endpoint. The Panel acknowledged that, in some cases, the window of exposure may precede the window of expression of toxicity. Epidem iological Study Selection Developm ental Neurotoxicity The Panel recognized that an unusual feature for this endpoint is that there are sufficient human data for dose-response modeling (e.g., Dutch children [Huisman et al., 1995; Patandin et al., 1999] and U.S. children [Jacobson and Jacobson, 1996]) and there is an internal dose metric (serum concentrations). Additionally, the Panel discussed recent studies that address this endpoint in humans (from Japan [reference not provided] and Holland [e.g., Koopman-Esseboom et al., 1996; Vreugdenhil et al., 2002]). For continued investigation into this endpoint, the Panel raised two issues to the U.S. EPA: Conduct an evaluation of whether a modeled effect can be attributed to TCDD and not some other persistent organic pollutant (POP), although the Panel recognized that it is unlikely U.S. EPA will be able to distinguish among these exposures because other POPs are intrinsic confounders in the Dutch study. Allow animal data to inform the dose-response modeling of epidemiological data. Thyroid Dysfunction The Panel identified the availability of human data for this endpoint (e.g., Calvert et al., 1999; Koopman-Esseboom et al., 1994). Much of the thyroid dysfunction literature has been published since the 2003 Reassessment (e.g., Wang et al., 2005; Baccarelli et al., 2008). The Panel also noted the availability of an internal dose metric (serum concentrations). Additionally, the Panel discussed the mechanistic studies in animals that link TCDD to thyroid dysfunction. For continued investigation into this endpoint, the Panel raised three issues for the U.S. EPA to consider: Consider the newly available human data since the Reassessment. Investigate and clarify of the role of TCDD-induced thyroid dysfunction in developmental neurotoxicity. Evaluate and determine whether an effect can be attributed to TCDD or other contaminants. D ia b etes The Panel discussed that data suggest that diabetes incidence in those under 55 years old may be associated with exposure to PCBs. They acknowledged that whether this is a dioxin-like compound (DLC) mediated effect or whether other POPs are responsible is still undetermined. The Panel also acknowledged that no animal model exists for the investigation of xenobioticinduced diabetes, and that separating the injury dose level from the current body burdens would depend on good pharmacokinetics in humans. For continued investigation into this endpoint, the Panel listed two issues for the U.S. EPA to consider: Results from the Anniston study and the Great Lakes Fishermen study (references not provided) should be examined for dose metrics (both studies examine human PCB exposures). This document is a draftfor review purposes only and does not constitute Agency policy. A-9 DRAFT--DO NOT CITE OR QUOTE  Changes of adipose tissue status need to be considered, given that dieting can cause release of lipid-soluble contaminants. References Baccarelli, A., S.M. Giacomini, C. Corbetta et al. 2008. Neonatal thyroid function in Seveso 25 years after maternal exposure to dDioxin. PLoS Med. 5(7):e161. doi:10.1371/journal.pmed.0050161. Calvert, G.M., M.H. Sweeney, J. Deddens and D.K. Wall. 1999. Evaluation of diabetes mellitus, serum glucose, and thyroid function among United States workers exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Occ. Env. Med. 56:270-276. Huisman, M., C. Koopman-Esseboom, V. Fidler et al. 1995. Perinatal exposure to polychlorinated biphenyls and dioxins and its effect on neonatal neurological development. Early Hum. Devel. 41(2):111-127. Jacobson, J.L. and S.W. Jacobson. 1996. Intellectual impairment in children exposed to polychlorinated biphenyls in utero. N. Engl. J. Med. 335:783-789. Koopman-Esseboom, C., N. Weisglas-Kuperus, M.A.J. de Ridder, C.G. Van der Paauw, L.G.M.Th. Tuinstra and P.J.J. Sauer. 1996. Effects of polychlorinated biphenyl/dioxin exposure and feeding type on infants' mental and psychomotor development. J. Pediatr. 97(5):700-706. Koopman-Esseboom, C., D.-C. Morse, N. Weisglas-Kuperus et al. 1994. Effects of dioxins and polychlorinated biphenyls on thyroid hormone status of pregnant women and their infants. Pediatr. Res. 36:468-473. Patandin, S., C.I. Lanting, P.G.H. Mulder, E.R. Boersma, P.J.J. Sauer and N. Weisglas-Kuperus. 1999. Effects of environmental exposure to polychlorinated biphenyls and dioxins on cognitive abilities in Dutch children at 42 months of age. J. Pediatr. 134:33-41. NAS (National Academy of Sciences). 2006. Health Risks from Dioxin and Related Compounds: Evaluation of the EPA Reassessment. National Academies Press, Washington, DC (July). Available at http://www.nap.edu/catalog.php7record id=11688. U.S. EPA (U.S. Environmental Protection Agency). 2003. Exposure and Human Health Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. NAS Review Draft (EPA/600/P-00/001Cb). U.S. Environmental Protection Agency, National Center for Environmental Assessment, Washington, DC. Available at http://www.epa.gov/nceawww1/pdfs/dioxin/nas-review/. Vreugdenhil, H.J., C.I. Lanting, P.G. Mulder, E.R. Boersma and N. Weisglas-Kuperus. 2002. Effects of prenatal PCB and dioxin background exposure on cognitive and motor abilities in Dutch children at school age. J. Pediatr. 140:48-56. This document is a draftfor review purposes only and does not constitute Agency policy. A-10 DRAFT--DO NOT CITE OR QUOTE Wang S.L., P.H. Su, S.B. Jong, Y.L. Guo, W.L. Chou and O. Papke. 2005. In utero exposure to dioxins and polychlorinated biphenyls and its relations to thyroid function and growth hormone in newborns. Environ. Health Perspect. 113:1645-1650. SESSION 3B: DOSE-RESPONSE FO R CARDIOVASCULAR TOXICITY AND H E P A T O T O X IC IT Y The U.S. EPA plans to consider development of a quantitative dose-response assessment for cardiovascular and/or hepatic effects associated with TCDD exposure. Such an assessment would be based on information in U.S. EPA (2003), NAS (2006) and key studies identified in this workshop. The purpose of this session was to identify and discuss key issues pertaining to dose-response assessment for dioxin-induced cardiovascular and/or hepatic effects. Session 3B Panelists (Session Co-Chairs are identified by asterisk) Bob Budinksy, Dow Chemical Manolis Kogevinas, Centre for Research in Environmental Epidemiology Rob McDowell, U.S. Department of Agriculture Jim Olson, State University of New York, University at Buffalo Marian Pavuk, Agency for Toxic Substances and Disease Registry *Jeff Swartout, U.S. EPA *Mary Walker, University of New Mexico Nigel Walker, National Institute of Environmental Health Sciences Please note that the use o f the term "concluded" or "recommended" in this summary does not mean that a consensus was reached. Session Summaries were written from the material prepared by the non-EPA/ANL Co-chair and represents a synopsis of the panel discussions. Key Study Selection Criteria The Panel initially focused on the draft key study selection criteria offered by the U.S. EPA (Appendix C). The panel recommended that for cardiovascular effects, which are not usually observed in rodents, the use of knockout mouse models (ApoE KO and LDLR KO) be moved to the "primary" column because only these studies establish the cardiovascular toxicity model in mice. The panel also was concerned that the gavage procedure can increase mouse blood pressure. Consequently, the panel recommended that gavage studies not be used for the blood pressure endpoint (i.e., only dietary dosing studies should be considered). Human Health Endpoints In relation to the hepatic endpoint, the Panel acknowledged the large body of dose response information on hepatic effects in rodents and that enzyme (mostly CYP1A1) induction was a sensitive effect. However, the Panel cited the lack of linkage of CYP1A1 to downstream events, which complicates the toxicological interpretation of this endpoint, and concluded that This document is a draftfor review purposes only and does not constitute Agency policy. A-11 DRAFT--DO NOT CITE OR QUOTE the more important liver effects in rodents are probably on the "road to cancer." The Panel noted that hepatic effects were not seen in the epidemiological studies, but acknowledged that these studies were not designed to detect them. In relation to the cardiovascular endpoint, the Panel identified hypertension and ischemic heart disease (IHD) as two key endpoints from the epidemiological studies. The Panel recommended that the U.S. EPA perform a meta-analysis of these data. The Panel also commented that recent animal studies support the observations linking TCDD exposure to IHD and hypertension. In particular, the National Toxicology Program (NTP) study shows inflammatory and structural effects on resistant vascular arterioles (NTP, 2006). Additional evidence from the study suggests that the vascular effects may be CYP1A1-dependent. The Panel suggested that the NTP study data might be used as a surrogate for dose-response modeling of hypertension and that such an approach would be supported by data on the role of AhR in vascular function and remodeling. POD Issues The Panel was not supportive of 1% of maximal response (ED0i), which was utilized in the 2003 Reassessment. The Panel concluded that the POD should depend on the specific endpoint and recommended the following to the U.S. EPA: For continuous measures, base the BMR on difference from control. Consider the adversity level--at what point does the endpoint become adverse? For incidence data, set the BMR to a fixed-risk level. Supporting Inform ation The Panel posed several suggestions to the U.S. EPA for reducing uncertainty and improving the knowledge base for TCDD toxicity. Use in vitro data to define uncertainties, such as the relative sensitivity between rodents and humans and around the definition of a POD. Consider studies on dioxin-like compounds (DLCs). Use PK modeling to define the dose metric for hepatic effects. Use body burden or serum concentrations for cardiovascular endpoints. Finally, the Panel recommended that U.S. EPA finish the reassessment quickly and establish a definitive plan to review and incorporate new data as they become available. References NAS (National Academy of Sciences). 2006. Health Risks from Dioxin and Related Compounds: Evaluation of the EPA Reassessment. National Academies Press, Washington, DC (July). Available at http://www.nap.edu/catalog.php?record id=11688. This document is a draftfor review purposes only and does not constitute Agency policy. A-12 DRAFT--DO NOT CITE OR QUOTE NTP (National Toxicology Program). 2006. Toxicology and Carcinogenesis Studies of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) (CAS No. 1746-01-6) in Female Harlan SpragueDawley Rats (Gavage Studies). U.S. Department of Health and Human Services. NTP TR 521. Research Triangle Park, NC (April). U.S. EPA (U.S. Environmental Protection Agency). 2003. Exposure and Human Health Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. NAS Review Draft (EPA/600/P-00/001Cb). U.S. Environmental Protection Agency, National Center for Environmental Assessment, Washington, DC. Available at http://www.epa.gov/nceawww1/pdfs/dioxin/nas-review/. SESSION 4A: DOSE-RESPONSE FO R CANCER The U.S. EPA plans to consider development of a quantitative dose-response assessment for cancer associated with TCDD exposure. Such an assessment would be based on information in U.S. EPA (2003), NAS (2006) and key studies identified in this workshop. The purpose of this session was to identify and discuss key issues pertaining to dose-response assessment for dioxin-induced cancer. Session 4A Panelists (Session Co-Chairs are identified by asterisk) Lesa Aylward, Summit Toxicology Kenny Crump, Louisiana Tech University Dale Hattis, Clark University *Janet Hess-Wilson, U.S. EPA Karen Hogan, U.S. EPA Manolis Kogevinas, Centre for Research in Environmental Epidemiology Marian Pavuk, Agency for Toxic Substances and Disease Registry Chris Portier, National Institute of Environmental Health Sciences Lorenz Rhomberg, Gradient Jay Silkworth, General Electric *Nigel Walker, National Institute of Environmental Health Sciences Please note that the use o f the term "concluded" or "recommended" in this summary does not mean that a consensus was reached. Session Summaries were written from the material prepared by the non-EPA/ANL Co-chair and represent a synopsis o f the panel discussions. K ey Study Selection The Panel discussed both human and rodent studies. In reviewing the epidemiological data, the Panel agreed the EPA should focus on four cohort studies (Dutch cohort, NIOSH cohort, BASF accident cohort, and Hamburg cohort) and pointed out that there are numerous updates and reevaluations of data now in the literature and others will be published soon. The Panel stated that it is appropriate for the U.S. EPA to consider the increase in total cancers for modeling human cancer data, however, Non-Hodgkin's lymphoma, and lung tumors are the main TCDD-related cancer types seen in humans exposed to TCDD. The Panel suggested the U.S. EPA focus the quantitative dose-response modeling on the human data. This document is a draftfor review purposes only and does not constitute Agency policy. A-13 DRAFT--DO NOT CITE OR QUOTE In reviewing the rat data, the Panel identified four new NTP rodent cancer bioassays with liver and lungs as the main target organs. However, they suggested that dose-response modeling efforts should model "all cancers" from these NTP data sets as well and use tumor incidence-- not individual rats as measures. Key Study Selection Criteria The Panel discussed whether data for TCDD only should be used or if PCB126 could be used to develop a dose-response curve. From this discussion, the Panel reached a general agreement that limiting the dose-response modeling and cancer assessment to TCDD only would be the best approach. Regarding the oral dosing regimens, the Panel discussed the differences in results from different bioassays. They concluded that there were insufficient data to pick between oral feed (Kociba et al., 1978) and oral gavage (NTP, 2006) studies, but stated "If all aspects of studies were equal, an oral feed study is preferred." However, given that current data sets are not equal, they agreed that U.S. EPA should consider both feed and gavage studies. The Panel put forth the recommendation that studies that include initiation-promotion model data and TgAC transgenic model data from oral exposure studies should be excluded from the primary category in the key study selection criteria (Appendix C lists the draft study selection criteria distributed prior to the meeting). Studies from both classifications should be moved to the second tier. The Panel was also unsupportive of the "response magnitude outside the range of normal variability" criterion, as they did not believe it was applicable to a cancer endpoint. Critical Endpoints to Consider The Panel recognized that the MOA for TCDD includes cell growth/differentiation dysregulation, that different endpoints (tumor types) across species may be expected, and that there are differences in tumor sites across species. The Panel further acknowledged that there is insufficient information to determine if rodent tumor types observed are relevant to humans. Thus, the Panel suggests the following: U.S. EPA should consider all the observed cancer endpoints in its evaluation. Nonlinear (aka threshold) Versus Linear Dose-Response M odeling The Panel agreed that NTP bioassays appear to demonstrate nonlinear dose response, but they expressed concern about using animal data to infer slope and dose response for humans. The Panel pointed out that there are differences in slopes across different bioassays, and specifically, that some appear linear while others appear nonlinear. Given the observation of both nonlinear vs. linear, the Panel concluded that neither could be ruled out for extrapolation below the POD simply based on the available data. One panelist noted that U.S. EPA Cancer Guidelines (U.S. EPA, 2005) state that only if one can demonstrate that the MOA has a threshold dose-response shape, and can exclude all other potential linear MOAs, can one use a nonlinear model. Lastly, the Panel noted that there are data and rationales to support use of both linear and This document is a draftfor review purposes only and does not constitute Agency policy. A-14 DRAFT--DO NOT CITE OR QUOTE nonlinear response below POD. From this discussion, the Panel raised one possibility to the U.S. EPA: Both linear and nonlinear model functions should be considered in the dose-response analysis. Dose M etrics In considering human data, the Panel expressed a preference for lipid-adjusted serum levels over body burden (BB), and they expressed concerns over the assumptions used in the back calculation of the BB in the epidemiologic cohorts. In considering the rat data, the Panel supported the use of BB--especially lipid-adjusted BB. The Panel, however, did express concern over the sequestering of TCDD in liver and then the use of liver levels in BB calculations. Supporting Inform ation-- Biologically-Based Dose-Response (BBDR) M odels and M OA The Panel discussed BBDR. Though once considered an attractive proposition, BBDR models may mask uncertainty within the models, necessitating them to be used with greater caution. The Panel suggested two issues for the U.S. EPA to consider: If there is a published model, use it if it is valid--do not generate a new model. Focus on the actual experimental data to drive the analysis. References Kociba, R.J., D.G. Keyes, J.E. Beyer et al. 1978. Results of a two-year chronic toxicity and oncogenicity study of 2,3,7,8-tetrachlorodibenzo-p-dioxin in rats. Toxicol. Appl. Pharmacol. 46:279-303. NAS (National Academy of Sciences). 2006. Health Risks from Dioxin and Related Compounds: Evaluation of the EPA Reassessment. National Academies Press, Washington, DC (July). Available at http://www.nap.edu/catalog.php7record id=11688. NTP (National Toxicology Program). 2006. Toxicology and Carcinogenesis Studies of 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) (CAS No. 1746-01-6) in Female Harlan SpragueDawley Rats (Gavage Studies). U.S. Department of Health and Human Services. NTP TR 521. Research Triangle Park, NC (April). U.S. EPA (U.S. Environmental Protection Agency). 2003. Exposure and Human Health Reassessment of 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds. NAS Review Draft (EPA/600/P-00/001Cb). U.S. Environmental Protection Agency, National Center for Environmental Assessment, Washington, DC. Available at http://www.epa.gov/nceawww1/pdfs/dioxin/nas-review/. U.S. EPA (U.S. Environmental Protection Agency). 2005. Guidelines for Carcinogen Risk Assessment. U.S. Environmental Protection Agency Risk Assessment Forum. EPA/630/P-03/001F. This document is a draftfor review purposes only and does not constitute Agency policy. A-15 DRAFT--DO NOT CITE OR QUOTE SESSION 4B: DO SE-RESPO NSE FO R REPRO DUCTIVE/DEVELO PM ENTAL T O X IC IT Y The U.S. EPA plans to consider development of a quantitative dose-response assessment for reproductive and developmental effects associated with TCDD exposure. Such an assessment would be based on information in U.S. EPA (2003), NAS (2006) and key studies identified in this workshop. The purpose of this session was to identify and discuss key issues pertaining to dose-response assessment for dioxin-induced reproductive and developmental effects. Session 4B Panelists (Session Co-Chairs are identified by asterisk) Barbara Abbott, U.S. EPA Bruce Allen, Bruce Allen Consulting Roger Cooke, Resources for the Future George Daston, Procter & Gamble Mike DeVito, U.S. EPA Rob Goble, Clark University *Fumio Matsumura, University of California-Davis Paolo Mocarelli, University of Milan Brian Petroff, University of Kansas *Glenn Rice, U.S. EPA Marie Sweeney, National Institute of Occupational Safety and Health Mary Walker, University ofNew Mexico Bernie Weiss, University of Rochester Please note that the use o f the term "concluded" or "recommended" in this summary does not mean that a consensus was reached. Session Summaries were written from the material prepared by the non-EPA/ANL Co-Chair and represent a synopsis o f the panel discussions. A M ajor Question Posed During this W orkshop Session was "Are Hum an Em bryos and Infants Less Sensitive to Dioxin Exposures Than Some Experim ental A nim als?" The Panel recognized that animal data show a wide range of species sensitivity to dioxin for a given developmental or reproductive endpoint. Presently, there are data for some endpoints that show that human sensitivity is comparable to experimental animals (e.g., semen quality), and for other endpoints the data demonstrate that humans are insensitive compared to other species (e.g., cleft palate). Lastly, the Panel recognized that there are some endpoints for which relative human sensitivity remains uncertain. K ey Study Selection The Panel reviewed the charge questions (Appendix B), discussed them, and listed two issues for the U.S. EPA to consider: Concerning key study determination, use a stepwise approach that is dependent upon the information available and needed to address the question. This document is a draftfor review purposes only and does not constitute Agency policy. A-16 DRAFT--DO NOT CITE OR QUOTE  C oncerning th e key studies inform ing th e P O D and th e P O D endpoint choice, use the P O D to d ep art from w h at is certain and u se a h ig h -co n fid en ce study th at h as found effects at a lo w enough lev el at w h ich o th er effects are p ro tected . T h e P a n e l a lso d e v e lo p e d T a b le 1, b a s e d o n th e in fo rm a tio n p re s e n te d in th is s e ssio n . T a b le 1 iden tifies specific rep ro d u ctiv e and developm ental effects o f concern, listin g w h eth er an effect h a s b e e n o b s e r v e d i n t e s t a n i m a l s a n d e p i d e m i o l o g i c c o h o r ts . I t a l s o i d e n t i f i e s t h e E D i0 estim ated by th e U .S. E P A (2 0 0 3 ) fo r h ealth effects o b serv ed in ro d en t b io assay s. I f th e U .S . E P A d id n o t r e p o r t a n E D i0 f o r a n e f f e c t, th e ta b le id e n tif ie s a s tu d y w h e r e th e e f f e c t w a s rep o rted and th e lo w est study dose w h ere th e effect w as observed. T able 1 also identifies the ep idem iologic cohort w h ere th e specific rep ro d u ctiv e and developm ental effects w ere observed. Epidem iological Study Utility T he P an el review ed th e charge q u estio n s (A p p en d ix B ), discussed them , and m ade tw o su g g estio n s to th e U .S. E PA : C o n cern in g th e ability o f epidem iological studies to inform critical effects, start w ith co n co rd an ce acro ss species (in clu d in g h u m an s) fo r th e spectrum o f effects. C o n cern in g th e ability o f epidem iological studies to in form d o se-resp o n se m odeling, start w ith th e ep id em io lo g y an d th en go to anim al d ata if th e d o se resp o n se has n o t b een w ell characterized fo r an endpoint o f in terest and com pare to anim al data as a reality check. Anim al M odel Utility T h e P a n e l re v ie w e d a n d d is c u s s e d th e c h a rg e q u e s tio n s (A p p e n d ix B ). T a b le 1, w h ic h identifies the effects th at occur in anim als and also have relev an ce to hum ans, sum m arizes m uch o f th is discussion. R eg ard in g th e in flu en ce o f m o d e o f action (M O A ) on anim al m odel choice, th e P an el co n clu d ed th at by ev alu atin g co n co rd an ce am ong h ealth effects rep o rted in ep id em io lo g ic and anim al b io assay data, th e U .S . E P A could id en tify a set o f p lau sib le rep ro d u ctiv e and developm ental effects to consider. A ctual anim al and h um an M O A in fo rm atio n is helpful in th at it creates co m fo rt w ith th e anim al m o d els and in d efin in g the b o u n d aries o f p o ssib le effects. This document is a draftfor review purposes only and does not constitute Agency policy. A-17 DRAFT--DO NOT CITE OR QUOTE TABLE 1 Reproductive/Developmental Effects of Concern for Human Health Endpoint Rodent (ED10 ng/kg-d) Human Notes Sperm Count/M otility Yes (6.2-28; 66-200) Yes E D 10b ases M abley et al. (1992a,b) caudal sperm count and daily sperm production range from 6 .2-28; Gray et al. (1997) epididymal sperm count and total testis sperm counts range from 66-200. Sex Ratio No Yes, Seveso Delayed Puberty M ales Yes (94) Y u-cheng E D 1 0 basis rat m ale puberty delay Gray et al. (1997). Need to qualify epidemiology data because of cohort PCDD/PCDFs exposures. Delayed Puberty in Females Yes No in Seveso Gray and Ostby (2002) report delayed puberty in female offspring o f pregnant rats receiving a single dose o f 1 pg TCDD/kg on GD 15. Cleft Palate Y e s(6300-6400) No E D 10b asis B irnbaum et al. (1989). Premature Senescence Yes No, Seveso Franczak et al. (2006) report that rats prematurely entered reproductive senescence, after receiving cumulative TCDD doses as low as 1.7 pg TCDD/kg. They considered first occurrence o f prolonged interestrous interval (>6 d) as evidence of onset of reproductive senescence. Hormones E2 Yes Yes, M ales-- Seveso Li et al. (1995) report serum estradiol-17f (E2) concentrations induced by equine Chorionic Gonadotropin injection were significantly elevated in female rats orally adm inistered 10 pg/kg T C D D o n PN D 22. W hile E2 decreased dramatically in control animals during the preovulatory LH surge, it did not in TCD D -treated rats. Low Birth W eight Yes (190) Suggestive effect in Seveso in first 8 years after exposure E D i 0b asis Gray et al. (1997). Reproductive Cycling (prolongation) Yes Yes, Seveso Prepubertal exposure Franczak et al. (2006) report loss o f normal cyclicity in female rats at 8 m onths o f age follow ing a cum ulative dose o f 1.7 pg TCDD/kg. This document is a draftfor review purposes only and does not constitute Agency policy. A-18 DRAFT--DO NOT CITE OR QUOTE Supporting Inform ation The Panel reviewed the charge questions (Appendix B), discussed them, and made two suggestions to the U.S. EPA: Concerning deviation from default approaches for noncancer endpoints, there needs to be a careful assessment of the POD and the application of uncertainty factors in light of PK/pharmacodynamics (PD), population characteristics and variability, and MOA information. Concerning the MOA's ability to clarify endpoint and the incorporation of a cascade of cellular event into dose-response for noncancer endpoint, any study that helps inform the dose response should be considered--including studies not specific to dioxins. Complicated mechanistic models need not be developed. Standard dose-response models can be applied. One can look at the cascade of events in a stepwise, simple way. References Birnbaum, L.S., M.W. Harris, L.M. Stocking et al. 1989. Retinoic acid and 2,3,7,8tetrachlorodibenzo-p-dioxin selectively enhance teratogenesis in C57BL/6N mice. Toxicol. Appl. Pharmacol. 98:487-500. Franczak, A., A. Nynca, K.E. Valdez, K.M. Mizinga and B.K. Petroff. 2006. Effects of acute and chronic exposure to the aryl hydrocarbon receptor agonist 2,3,7,8-tetrachlorodibenzop-dioxin on the transition to reproductive senescence in female Sprague-Dawley rats. Biol. Reprod. 74:125-130. Gray, L.E. and J.S. Ostby. 2002. In utero 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) alters reproductive morphology and function in female rat offspring. Toxicol. Appl. Pharmacol. 133(2):285-294. Gray, L.E., J.S. Ostby and W.R. Kelce. 1997. A dose-response analysis of the reproductive effects of a single gestational dose of 2,3,7,8-tetrachlorodibenzo-p-dioxin in male Long Evans Hooded rat offspring. Toxicol. Appl. Pharmacol. 146:11-20. Li, X., D.C. Johnson and K.K. Rozman. 1995. Reproductive effects of 2,3,7,8tetrachlorodibenzo-p-dioxin (TCDD) in female rats: ovulation, hormonal regulation, and possible mechanism(s). Toxicol. Appl. Pharmacol. 133:321-327. Mably, T.A., D.L. Bjerke, R.W. Moore et al. 1992a. In utero and lactational exposure of male rats to 2,3,7,8-tetrachlorodibenzo-p-dioxin. 3. Effects on spermatogenesis and reproductive capability. Toxicol. Appl. Pharmacol. 114:118-126. Mably, T.A., R.W. Moore, R.W. Goy et al. 1992b. In utero and lactational exposure of male rats to 2,3,7,8-tetrachlorodibenzo-p-dioxin. 2. Effects on sexual behavior and the regulation of luteinizing hormone secretion in adulthood. Toxicol. Appl. Pharmacol. 114:108-117. This document is a draftfor review purposes only and does not constitute Agency policy. A-19 DRAFT--DO NOT CITE OR QUOTE NAS (National Academy of Sciences). 2006. Health Risks from Dioxin and Related Compounds: Evaluation of the EPA Reassessment. National Academies Press, Washington, DC (July). Available at http://www.nap.edu/catalog.php7record id=11688. U .S. E P A (U .S . E n v iro n m en tal P ro tectio n A gency). 2003. E x p o su re and H u m an H ealth R eassessm en t o f 2 ,3 ,7 ,8 -T etrach lo ro d ib en zo -p -D io x in (T C D D ) and R elated C om pounds. N A S R ev iew D ra ft (E P A /6 0 0 /P -0 0 /0 0 1 C b ). U .S. E n v iro n m en tal P ro tectio n A g en cy , N atio n al C en ter fo r E n v ironm ental A ssessm ent, W ashington, D C . A v ailab le at h ttp ://w w w .ep a.g o v /n ceaw w w 1 /p d fs/d io x in /n as-rev iew /. SESSIO N 5: Q UAN TITA TIV E U N C ER TA IN TY A N A LY SIS OF D O SE-R ESPO NSE T his session addressed th e uncertainty analysis to b e considered fo r th e dose-response assessm ents. T he session opened w ith a p resen tatio n on cu rren t estim ates o f d ioxin exposure levels. T hen it fo cu sed on th e factors to in clu d e in th e scope o f an u n certain ty analysis in clu d in g dioxin kinetics. Session 5 Panelists (Session Co-Chairs are identified by asterisk) B ru ce A llen, B ru ce A llen C onsulting L esa A ylw ard, Sum m it T oxicology R o g er C ooke, R eso u rces fo r th e F uture K enny C rum p, L o u isian a T ech U niversity M ik e D eV ito, U .S. E P A D ale H attis, C lark U niversity *R ick H ertzberg, B io m ath C onsulting N an cy K erk v liet, O reg o n S tate U n iv ersity L eo n id K opylev, U .S. E P A R ob M cD o w ell, U .S. D ep artm en t o f A g ricu ltu re L orenz R hom berg, G radient W o o d y Setzer, U .S . E P A M arie S w eeney, N atio n al In stitu te o f O ccu p atio n al S afety and H ealth *L inda T euschler, U .S. E P A Please note that the use of the term "concluded" or "recommended" in this summary does not m ean that a consensus was reached. Session Summaries were written from the material prepared by the non-EPA/ANL Co-Chair and represent a synopsis of the panel discussions. T he P anel sum m arized the N A S com m ents regarding uncertainty. A reas fo r im provem ent in c lu d e : E n su re "tran sp aren cy , tho ro u g h n ess, and clarity in q u an titativ e u n certain ty an aly sis." D escrib e and define (q u an titativ ely to the ex ten t p o ssib le) th e v ariab ility and u n certain ty fo r key assum ptions u sed fo r each key en d p o in t-sp ecific risk assessm ent, in clu d in g ch o ices o f d ata set, p o in t o f d ep artu re, d o se-resp o n se m o d el, and do se m etric. Incorporate pro b ab ilistic m odels to rep resen t th e range o f p lausible values. This document is a draftfor review purposes only and does not constitute Agency policy. A-20 DRAFT--DO NOT CITE OR QUOTE  A ssess g o o d n ess-o f-fit o f d o se-resp o n se m odels. P ro v id e u p p er and lo w er b o u n d s on central ten d en cy estim ates fo r all statistical estim ates. W h e n q u an tifica tio n is n o t p o ssib le, clearly state it, an d ex p lain w h a t w o u ld b e req u ired to ach iev e quantification. Identification of Im portant Uncertainties T he P anel review ed th e charge questions (A ppendix B ), discussed them , and listed eight issu es fo r co n sid eratio n by th e U .S. E P A : C o n cern in g species and strain d ifferen ces in th e U .S . E P A 's R esp o n se to N A S , cu rren t U .S. E P A p ro ced u res do n o t tak e th is in to acco u n t w h en selectin g one d ata set fo r risk assessm ent. Issu es in clu d e "W h ere are hum ans in th e distrib u tio n o f p o ten cies th at can b e g en erated ? H o w lik ely is it th a t h u m an resp o n se is sim ilar to th e selected data? C an w e in fer inter-individual v ariability from th ese differences?" C o n cern in g th e u se o f anim al d ata fo r cross species ex trap o latio n to hu m an s (P K and P D u ncertainties), issues to co n sid er in clude differences in d istribution and responses fo llo w in g b o lu s doses from th o se o f su b ch ro n ic and chro n ic protocols; u n certain ty in liv er doses due to sequestration; d ifferen ces in recep to r b in d in g affinity am ong co n g en ers; and age facto rs (e.g., assu m p tio n o f a lifetim e co n stan t daily d o se fo r a can cer e x tra p o la tio n ). C oncerning th e description o f A h R response, biochem ical changes occur at lo w er doses th an to xicological changes. T here should be an effort to identify th e biochem ical changes th at w o u ld m ark A h recep to r b in d in g to in form th e B M R , and, thus, p rev en t toxicity. C oncerning m odel uncertainty, th e m athem atical m odel choice depends on endpoint. T here should b e an effo rt to w ard s d eterm in in g w h at is the m o st sensitive end p o in t(s) for hum ans and co nducting anim al studies to m odel th at endpoint(s). C o n cern in g ex p o su re and dose resp o n se in h u m an studies, en su re en o u g h sim ilarity to current hum an exposure p rofiles (m ixture com position) so th at a dose-response assessm en t can b e done. In co rp o rate n ew epidem iological studies. E v alu ate co n co rd an ce w ith anim al d ata and consistency across studies. P an el-ack n o w led g ed uncertainties include exposure estim ates from person to person, shape o f hum an doseresp o n se curve, h ealth y w o rk e r effect, and age dependence. C oncerning P O D determ ination, uncertainty factors are inherently m athem atically inconsistent and th at should be conveyed in the discussion o f uncertainties w hen interpreting th e PO D . C o n cern in g dose m etric, tissu e co n cen tratio n is p referred. It should b e evaluated ag ain st a b ack g ro u n d o f v ariab ility in A h R -b in d in g expression. T here is u n certain ty in w h at level o f b in d in g sho u ld b e considered, in d ifferen t cell ty p es, tissu es, life stage (developm ent). T he relatio n sh ip b etw een dose m etric and cau satio n o f ad v erse effects should be exam ined. This document is a draftfor review purposes only and does not constitute Agency policy. A-21 DRAFT--DO NOT CITE OR QUOTE Low-Dose Extrapolation T he P anel review ed th e charge questions and discussed th em (A ppendix B ). T he P anel concluded th at cu rve-fitting u n certain ty (fo r a given dataset, dose m etric, and m odel) can be ch aracterized and is u sefu l, b u t, b y itself, it is an in co m p lete ch aracterizatio n of u ncertainty. T he P anel acknow ledged th e difficulty of fu lly characterizin g u n certain ty , especially quantitatively. S o m e p an elists argued th at th e p ro b lem is in su rm o u n tab le and th at no m eaningful u n certain ty analysis is likely to b e perform able. O th er p an elists co ntended that, th e difficulties notw ithstanding, " go o d -faith " efforts to do som ething practical and forth rig h t to characterize uncertainty in low -dose extrapolation w o u ld b e useful and im portant. T he P anel clarified "good fa ith " as m ean in g a ch aracterizatio n th at is u sefu l and n o t m islead in g to d ecisio n m ak ers and is inclu siv e o f approaches th at have m eaningful support in th e scientific com m unity as a w hole. B e in g in " g o o d fa ith " is m o re im p o rta n t th a n b e in g co m p lete (i.e., ad d ressin g ev ery u n c e rta in elem ent), especially since com pleteness is n o t a realistic goal. F rom th is discussion, th e P anel listed fo u r issu es fo r co n sid eratio n by th e U .S . E PA : R ev iew altern ativ e d ata sets, dose m etrics, and m o d els to see w h ere co n seq u en tial u n ce rtain tie s an d im p acts on lo w -d o se im p lica tio n s arise. C o n sid er th e im p acts o f ch o ices am ong p lau sib le altern ativ e d ata sets, d o se m etrics, m odels, and o ther m ore qualitative choices-- issues include how m uch difference the choices m ake and also ho w m u ch relative credence should be p u t to each alternative as a w ay o f g auging and d escribing th e landscape o f im p erfect know ledge regarding possibilities fo r th e tru e dose-response. H ard to do quan titativ ely , since th e facto rs are n ot readily ex p ressed as statistical distributions, b u t can d escribe th e ratio n ale fo r b eliev in g /d o u b tin g each altern ativ e in term s o f available supporting evidence, contrary evidence, and needed assum ptions. E x p ert ju d g m en t m ethods m ay b e helpful in characterizin g th e relative w eig h ts o f scientific credibility am ong alternatives. T he ex p ert ju d g m en t process, w h en conducted system atically, can be th o u g h t o f as adding data to th e assessm ent of cred ib ility o f altern ativ es, rath er th an as ju s t an o p in io n poll. In form ation on p lausibility o f alternative low -dose ex trapolation approaches can com e from ex tern al co n sid eratio n s o f m o d e o f action, an d n o t ju st fro m statistical su ccess at fittin g p artic u la r (h ig h -d o se) d ata sets. C haracterizing uncertainty th ro u g h a v ariety o f approaches could b e tried, and th eir relative m erits and shortcom ings discussed, as a w ay forw ard. C o n sid er th e so u rces o f p o ten tial error, p articu larly in ep id em io lo g ical d ata (e.g., T E F uncertainty and v ariatio n in co n g en er m ix tu res) and if possible quantify th eir im pact on th e d o se-resp o n se assessm ent. Considerations for Conducting Uncertainty Analysis O verall, th e P an el w as split on w h e th e r U .S. E P A sh o u ld do q u an titativ e u n certain ty analyses. T he P an el n o ted th at if d one on only som e o f th e u n certain ties, th en resu lts w o u ld be m islead in g and could b e m isused. U ltim ately, th e P anel listed seven issues fo r consideration by th e U .S. E P A : This document is a draftfor review purposes only and does not constitute Agency policy. A-22 DRAFT--DO NOT CITE OR QUOTE  T he P an el recap p ed w h at som e co n sid er as b ein g th e first in teg rated risk assessm ent, w ith stru ctu re d e x p e rt ju d g m e n t an d u n c e rta in ty an aly sis, i.e., th e R a sm u sse n R e p o rt (W A S H -1 4 0 0 ; U .S. N u c le ar R eg u lato ry C om m issio n , 1975). In th eir d iscu ssio n o f th e report, th e P anel n o ted th at in ad dition to standard event tree/fau lt tree m odeling, this rep o rt also tack led difficult m odel uncertain ty issues involved in accid en t progression, d isp ersio n o f released p o llu tan ts in th e atm osphere, environm ental transport, exposure, health, and eco n o m ic im pacts. A n d th o u g h th e P an el also reco g n ized th at th is m ethod w as no longer state-of-the-art, the P anel contended th at it represents a good exam ple o f a structured approach and m ethodology th at could be b u ilt upon. T he P anel also discussed T E Q s u sed in epidem iological studies, b ased on intake, and reco g n ized th at th e key u n certain ty in w h at w as m easu red w as n o t ju st in tak e b u t also involved P K /P D issues. T he P anel acknow ledged th at th e T E Q system is regularly used on a co n cen tratio n b asis, b u t th ey ex p ressed co n cern th at th e q u alificatio n b eco m es lost. T E Q s ignore pharm acokinetics and the com m on practice o f rounding to orders o f m agnitude introduces m ore error. S tructure th e risk assessm en t along M O A steps-- id en tify key b io ch em ical m easu res (~ 5 -1 0 ) co m m o n across to x ic en d p o in ts an d id en tify th e d eg ree o f m ean in g fu l ch an g e in effect o r effec t v arian ce. M ak e a tab le w ith all o p tio n s fo r d ata set, m o d el, etc.; m ak e b est estim ates/choices and determ ine w h ich o f these choices m atter th e m o st to the answ er. U se expert p anels-- expert ju d g m en t can be collected scientifically (procedures are published). B u t th ere are k n o w n biases; central ten d en cy estim ates w o rk m u ch b etter th an extrem es. U se su p p o rtin g stu d ies to fill in critical d ata g ap s-- In fo fillin g m eth o d s do ex ist (e.g., P K m odeling). P u t short-term studies in to th e " su p p o rtin g in fo " category (unless, o f course, th e risk assessm en t is fo r acu te ex p o su res, such as ch em ical spills). B e creative in th e analysis o f uncertainty. Interm ediate steps betw een A h R b inding and th e end p ro cesses can b e h y p o th esized b ased on data, experiences, and analogies related to o th er chem icals. T he 2003 R eassessm en t presen ted p o tency estim ates on w id e v ariety o f en d p o in ts/m o d els; n eed ed to b e m o re tran sp aren t in th a t discussion. S tatistical g rap h ics can be u sed to convey uncertainties. Reference U .S . N u c le a r R eg u lato ry C o m m issio n . 1975. R eac to r S afety Study: A n A ssessm en t o f A ccid en t R isk s in U .S. C om m ercial N u clear P o w er P lants. W A S H -1 4 0 0 (N U R E G -75-014). W ash in g to n , DC. This document is a draftfor review purposes only and does not constitute Agency policy. A-23 DRAFT--DO NOT CITE OR QUOTE APPENDIX A: 2009 U.S. EPA DIOXIN WORKSHOP AGENDA SCIENTIFIC W ORKSHOP TO INFO RM TH E TECH NICAL W O R K PLAN FO R U.S. E PA 'S RESPO NSE TO NAS COM M ENTS ON THE HEALTH EFFECTS OF DIOXIN PR ESENTED IN U.S. E P A 'S DIO XIN REASSESSM ENT Cincinnati, OH Date: February 18-20, 2009 BACK GROUND/W O RK SH OP OBJECTIVE At the request of the U.S. Environmental Protection Agency (U.S. EPA), the National Academy of Sciences (NAS) prepared a report, H ealth R isks fr o m D ioxin a n d R ela ted C om pounds: E va luation o f the E P A R ea ssessm en t (NAS, 2006), that made a number of recommendations to improve the U.S. EPA's risk assessment for 2,3,7,8-tetrachlorodibenzop-dioxin (TCDD). In response, the U.S. EPA will prepare a technical report that addresses key comments on the dose-response assessment for TCDD. The U.S. EPA intends to develop its response through a transparent process that provides multiple opportunities for input. To assist in this effort, a Workshop will be held to inform the U.S. EPA's evaluation of the NAS recommendations. The Workshop will be open to the public. At the Workshop, the U.S. EPA will solicit input from expert scientists and the public. The goal of the Workshop is to ensure that the U.S. EPA's response to the NAS comments focuses on the key issues and reflects the most meaningful science. The three main objectives of the Workshop are to (1) identify and discuss the technical challenges involved in addressing the NAS key comments on the TCDD dose-response assessment in the U.S. EPA Reassessment (U.S. EPA, 2003), (2) discuss approaches for addressing these comments, and (3) identify key published, independently peer-reviewed literature, particularly studies describing epidemiologic and in vivo mammalian bioassays, which are expected to be most useful for informing the U.S. EPA response. Workshop participants will be encouraged to think broadly about the body of scientific information that can be used to inform the U.S. EPA's response and to participate in open dialogue regarding ways in which the science can best be used to address the key dose-response issues. This Workshop is similar to scientific workshops being conducted under the new review process for the National Ambient Air Quality Standards (NAAQS)1that assess health-related information for criteria pollutants. 1Please see http://www.epa.gov/ttn/naaqs/ for more inform ation on the new NAAQS review process. This document is a draftfor review purposes only and does not constitute Agency policy. A-24 DRAFT--DO NOT CITE OR QUOTE T he W orkshop discussions are expected to build upon tw o p rio r publications: 1. Exposure a n d H um an H ealth R eassessm ent o f 2,3,7,8-Tetrachlorodibenzo-p-D ioxin (TCDD) a n d R elated Com pounds (U .S . E P A , 2 0 0 3 ). T h is e x tern al re v ie w d ra ft p ro v id es a co m p reh en siv e reassessm en t o f d io x in ex p o su re and h u m an h ealth effects. T his " dioxin reassessm en t" w as subm itted in O ctober 2004 to the N ational A cadem y o f S ciences (N A S ) fo r review . 2. H ealth Risks fro m D ioxin a n d R elated Com pounds: Evaluation o f the EP A Reassessment (N A S, 2006). W orkshop p articipants are encouraged to rev iew b o th o f th ese docum ents and other re le v an t m aterials (e.g., th e N atio n al T o x ico lo g y P ro g ram rep o rt on T C D D [N T P, 2 0 0 6 ]) b efo re th e m eeting b ecau se th ey pro v id e im p o rtan t in sig h ts into th e key q u estio n s and challenges. T here are a nu m b er o f open com m ent p eriods th at are in tended to facilitate a b ro ad d iscussion o f th e issues. S cientists w ith sig n ifican t ex p ertise and ex p erien ce relev an t to the h ealth effects o f T C D D or dioxin-like com pounds and associated topics w ill b e asked to serve on "expert p anels" fo r discussions th ro u g h o u t th e W orkshop. W orkshop panelists w ill include a w id e range o f ex p erts re p re se n tin g m an y scien tific areas n e e d e d to assess T C D D d o se-resp o n se (e.g., epidem iology, hum an and anim al toxicology, n u clear recep to r biology, d ose-response m odeling, risk assessm ent, and u n certain ty analysis). T he W o rk sh o p p an elists w ill b e asked to h ig h lig h t sig n ifican t and em erging research and to m ak e reco m m en d atio n s to th e U .S . E P A reg ard in g the design and scope o f th e technical response to N A S com m ents on the dose-response analysis for T C D D -- in clu d in g , b u t n o t lim ited to, re co m m en d atio n s fo r ev alu atin g asso ciated u n certain ty . O pen co m m en t p erio d s w ill fo llo w each panel d iscu ssio n session. P u b lic p articip atio n w ill be en co u rag ed b y w ay o f th ese d esig n ated o pen co m m en t p erio d s and, also, b y p articip atio n in th e scien tific p o ste r sessio n p la n n e d fo r th e sec o n d e v e n in g (F eb ru ary 19). U .S. E P A w ill u se th e in p u t receiv ed durin g this W o rk sh o p as th e fo u n d atio n fo r its d evelopm ent o f a technical w o rk plan fo r responding to th e N A S com m ents on th e T C D D doseresp o n se analysis. T he w o rk plan w ill outline th e schedule, process, and approaches fo r ev alu atin g th e relev an t scien tific in fo rm atio n and ad d ressin g th e key issues. T he w o rk plan also w ill id en tify th e key literatu re to b e u tilized in U .S . E P A 's response. A s a fo llo w -o n activity to this W orkshop, a panel is b ein g estab lish ed u n d er th e F ederal A d v iso ry C o m m ittee A ct (F A C A ) to g u id e an d rev iew th e U .S. E P A 's resp o n se to N A S com m ents. T he F A C A panel w ill be asked to conduct a consultation w ith th e A gency on the d raft tech n ical w o rk plan. A t th e sam e tim e, th e p u b lic w ill also hav e th e o p portunity to pro v id e com m ents to the F A C A panel on th e w o rk plan. T he final tech n ical w o rk plan w ill g u id e the d ev elo p m en t o f th e tech n ical rep o rt th at w ill co n stitu te th e U .S . E P A 's resp o n se to N A S com m ents. D u rin g th e d ev elo p m en t o f th is response, th e U .S. E P A w ill seek ad v ice from th e F A C A panel and th e p u b lic several tim es. F inally, th e F A C A panel w ill b e asked to rev iew the technical report in a p ublic forum . T he prelim inary A g en d a presen ted on th e follo w in g pages m ay be revised prio r to the W orkshop follow ing rev iew by th e session C o-C hairs; th e dates and general tim in g o f the This document is a draftfor review purposes only and does not constitute Agency policy. A-25 DRAFT--DO NOT CITE OR QUOTE sessions, however, will not change. A final Agenda and a set of charge questions, intended to provide general direction for the Workshop discussions, will be posted on the Workshop Internet site (http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=199923) prior to the meeting. A p o ste r sessio n w ill b e h eld on th e ev e n in g o f th e seco n d day (F eb ru ary 19). T h e p u rp o se o f th is p o ster session is to pro v id e a forum fo r scientists to p resen t recen t studies relevant to T C D D dose-response assessm ent and to encourage open discussion about these p re se n ta tio n s. REFERENCES N A S (N ational A cad em y o f S ciences). 2006. H ealth R isk s from D io x in and R elated C om pounds: E v alu atio n o f th e E P A R eassessm ent. N ational A cadem ies P ress, W ashington, D C (July). A v ailab le at h ttp ://w w w .n ap .ed u /catalo g .p h p ?reco rd id = 1 1 6 8 8 . N T P (N ational T ox ico lo g y P rogram ). 2006. T ox ico lo g y and C arcin o g en esis S tudies o f 2 ,3 ,7 ,8 -T etrach lo ro d ib en zo -p -d io x in (T C D D ) (C A S N o. 1746-01-6) in F em ale H arlan S pragueD aw ley R ats (G av ag e S tudies). U .S. D ep artm en t o f H ealth an d H u m an S ervices. N T P T R 521. R esearch T rian g le P ark, N C (A pril). U .S. E P A (U .S . E n v iro n m en tal P ro tectio n A gency). 2003. E x p o su re and H u m an H ealth R eassessm en t o f 2 ,3 ,7 ,8 -T etrach lo ro d ib en zo -p -D io x in (T C D D ) and R elated C o m p o u n d s, N A S re v ie w d ra ft, V o lu m e s 1-3 (E P A /6 0 0 /P -0 0 /0 0 1 C b , V o lu m e 1). U .S . E n v iro n m e n ta l P ro te c tio n A gency, N ational C en ter fo r E n v ironm ental A ssessm ent, W ashington, D C (D ecem ber). A vailable at h ttp ://w w w .ep a.g o v /n ceaw w w 1 /p d fs/d io x in /n as-rev iew /. This document is a draftfor review purposes only and does not constitute Agency policy. A-26 DRAFT--DO NOT CITE OR QUOTE WORKSHOP AGENDA Day 1 8 :00-9:00 9 :0 0 -9 :3 0 9 :3 0 -9 :4 5 R eg istra tio n W elcom e/Purpose of M eeting/Docum ent Developm ent Process Panel Com m ents/Questions on Charge 9:45-2:45 9 :4 5 -10:10 1 0 :1 0 -1 0 :3 5 1 0 :3 5 -1 1 :3 0 1 1 :3 0 -1 :0 0 1 :0 0 -2 :0 0 2 :00-2:45 S e ssio n 1: Q u a n tita tiv e D o se -R e sp o n se M o d e lin g Issu es (H all o f M irrors) Background/Introductory Remarks TCDD K inetics: C onverting Adm inistered Doses in A nim als to Human Body Burdens P resenter: M ich ael D ev ito Panel Discussion Lunch Panel Discussion cont. Open Com m ent Period 2:45-3:05 Break 3:05-5:15 3 :05-3:15 3 :15-4:45 4 :45-5:15 Session 2: Im m unotoxicity (H all o f M irrors) Background/Introductory Remarks Panel Discussion Open Com m ent Period This document is a draftfor review purposes only and does not constitute Agency policy. A-27 DRAFT--DO NOT CITE OR QUOTE 8:00-8:30 8 :00-8:15 8 :1 5 -8 :3 0 Day 2 R eport-O uts for Sessions 1 and 2 (H all o f M irrors) R ep ort-O u t for 1: Q u an titative D ose-R esp on se M od elin g Issues R eport-O ut for 2: Im m unotoxicity 8:30-11:30 Sessions 3A and 3B (concurrent sessions) 8:30-11:30 Session 3A: D ose-R esponse for N eurotoxicity and N onreproductive E ndocrine Effects (H all o f M irrors) 8 :30-8:45 Background/Introductory Remarks 8 :4 5 -1 1 :0 0 Panel Discussion 1 1 :0 0 -1 1 :3 0 Open Com m ent Period 8:30-11:30 Session 3B: D ose-R esponse for C ardiovascular T oxicity and H epatotoxicity (R ookw ood R oom ) 8:30 -8 :4 5 Background/Introductory Remarks 8 :4 5 -1 1 :0 0 1 1 :0 0 -1 1 :3 0 Panel Discussion Open Com m ent Period 11:30-1:00 Lunch 1:00-2:00 R eport-O uts for Sessions 3A and 3B (H all o f M irrors) T he structure o f th e session report-outs w ill in clude th e follow ing: S um m ary o f session presen tatio n including m inority opinion P u b lic com m ents D iscussion 1 :0 0 -1 :1 5 Report-O ut for 3A: Dose-Response for Neurotoxicity and Nonreproductive Endocrine Effects 1 :1 5 -1 :3 0 Open Com m ent Period This document is a draftfor review purposes only and does not constitute Agency policy. A-28 DRAFT--DO NOT CITE OR QUOTE 1 :3 0 -1 :4 5 1 :4 5 -2 :0 0 Report-O ut for 3B: Dose-Response for Cardiovascular Toxicity and Hepatotoxicity Open Com m ent Period 2:00-5:15 2:00-5:15 2 :00-2:15 2:15 -4 :4 5 4 :45-5:15 Sessions 4A and 4B (concurrent sessions) Session 4A: D ose-R esponse for C ancer (H all o f M irrors) Background/Introductory Remarks Panel Discussion Open Com m ent Period 2:00-5:15 2 :00-2:15 2:15 -4 :4 5 4 :45-5:15 Session 4B: D ose-R esponse for R eproductive/D evelopm ental T oxicity (R ookw ood R oom ) Background/Introductory Remarks Panel Discussion Open Com m ent Period 6:45-8:15 P oster Session (R osew ood R oom ) Day 3 8:30-9:30 8 :30-8:45 8 :4 5 -9 :0 0 9:00 -9 :1 5 9 :1 5 -9 :3 0 R eport-O uts for Sessions 4A and 4B (H all o f M irrors) Report-Out for 4A: Dose-Response for Cancer Open Com m ent Period Report-O ut for 4B: Dose-Response for Reproductive/Developm ental Toxicity Open Com m ent Period This document is a draftfor review purposes only and does not constitute Agency policy. A-29 DRAFT--DO NOT CITE OR QUOTE 9:30-3:30 Session 5: Q uan titative U n certainty A n alysis o f D oseR esponse (H all o f M irrors) 9 :3 0 -9 :4 0 Background/Introductory Remarks 9 :4 0 -1 0 :1 0 Evidence of a Decline in Background Dioxin Exposures in Am ericans Between the 1990s and 2000s P resenter: M att L o rb er 10:10-10:30 Break 1 0 :3 0 -1 1 :3 0 Panel Discussion 1 1 :3 0 -1 :0 0 Lunch 1 :0 0 -2 :1 5 Panel Discussion cont. 2 :1 5 -2 :3 0 B reak 2 :30-3:00 Open Com m ent Period 3:00 -3 :1 5 R eport-O ut for 5: Q uantitative U ncertainty A nalysis o f DoseResponse 3 :1 5 -3 :3 0 Closing Rem arks 3:30 A d jou rn This document is a draftfor review purposes only and does not constitute Agency policy. A-30 DRAFT--DO NOT CITE OR QUOTE A PPEN D IX B: 2009 U.S. EPA DIO XIN W O RK SH O P QUESTIONS TO GUIDE PANEL DISCUSSIONS SESSION 1 Dose M etric C o n sid erin g all o f th e en d p o in ts o r targ e t tissu es, and sp ecies th a t U .S . E n v iro n m en tal P ro tectio n A g en cy (U .S . E P A )'s d o se-resp o n se m o d elin g m ig h t evaluate, w h a t are th e b e st m easu res o f do se (e.g., in g ested , tissu e co n cen tratio n s, b o d y b u rd en , re cep to r o ccu p an cy , o th er su rro g ate) and w hy? Developing Dose-Response M odels from M am m alian Bioassays H o w b est can th e p o in t o f d ep artu re (P O D ) b e d eterm in ed w h en th e resp o n se ran g e is in co m p letely ch aracterized (i.e., h ig h resp o n se at th e lo w est d o se o r lo w resp o n se at th e h ig h est dose; o b serv ed in several key 2 ,3 ,7 ,8 -T etrach lo ro d ib en zo -p -D io x in [T C D D ] studies)? If considered to be b iologically plausible, h o w can a th resh o ld be inco rp o rated into a dosere sp o n se fu n c tio n (e.g., fo r T C D D c a n c e r data)? H o w can n o n m o n o to n ic resp o n ses b e in co rp o rated into th e d o se-resp o n se function? Developing Dose-Response M odels from Epidem iological Studies H o w can th e epidem iological data be u tilized b est to inform th e T C D D exposure-response m odeling? W h ich epid em io lo g ical studies are m o st relevant? Supporting Inform ation F o r those to xicological endpoints th at are A h receptor-m ediated, h o w w o u ld th e recep to r kinetics in flu en ce the shape o f the d o se-resp o n se curve? H o w w o u ld d o w n stream cellu lar events affect th e shape o f th e d ose-response curve? H o w can th is cascade o f cellular events b e in corporated into a quantitative m odel o f dose-response? This document is a draftfor review purposes only and does not constitute Agency policy. A-31 DRAFT--DO NOT CITE OR QUOTE SESSIONS 2, 3A, 3B, 4A, AND 4B K ey Study Selection F o r this endpoint, w h at refinem ents should be m ad e to th e d raft criteria fo r selection o f key stu d ie s? W h at are th e specific effects o f concern fo r h u m an h ealth fo r this en d p o in t? B ased on th e draft criteria fo r th e selection o f key studies, w h at are th e key studies in fo rm in g the shape o f the dose-response curve above the P O D and the choice o f the P O D for this endpoint? Epidem iological Study Utility H o w and to w h at ex ten t do th e epid em io lo g ical data in form th e choice o f critical effect? H o w can the epidem iological data inform the quantitative dose-response m odeling? Anim al M odel Utility A re there types o f effects observed in anim al m odels th at are m ore relev an t to hum ans than others? T o w h at extent does inform ation on m ode o f action (M O A ) influence th e choice o f anim al m odel (species, strain, sex)? Supporting Inform ation A re there studies th at establish a sufficient ju stificatio n fo r departure from the d efault procedures th at address the shape o f the dose-response curve b elo w the P O D u n d er the cancer guidelines? A re th ere studies th at estab lish a su fficien t ju stific atio n fo r d ep artin g from U .S. E P A 's defau lt approaches fo r noncancer endpoints? T o w h at extent can M O A inform ation clarify the identification o f endpoints o f concern and doseresponse m etric fo r th is endpoint? H o w can th e cascade o f cellular events fo r this endpoint be incorporated into a quantitative m odel o f dose response? This document is a draftfor review purposes only and does not constitute Agency policy. A-32 DRAFT--DO NOT CITE OR QUOTE SESSION 5 F o r can cer and n o n can cer T C D D d o se-resp o n se assessm en ts, U .S . E P A is in terested in d eveloping a q u an titativ e u n certain ty analysis addressing b o th p aram eter and m o d el uncertainty, if feasible. U n certain ties w ill include, am ong others, ch o ice o f endpoint; u n d erly in g study u n certainties; choice o f dose m etric; interspecies extrapolations such as k inetic u n certainties; and ch o ice o f d o se-resp o n se m odel, in clu d in g th resh o ld m odels. T h e U .S. E P A is curren tly exam ining tech n iq u es and tools fo r uncertainty analysis-- including B ay esian and frequentist approaches. Identification of Im portant Uncertainties W h at are th e m ajo r un certain ties p ertaining to m odeling th e anim al data? C o n sid er th e dose m etric (species or tissu e specificity), v eh icle o f adm inistration, ex p o su re freq u en cy , ex p o su re d u ratio n , an d P O D d eterm in atio n (e.g., b en ch m ark response selection or no -o b serv ed -ad v erse-effect lev el/lo w est-o b serv ed -ad v erse-effect level identification). W h at are the m ajo r un certain ties pertain in g to d ose-response m odeling b elo w the P O D ? C o n sid er h o w recep to r kinetics and dow nstream cellu lar event inform ation m ig h t b e used to b o u n d th e u n certain ties associated w ith dose-response m odeling b elo w the PO D . W h a t are th e m a jo r u n c e rta in tie s in cro ss-sp ecies e x tra p o latio n (e.g., h alf-liv e s, tissu e distribution, and to x ico d y n am ics)? C o n sid er th e p rim ary species d o sed w ith T C D D : m ice, h am sters, rats, g u in ea pigs, and m onkeys. W h at are th e m ajo r un certain ties pertain in g to in trah u m an v ariab ility ? C o n sid er w h at d ata sets w o u ld b e usefu l to rep resen t sen sitiv e subpopulations. W h at are o th er significant sources o f uncertainty fo r th e can cer and n o n can cer assessm ents? Considerations for Conducting Uncertainty Analysis W h at d ata sets co u ld b e u sed to q u antify u n certain ties in can cer and n o n can cer T C D D doseresp o n se assessm ents? C o n sid er d io x in -lik e co m p o u n d d o se-resp o n se data. C o n sid er M O A inform ation. W h at are th e app ro p riate tech n iq u es fo r th e T C D D d o se-resp o n se u n certain ty analysis, and w h at are th eir respective strengths and w eak n esses o f these approaches as ap plied to T C D D ? This document is a draftfor review purposes only and does not constitute Agency policy. A-33 DRAFT--DO NOT CITE OR QUOTE A P PE N D IX C: 2009 U.S. E PA D IO X IN W O R K SH O P D R A FT SE L E C T IO N C R IT E R IA TO ID E N T IFY K E Y IN VIVO M A M M A L IA N ST U D IE S T H A T IN F O R M D O SE -R E SP O N S E M O D E L IN G F O R 2,3,7,8-T E T R A C H L O R O D IB E N Z O -p-D IO X IN (T C D D )a This document is a draftfor review purposes only and does not constitute Agency policy. Study Feature Chemical, purity, matrix/medium Peer review P rim ary b TCDD-only doses included, purity specified, matrix in which TCDD is administered is identified Independently peer-reviewed, publicly available Study design, execution, and reporting Study subject: species, strain, and sensitivity for given endpoint; litter; life stage; gender Exposure route Clearly documented and consistent with standard toxicological principles, testing protocols, and practice (i.e., endpoint-appropriate, particularly for negative findings) Mammalian species Strain and gender identified Animal age at beginning of treatment identified Litter confounders (within/between) accounted for Oral Dose level Exposure frequency, duration, and timing Controls Response Statistical evaluation Lowest dose <200 ng/kg-d for noncancer endpoints and <1 pg/kg-d for cancer Dosing regimen characterized and explained Appropriate and well characterized Effect relevant to human health Magnitude outside range of normal variability Clearly described and appropriate to the endpoint and study design (e.g., per error variance, magnitude of effect) Selection Rationale Secondaryc TCDD purity or matrix not clearly identified Supplementary materials accompanying peer-reviewed publication Testing protocol provides incomplete coverage of relevant endpoint-specific measures, particularly for negative findings Currently Excluded Studies of dioxin-like compounds (DLCs) or mixtures Not formally peer-reviewed; literature not publicly available Studies not meeting standard principles and practices Mammalian species, in vivo, but only studying an artificially sensitive subject (e.g., knockout mouse) Non-mammalian or not in vivo Parenteral (e.g., intravenous, intramuscular, intraperitoneal, subcutaneous) Lowest dose >200 ng/kg-d for noncancer endpoints, or >1.0 pg/kg-d for cancer Effect reported, but with no negative control Precursor effects, or adaptive responses potentially relevant to human health Limited statistical context Inhalation, dermal, ocular Characterization/explanation missing or cannot be determined Lethality A -34 D R A FT: D O N O T C IT E O R Q U O T E a NAS (2006) commented that the selection of data sets for quantitative dose-response modeling needed to be more transparent. These draft criteria are offered for consideration at the kickoff workshop. These criteria would be used to identify candidate studies of non-human mammals that would be used to define the point-of-departure (POD). These criteria are not designed for hazard identification or weight-of-evidence determinations. Studies addressing data other than direct TCDD dose-response in mammals (including toxicokinetic data on absorption, distribution, metabolism, or elimination; information on physiologically-based pharmacokinetic [PBPK] modeling, and mode of action data) will be evaluated separately. b Presents preliminary draft criteria for evaluating a study being considered for estimating a POD in a TCDD dose-response model. c Presents preliminary draft criteria that could qualify a study as primary with support from other lines of evidence (e.g., PBPK modeling), when no study for an endpoint meets the "primary" criteria. [T his p ag e in ten tio n ally left blank.] DRAFT D O N O T C IT E O R Q U O T E M ay 2010 E xternal R ev iew D raft APPENDIX B Evaluation of Cancer and Noncancer Epidemiological Studies for Inclusion in TCDD Dose-Response Assessment N O T IC E T H IS D O C U M E N T IS A N E X T E R N A L R E V IE W D R A F T . It has n o t been form ally released by the U .S. E n v iro n m en tal P ro tectio n A g en cy and should n o t at th is stage b e co n stru ed to rep resen t A gency policy. It is b ein g circu lated fo r co m m en t on its technical accuracy and policy im plications. N ational C en ter fo r E n v iro n m en tal A ssessm ent O ffice o f R esearch and D evelopm ent U .S. E n v iro n m en tal P ro tectio n A g en cy C incinnati, O H CONTENTS--Appendix B: Evaluation of Cancer and Noncancer Epidemiological Studies for Inclusion in TCDD Dose-Response Assessment L I S T O F T A B L E S ....................................................................................................................................................................... B - i i i A PPE N D IX B. E V A L U A T IO N OF C A N C E R A N D N O N C A N C E R E P ID E M IO L O G IC A L STU D IE S F O R IN C L U S IO N IN T C D D D O S E - R E S P O N S E A S S E S S M E N T ...............................................................................................B -1 B .1 . E V A L U A T I O N O F C A N C E R S T U D I E S ........................................................................................B -1 B .1 .1 . N I O S H C o h o r t S t u d i e s ..............................................................................................................B -1 B .1 .2 . B A S F C o h o r t S t u d i e s .................................................................................................................. B - 8 B .1 .3 . T h e H a m b u r g C o h o r t ................................................................................................................B - 1 1 B .1 .4 . T h e S e v e s o C o h o r t S t u d i e s ...................................................................................................B - 1 6 B .1 .5 . T h e C h a p a e v s k S t u d y .............................................................................................................. B - 2 2 B .1 .6 . T h e A i r F o r c e H e a l t h ( " R a n c h H a n d s " ) S t u d y ........................................................B - 2 3 B .1 .7 . O t h e r S tu d i e s o f P o t e n t i a l R e l e v a n c e t o D o s e - R e s p o n s e M o d e l i n g ..........B - 2 6 B .2 . E V A L U A T I O N O F N O N C A N C E R S T U D I E S ........................................................................B - 3 1 B .2 .1 . N I O S H C o h o r t ............................................................................................................................. B - 3 1 B .2 .2 . B A S F C o h o r t ................................................................................................................................ B - 3 3 B .2 .3 . H a m b u r g C o h o r t ......................................................................................................................... B - 3 5 B .2 .4 . T h e S e v e s o W o m e n ' s H e a l t h S t u d y ...............................................................................B - 3 6 B .2 .5 . O t h e r S e v e s o N o n c a n c e r S t u d i e s ......................................................................................B - 4 5 B .2 .6 . C h a p a e v s k S t u d y ..........................................................................................................................B - 5 5 B .2 .7 . A i r F o r c e H e a l t h ( " R a n c h H a n d s " ) S t u d y ..................................................................B - 5 6 B .2 .8 . O t h e r N o n c a n c e r S tu d i e s o f D i o x i n ................................................................................ B - 5 7 B .3 . R E F E R E N C E S ............................................................................................................................................... B - 6 1 This document is a draftfor review purposes only and does not constitute Agency policy. B-ii DRAFT--DO NOT CITE OR QUOTE LIST OF TABLES B -1. F i n g e r h u t e t a l., 1 9 9 1 -- A ll c a n c e r s ite s , s i t e - s p e c i f i c a n a l y s i s ...................................................B -1 B -2. S te e n l a n d e t a l., 1 9 9 9 -- A ll c a n c e r s ite s c o m b i n e d ,s i t e - s p e c i f i c a n a l y s i s .............................. B - 2 B -3. S t e e n l a n d e t a l., 2 0 0 1 -- A ll c a n c e r s ite s c o m b i n e d ............................................................................... B - 4 B -4. C h e n g e t a l., 2 0 0 6 -- A ll c a n c e r s ite s c o m b i n e d ...................................................................................... B - 5 B -5. C o l l i n s e t a l., 2 0 0 9 -- A ll c a n c e r s i t e s c o m b i n e d , s i t e - s p e c i f i c a n a l y s i s ...................................B - 6 B -6. Z o b e r e t a l., 1 9 9 0 -- A ll c a n c e r s ite s c o m b i n e d , s i t e - s p e c i f i c a n a l y s i s ..................................... B - 8 B -7. O t t a n d Z o b e r , 1 9 9 6 -- A ll c a n c e r s i t e s c o m b i n e d ..................................................................................B - 9 B -8. M a n z e t a l., 1 9 9 1 -- A ll c a n c e r s i t e s c o m b i n e d , s i t e - s p e c i f i c a n a l y s e s ................................. B - 1 1 B -9. F le sc h -Ja n y s e t al., 1995; F le sc h -Ja n y s e t al., 1996 erratu m -- A ll c a n c e r sites c o m b i n e d .................................................................................................................................................................... B - 1 2 B - 1 0 . F l e s c h - J a n y s e t a l., 1 9 9 8 -- A ll c a n c e r s ite s c o m b i n e d , s i t e - s p e c i f i c a n a l y s i s ................B - 1 3 B - 1 1 . B e c h e r e t a l., 1 9 9 8 -- A ll c a n c e r s ite s c o m b i n e d ..................................................................................B - 1 5 B - 1 2 . B e r t a z z i e t a l., 2 0 0 1 -- A ll c a n c e r s i t e s c o m b i n e d , s i t e - s p e c i f i c a n a l y s e s ........................... B - 1 6 B - 1 3 . P e s a t o r i e t a l., 2 0 0 3 -- A ll c a n c e r s ite s c o m b i n e d , s i t e - s p e c i f i c a n a l y s e s .............................B - 1 7 B - 1 4 . C o n s o n n i e t a l., 2 0 0 8 -- A ll c a n c e r s i t e s c o m b i n e d , s i t e - s p e c i f i c a n a l y s e s .........................B - 1 8 B - 1 5 . B a c c a r e l l i e t a l., 2 0 0 6 -- S it e - s p e c i f ic a n a l y s i s ......................................................................................B - 2 0 B - 1 6 . W a r n e r e t a l., 2 0 0 2 -- B r e a s t c a n c e r i n c i d e n c e ...................................................................................... B - 2 1 B - 1 7 . R e v i c h e t a l., 2 0 0 1 -- A ll c a n c e r s i t e s c o m b i n e d , a n d s i t e - s p e c i f i c a n a l y s e s .................... B - 2 2 B - 1 8 . A k h t a r e t a l., 2 0 0 4 -- A ll c a n c e r s i t e s c o m b i n e d a n d s i t e - s p e c i f i c a n a l y s e s ....................... B - 2 3 B - 1 9 . M i c h a l e k a n d P a v u k , 2 0 0 8 -- A ll c a n c e r s ite s c o m b i n e d ............................................................... B - 2 5 B - 2 0 . `t M a n n e t j e e t a l., 2 0 0 5 -- A ll c a n c e r s ite s c o m b i n e d , s ite s p e c if ic a n a l y s e s .....................B - 2 6 B - 2 1 . M c B r i d e e t a l., 2 0 0 9 b -- A ll c a n c e r s i t e s c o m b i n e d , s i t e - s p e c i f i c a n a l y s i s .........................B - 2 7 B - 2 2 . M c B r i d e e t a l., 2 0 0 9 a -- A ll c a n c e r s ite s c o m b i n e d , s i t e - s p e c i f i c a n a l y s i s .........................B - 2 8 B - 2 3 . H o o i v e l d e t a l., 1 9 9 8 -- A ll c a n c e r s ite s c o m b i n e d , s i t e - s p e c i f i c a n a l y s i s ..........................B - 2 9 B - 2 4 . S t e e n l a n d e t a l., 1 9 9 9 -- M o r t a l i t y ( n o n c a n c e r ) ..................................................................................... B - 3 1 B -25. C o l l i n s e t a l., 2 0 0 9 -- M o r t a l i t y ( n o n c a n c e r ) .......................................................................................... B - 3 2 This document is a draftfor review purposes only and does not constitute Agency policy. B -iii D R A FT -- D O N O T C IT E O R Q U O T E LIST OF TABLES (continued) B - 2 6 . O t t a n d Z o b e r , 1 9 9 6 -- M o r t a l i t y ( n o n c a n c e r ) .......................................................................................B - 3 3 B -27. F le sc h -Ja n y s e t al., 1995; F le sc h -Ja n y s et al., 1996 erratu m -- M o rtality ( n o n c a n c e r ) ...............................................................................................................................................................B - 3 5 B - 2 8 . E s k e n a z i e t a l., 2 0 0 2 a -- M e n s t r u a l c y c le c h a r a c t e r i s t i c s .............................................................. B - 3 6 B - 2 9 . E s k e n a z i e t a l., 2 0 0 2 b -- E n d o m e t r i o s i s ....................................................................................................B - 3 8 B - 3 0 . E s k e n a z i e t a l., 2 0 0 3 -- B i r t h o u t c o m e s ..................................................................................................... B - 3 9 B - 3 1 . W a r n e r e t a l., 2 0 0 4 -- A g e a t m e n a r c h e ..................................................................................................... B - 4 0 B - 3 2 . E s k e n a z i e t a l., 2 0 0 5 -- A g e a t m e n o p a u s e ..............................................................................................B - 4 1 B - 3 3 . W a r n e r e t a l., 2 0 0 7 -- O v a r i a n f u n c t i o n ..................................................................................................... B - 4 3 B - 3 4 . E s k e n a z i e t a l., 2 0 0 7 -- U t e r i n e l e i o m y o m a ........................................................................................... B - 4 4 B - 3 5 . M o c a r e l l i e t a l., 2 0 0 8 -- S e m e n q u a l i t y ..................................................................................................... B - 4 5 B - 3 6 . M o c a r e l l i e t a l., 2 0 0 0 -- S e x r a t i o ..................................................................................................................B - 4 6 B - 3 7 . B a c c a r e l l i e t a l., 2 0 0 8 -- N e o n a t a l t h y r o i d f u n c t i o n .......................................................................... B - 4 8 B - 3 8 . A l a l u u s u a e t a l., 2 0 0 4 -- O r a l h y g i e n e ........................................................................................................ B - 4 9 B - 3 9 . B e r t a z z i e t a l., 2 0 0 1 -- M o r t a l i t y ( n o n c a n c e r ) .......................................................................................B - 5 0 B - 4 0 . C o n s o n n i e t a l., 2 0 0 8 -- M o r t a l i t y ( n o n c a n c e r ) .................................................................................... B - 5 2 B - 4 1 . B a c c a r e l l i e t a l., 2 0 0 5 -- C h l o r a c n e ..............................................................................................................B - 5 3 B - 4 2 . B a c c a r e l l i e t a l, 2 0 0 2 a n d 2 0 0 4 -- I m m u n o l o g i c a l e f f e c t s ...............................................................B - 5 4 B - 4 3 . R e v i c h e t a l., 2 0 0 1 -- M o r t a l i t y ( n o n c a n c e r ) a n d r e p r o d u c t i v e h e a l t h ....................................B - 5 5 B - 4 4 . M i c h a l e k a n d P a v u k , 2 0 0 8 -- D i a b e t e s .......................................................................................................B - 5 6 B - 4 5 . M c B r i d e e t a l., 2 0 0 9 a -- M o r t a l i t y ( n o n c a n c e r ) .................................................................................... B - 5 7 B - 4 6 . M c B r i d e e t a l., 2 0 0 9 b -- M o r t a l i t y ( n o n c a n c e r ) .................................................................................. B - 5 9 B - 4 7 . R y a n e t a l., 2 0 0 2 -- S e x r a t i o ............................................................................................................................ B - 6 0 This document is a draftfor review purposes only and does not constitute Agency policy. B-iv DRAFT--DO NOT CITE OR QUOTE 1 APPENDIX B. EVALUATION OF CANCER AND NONCANCER 2 EPIDEM IOLOGICAL STUDIES FOR INCLUSION IN TCDD 3 DOSE-RESPONSE ASSESSM ENT 4 5 6 B.1. EV ALUA TIO N OF C A N C ER STUDIES 7 B.1.1. NIO SH Cohort Studies 8 9 Table B-1. Fingerhut et al., 1991-- All cancer sites, site-specific analysis 10 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. The data sources to ascertain vital status and cause o f death inform ation were the Social Security death files, the National D eath Index, and the Internal Revenue Service. Vital status could be determ ined for 98% of the cohort. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration not satisfied. W hile the authors provide compelling arguments that suggest risks are not unduly biased by lack o f cigarette smoking data, they acknowledge potential biases that could exist for other occupational exposure (e.g., asbestos) for w hich data were lacking. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response 4. Consideration Consideration not satisfied. There was not a statistically significant linear trend o f increasing mortality w ith increased duration o f exposure. Exposure assessm ent methodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration not satisfied. This study used duration o f exposure, at an individual level, as a surrogate measure of TCDD. Duration o f exposure determined by num ber of years workers were involved in processes involving TCDD contamination. Exposure was determined by reviewing, at each plant, operating conditions, job duties, records o f TCDD levels in industrial hygiene samples, intermediate reactants, products, and wastes. Exposure assessm ent was lim ited and the uncertainty related to exposure measures not fully addressed. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. This is the largest o f the occupational cohorts that has been exposed to TCDD. The cohort consisted o f 5,172 w orkers and a total o f 265 cancer deaths. Sitespecific mortality analyses, including soft tissue sarcom a (n = 4), was lim ited by small numbers. 1. C riteria Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. This document is a draftfor review purposes only and does not constitute Agency policy. B-1 DRAFT--DO NOT CITE OR QUOTE Response 2. Criteria Response 3. C riteria Response Criteria satisfied. New England Journal o f M edicine, 1991; 324:212-218. Authors address the possibility o f bias from lack o f control for potential confounders such as smoking and other occupational exposures. They address limitations of using death certificates for identifying certain causes o f deaths, and limitations o f using duration o f em ploym ent as an exposure metric. Exposure must be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria not satisfied. Since this study used duration o f exposure as the exposure metric, dose-response relationships cannot be quantified. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose-is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria satisfied. M odels incorporated period o f latency, and a surrogate measure of cumulative TCDD exposure was modeled. The follow-up interval was sufficiently long (1942-1987). Conclusion Overall, quantitative exposure data are lacking on an individual-level basis. Further dose-response analysis should consider updated data for this cohort that includes serum-based measures o f TCDD, in addition to an extension o f the follow-up period. Given these limitations, this study is not further evaluated fo r TCDD dose-response assessment. 1 2 3 Table B-2. Steenland et al., 1999-- All cancer sites com bined, site-specific 4 analysis 5 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. The study evaluated mortality from all cancer sites (combined). As described in the paper, the sources o f vital status and cause of death information were received from the Social Security death files, the National D eath Index, and the Internal Revenue Service. Vital status was know n for 99.4% o f the cohort members, cause o f death information is available for 98% o f the decedents. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Occupational exposure to asbestos and 4-aminobiphenyl contributed to some excess cancer, but no evidence o f confounding for the relationship betw een TCDD and all cancer mortality was detected following removal of workers who died of bladder cancer. No inform ation is available for cigarette smoking, although dose-response patterns were stronger for nonsm oking related cancers. This finding suggests that smoking is not responsible for excess cancer risk that was observed in the cohort. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. This document is a draftfor review purposes only and does not constitute Agency policy. B-2 DRAFT--DO NOT CITE OR QUOTE Response 4. Consideration Response 5. C onsideration Response Consideration satisfied. W hen a 15-year lag interval was incorporated into the exposure m etric a statistically significant dose-response pattern was observed for all cancer sites com bined w ith both a continuous measure o f TCDD (p = 0.05) as well as one that was log-transform ed (p < 0.001). Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Consideration satisfied. The study conducted detailed sensitivity analyses and evaluated different assumptions regarding latency, log-transformed TCDD exposures, and half-life values for TCDD. Study size and follow-up are large enough to yield precise estim ates of risk and ensure adequate statistical power. Consideration satisfied. This is the largest o f the occupational cohorts w ith exposures to TCDD. The cohort consisted o f 5,132 male workers and a total o f 377 cancer deaths. This permits characterization o f risk for all cancer sites (combined). 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Journal o f the National Cancer Institute, 1999; 91(9):779-786. The authors discussed the potential for bias from smoking, and other occupational exposures for which data for both were lacking at an individual basis. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. Exposure scores assigned on an individual level using a job-exposure matrix. The job-exposure matrix was based on estimated factor of contact w ith TCDD in each job, level o f TCCD contamination o f materials at each plant over time, and proportion of day w orker could be in contact w ith materials. These factors were multiplied together to derive a daily exposure score, w hich was accum ulated over the working history o f each w orker to obtain a cumulative measure of TCDD. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria satisfied. The follow -up o f the cohort extended from 1942 until the end o f 1993. Greater than 25 years o f follow-up have accrued in cohort allowing for latency to be examined. Different assumptions on the half-life o f TCDD were evaluated and produced sim ilar results. Latency intervals were incorporated, w ith strongest associations noted w ith an interval o f 15 years. Conclusion 1 2 3 4 5 This study meets the criteria and considerations noted above but has been superseded and updated by Steenland et al. (2001). Therefore, this study was not considered for further dose-response analyses. This document is a draftfor review purposes only and does not constitute Agency policy. B-3 DRAFT--DO NOT CITE OR QUOTE 1 Table B-3. Steenland et al., 2001-- All cancer sites com bined 2 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. The study evaluated mortality from all cancer sites (combined). As described by Steenland et al., (1999) the sources o f vital status and cause o f death inform ation were received from the Social Security death files, the National D eath Index, and the Internal Revenue Service. Vital status was know n for 99.4% o f the cohort members, cause o f death inform ation is available for 98% o f the decedents. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Occupational exposure to asbestos and 4-aminobiphenyl contributed to some excess cancer, but no evidence o f confounding for the relationship betw een TCDD and all cancer mortality was detected following removal of workers who died of bladder cancer. No inform ation is available for cigarette smoking, although dose-response patterns were similar betw een smoking and nonsmoking related cancers. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. Increased risk estimates were observed in the higher cumulative exposure categories. The dose-response curve was not linear at higher doses. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. Exposure metrics considered included cumulative TCDD, log10TCDD, average exposure, and a cubic spline model was also evaluated. Exposure response relationships were also evaluated using TEQs. Exposure scores were assigned on an individual level using a job-exposure matrix. The job-exposure matrix was based on estimated factor of contact w ith TCDD in each job, level o f TCCD contamination o f materials at each plant over time, and proportion of day w orker could be in contact w ith materials. Serum levels were measured in 199 workers at one o f 8 plants in 1998. D ifferent estimate o f the half-life o f TCDD were used, and similar results were produced. The paper presented a range in risk estimates thereby conveying the range of uncertainties in risk estimates derived using different measures of exposure. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. This is the largest o f the occupational cohorts w ith exposures to TCDD. The cohort consisted o f 3,538 male workers and a total o f 256 cancer deaths. 1. C riteria Response 2. Criteria Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied Am J Epidem, 2001, 154(5):451-458. However, additional details to assess uncertainties associated w ith characterizing serum data in a subset o f workers to rem ainder of cohort are lacking. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. This document is a draftfor review purposes only and does not constitute Agency policy. B-4 DRAFT--DO NOT CITE OR QUOTE Response 3. C riteria Response Criteria satisfied. The metrics considered included cumulative TCDD, loglOTCDD, average exposure, and a cubic spline model was also evaluated. Exposure response relationships were also evaluated using TEQs. Serum lipid TCDD m easurem ents from 170 w orkers whose TC D D levels w ere greater th an 10 p p t (the u p p er ranges o f a back g ro u n d level) w ere used along w ith JEM information, w ork histories, and a pharmacokinetic elim ination model to estimate dose rates per unit exposure score. In this regression model, the estimated TCDD level at the time of last exposure was modeled as a function o f exposure scores. The coefficient relating serum levels and exposure scores was then used to estimate serum TCDD levels over time from occupational exposure (minus the background level) for all 3,538 workers. Tim e-specific serum levels were then integrated over time to derive a cumulative serum lipid concentration due to occupational exposure for each worker. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria satisfied. Greater than 25 years o f follow-up have accrued in cohort allowing for latency to be examined. Different assumptions on the half-life o f TCDD were evaluated producing sim ilar results. Conclusion Overall, criteria have b een satisfied. This study w as m odeled in the 2003 R eassessm ent and is considered for further dose-response evaluations herein. 1 2 3 Table B-4. Cheng et al., 2006-- All cancer sites com bined 4 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. The study evaluated cancer mortality. The vital status and the inform ation regarding the cause o f death were extracted from the Social Security death files, the National D eath Index, and the Internal Revenue Service (Steenland et al., 1999). Vital status w as know n fo r 99.4% o f the cohort m em bers, w hile cause o f death inform ation is available for 98% o f the decedents. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. This is the same data set used in the Steenland et al., (2001) paper. Occupational exposure to asbestos and 4-am inobiphenyl contributed to some excess cancer, but no evidence o f confounding for the relationship betw een TCDD and all cancer mortality was detected following removal of workers who died of bladder cancer. No inform ation is available for cigarette smoking, although dose-response patterns were similar between smoking and nonsmoking related cancers. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. Slope coefficients are available for all cancers com bined under a varying set of assumptions. Little evidence o f an association was found w hen lag interval was not taken into account. A ssociations strengthened w ith incorporation o f a 10 to 15 y ear lag interval. Dose-response was nonlinear at higher exposures, suggesting a nonlinear relationship or increased exposure m isclassification at higher levels. This document is a draftfor review purposes only and does not constitute Agency policy. B-5 DRAFT--DO NOT CITE OR QUOTE 4. Consideration Response 5. C onsideration Response Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. C onsideration satisfied. C om pared to the 1st order m odels, the concentration, and age dependent m odel (CADM ) provided a better fit for the serum sam pling data. CADM model exposure estim ates are higher than those based on an age only, constant 8.7-year half-life model. As discussed by A ylward et al. (2005b), m odel exposure estim ates are influenced not only by choice o f elim ination model, but also by choices in regression procedure (e.g., log transform ation, use o f intercept, and incorporation o f background dose term). Other limitations or uncertainties in exposure assessment include the following Job-exposure matrix based on limited sampling data, and subjective judgm ent on contact times and factors Inability to take into account interindividual variability in TCDD elim ination kinetics Dose-rate regressions are based on a small sample o f the cohort w ith serum measures; therefore, regression results may not be representative o f rem ainder o f the cohort. Study size and follow-up are large enough to yield precise estim ates of risk and ensure adequate statistical power. Consideration satisfied. Largest cohort o f TCDD exposed workers. The risk estimates are based on a total o f 256 cancer deaths. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Risk Analysis, 2006; 4:1,059-1,071. Additional details to assess uncertainties associated w ith characterizing serum data can be found in A ylw ard et al. (2005b); R isk Anal. 25(4):945-956. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. Cumulative serum lipid concentrations were estim ated for each worker. No other dioxin-like compounds were assessed in this analysis. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria satisfied. Concentration and age-dependence o f TCDD elim ination and two compartments (hepatic and adipose tissue) were taken into account w hen estimating TCDD exposures. Nearly 50 years o f follow-up were available permitting an evaluation of latency. Conclusion This study m et the m ain criteria and considerations. The study is considered for further dose-response analyses. 1 2 3 Table B-5. Collins et al., 2009-- All cancer sites com bined, site-specific 4 analysis 5 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. This document is a draftfor review purposes only and does not constitute Agency policy. B-6 DRAFT--DO NOT CITE OR QUOTE Response Consideration satisfied. Vital status complete for all but two workers. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. No inform ation collected on smoking status, but no excess in lung cancer or nonm alignant respiratory diseases noted. Analyses took into account potential for exposure to pentachlorophenol. 3. C onsideration Study dem onstrates an association b etw een T C D D an d adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. No dose-response pattern was observed w ith all cancer sites combined, however, a dose-response pattern was observed w ith soft tissue sarcoma. The study found no association between TCDD and death from m ost types of cancer. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. The authors used these serum from 280 form er TCP workers to estimate historical exposure levels of TCDD, furans, and polychlorinated biphenyls for all 1,615 workers. Exposure assessm ent included detailed w ork history, industrial hygiene monitoring, and the presence of chloracne cases among groups of workers. This data was integrated into a 1-compartment, first-order pharmacokinetic to determine the average TCDD dose associated w ith jobs in each group, after accounting for the presence of background exposures estim ated from the residual serum TCDD concentration in the sampled individuals. The authors did not evaluate departures from linearity, or examine skewness at higher exposures. Exposure levels were not provided. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk an d ensure adequate statistical power. Response Consideration satisfied. Largest study o f workers em ployed in one center, and a total of 177 deaths from cancer were observed. Lim ited precision in the relative risk estimate was noted for soft tissue sarcoma and TCDD exposures. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Published in Am J Epidemiol, 2009, 170(4):501-506. The authors discuss limitations o f using death certificates for identifying deaths from soft tissue sarcoma for which a positive association was noted, assumptions in exposure characterization, and effects of cigarette smoking. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. This study has the largest num ber o f serum samples obtained from a specific plant. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria satisfied. A lthough specific analyses o f latency were not reported, this cohort had a sufficient length o f follow-up for cancer mortality outcomes. This document is a draftfor review purposes only and does not constitute Agency policy. B-7 DRAFT--DO NOT CITE OR QUOTE Conclusion The authors found a statistically significant dose-response trend for soft tissue sarcoma mortality and TCDD exposures. The all-tum or results are not amenable to dose-response analysis because they found no effect. Therefore, this study is considered for quantitative dose-response analysis for the soft tissue sarcom a mortality results, only. 1 2 B.1.2. BASF Cohort Studies 3 4 Table B-6. Zober et al., 1990-- All cancer sites com bined, site-specific 5 analysis 6 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. A large com ponent o f the cohort (94 out o f 247 workers) was assembled by actively seeking out workers who were alive in 1986 through the "D ioxin Investigation Program m e." As a result, it is likely a num ber o f deaths were m issed due to the recruitm ent o f survivors. This underascertainm ent is supported by m uch lower all cancer SM R one com ponent o f the cohort (SM R = 0.48, 95% CI: 0.13-1.23) relative to the general population. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. See above discussion of underascertainment in mortality for some of the cohort members. A lthough it is likely that other coexposures occurred (e.g., am ong firefighters), confounding could only occur if these coexposures were associated w ith both the endpoint and exposure (TCDD) being considered. 3. C onsideration Study dem onstrates an association b etw een TC D D and adverse h ealth effect w ith evidence o f an exposure-response relationship. Response Consideration not satisfied. W orkers were not categorized on the basis o f their exposure, but rather their mortality experience com pared to control cohort and the general population. The design o f the study does not allow for dose-response to be examined. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration not satisfied. Although years since first exposure was examined, exposure assessm ent was based on w orking in various occupational cohorts. Since there was no quantitative assignment of TCDD exposures, the associated uncertainties could not be evaluated. 5. C onsideration Study size and follow -up are large enough to y ield p recise estim ates o f risk an d ensure adequate statistical power. Response Consideration not satisfied. There were only 23 cancer deaths in the entire cohort. As such, this study lacked adequate statistical pow er to detect cancer mortality differences that were moderate in magnitude. 1. C riteria Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. This document is a draftfor review purposes only and does not constitute Agency policy. B-8 DRAFT--DO NOT CITE OR QUOTE Response 2. Criteria Response 3. C riteria Response Criteria satisfied. Int A rch Occup Envir Health, 1990, 62:139-157. The authors address issues related to the healthy w orker effect, m ultiple comparisons, smoking, and small size of the cohort. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria not satisfied. Risks were derived by com paring mortality rates o f the three cohort subsets relative to a control cohort and the general population by time since first exposure categories. W orkers were not assigned exposures. There were no quantitative estimates of TCDD exposure. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria not satisfied. W hile the study was able to indirectly look at variations in risk estimates related to latency by using tim e since exposure, there were no quantitative estim ates o f TCDD exposure. Conclusion This study is not suitable for dose-response analysis, as it failed the inclusion criteria. M ost notably, the lack o f exposure data does not permit the use of these data for a dose-response analysis. 1 2 3 Table B-7. Ott and Zober, 1996-- All cancer sites com bined 4 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. M ortality ascertainment appeared to be fairly complete. The ascertainm ent o f cancer incidence is more difficult to judge as geographical area not covered by a cancer registry. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Inform ation was collected on smoking status, body mass index, and other occupational exposures, however a large portion of the cohort was firefighters who may have been exposed to other occupational carcinogens. However, the recruitment of survivors may results in under-ascertainm ent o f mortality. 3. C onsideration Study dem onstrates an association b etw een TC D D and adverse h ealth effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. Increased cancer incidence was observed in the highest TCDD cumulative exposure category. Risks were most pronounced w hen a period of 20 years since first exposure was incorporated into the model. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. This document is a draftfor review purposes only and does not constitute Agency policy. B-9 DRAFT--DO NOT CITE OR QUOTE Response 5. C onsideration Response Consideration satisfied. Cumulative measure of TCDD expressed was derived from serum measures. Exposure was also estimated by chloracne status o f the cohort members. The authors have not addressed the potential im plication o f deriving TCDD exposure estimates for the whole cohort using sera data that were available for only about half of the cohort. Study size and follow -up are large enough to yield precise estim ates o f risk and ensure adequate statistical power. C onsideration satisfied. F or all cancer sites com bined, there w ere 31 deaths. It is the sm allest o f the occupational cohorts, but the deaths can be grouped into quartiles to allow for evaluation o f dose-response relationships. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Occupational and Environm ental M edicine, 1996, 53:606-612. A large com ponent o f the cohort (94 out of 247 workers) was assembled by actively seeking out workers who were alive in 1986 through the "D ioxin Investigation Program m e." As a result, it is likely a num ber o f deaths were m issed due to the recruitm ent o f survivors. This underascertainm ent is supported by m uch low er all cancer SM R one com ponent o f the cohort (SM R = 0.48, 95% CI: 0.13-1.23) relative to the general population (Zober et al., 1990). Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. Serum samples, taken in 1989, were available for 138 surviving workers out of 254 and allowed for cumulative TCDD levels to be estimated using regression techniques in the rem ainder o f the cohort. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria satisfied. Exposure assignm ent took into the affect that body mass index had on TCDD half-lives. TCDD levels estimates through back-extrapolation o f serum levels based on half-life estimates obtained from previous studies. Latency was considered w ith stronger association observed in external comparisons incorporating a latency o f 20 years. The follow-up o f the cohort was lengthy (>50 years). Conclusion Given a part o f the cohort was based solely on survivors in the in the mid-1980s, the SMR statistic derived from this study underestim ates excess mortality relative to the general population. The cohort also includes some firefighters who are recognized to be exposed to other carcinogenic agents-- these exposures may be confounding the associations that were reported. However, exposure to TCDD was quantified and the effective dose and oral exposure estimable. Overall, criteria have been satisfied. This study was m odeled in the 2003 Reassessm ent and is considered for further dose-response evaluations herein. 1 2 3 4 5 6 This document is a draftfor review purposes only and does not constitute Agency policy. B-10 DRAFT--DO NOT CITE OR QUOTE 1 B.1.3. The H am burg Cohort 2 3 Table B-8. M anz et al., 1991-- All cancer sites com bined, site-specific 4 analyses 5 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. Deaths were identified through medical records o f the cohort members. A review o f death certificates o f the identified cancer deaths found a high degree of concordance (51/54). One o f the 136 noncancer death certificates examined indicated an "occult" neoplasm. 2. Consideration Response 3. C onsideration Response 4. Consideration Response 5. C onsideration Response Risk estimates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Consideration satisfied. Smoking data were similar betw een exposed and nonexposed cohort based on independent samples. Occupational exposure for w hich individual data are lacking unlikely to explain dose-response w ith TCDD. Study dem onstrates an association betw een TCDD and adverse health effect w ith evidence of an exposure-response relationship. Consideration satisfied. Dose-response patterns across three levels o f exposure observed among those who started work before 1954, and among those who worked for 20 years or longer. Dose-response patterns not evident across whole cohort, among those w ith less than 20 years of employment, or among those who started after 1954. Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Consideration satisfied. Categorical exposures were based on TCDD concentrations in precursor materials, products, waste, and soil from the plant grounds, measured after the plant closed in 1984. Exposure uncertainty examined using a separate group o f 48 workers who provided adipose tissue samples. Other surrogate measures o f exposure were considered in this study, including duration o f exposure and year o f first employment. Study size and follow -up are large enough to yield precise estim ates o f risk and ensure adequate statistical power. Consideration satisfied. For all cancer sites combined, there were 65 cancer deaths for the com parison to the com parison cohort o f gas workers. The study is underpowered to look at site-specific cancers. 1. C riteria Response 2. Criteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Lance,t 1991, 338:959-964. The authors discussed potential for m isclassification using death certificates, healthy w orker effect and their related use o f a com parison cohort o f gas supply workers, other occupational exposures present at the plant, potential im pact and the lack o f sm oking data. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria not satisfied. Exposure consisted o f a large DLC com ponent that was not quantified. Given crude TCDD exposure categorization data, no quantitative exposure m etric was derived. This document is a draftfor review purposes only and does not constitute Agency policy. B-11 DRAFT--DO NOT CITE OR QUOTE 3. C riteria Response The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria satisfied. Exposure metrics were constructed that took into account duration of exposure, and periods w hen exposure was highest. However, exposure estimates did not consider lagged exposure. Conclusion This study is not amenable to further TCDD dose-response analysis and is not considered further here because it consisted of a large DLC component that was quantified and no quantitative exposure m etric was derived. The dose-response patterns o f risks observed across the three exposure groups provide compelling support for an association betw een TCDD and cancer mortality, particularly, given the associations observed w hen analyses restricted to those who were hired w hen TCDD exposures were know n to be m uch higher, and am ong those who worked for at least 20 years. Subsequent studies improved the exposure assessment through the use of serum measures. 1 2 3 Table B-9. Flesch-Janys et al., 1995; Flesch-Janys et al., 1996 erratum -- All 4 cancer sites com bined 5 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. M edical records used to identify deaths over the period 1952-1992. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Similarity in smoking rates betw een control cohort and the exposed workers was similar based on independent surveys. Occupational exposures to benzene, and dimethyl sulfate were unlikely to bias dose-response pattern observed as these exposures occurred in production departments w ith low-medium levels of exposure. 3. C onsideration Study dem onstrates an association b etw een TC D D and adverse h ealth effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. Dose-response relationship observed across 6 exposure categories, w ith the cohort o f gas supply workers used as the referent. 4. Consideration Consideration satisfied. Exposure assessm ent methodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessment are considered. Response The exposure measure was an integrated TCDD concentration over time estimate that back calculated TCDD exposures to the end of the employment. Categorical and continuous TCDD exposures were examined in relation to the health outcome. These efforts improve the exposure assessm ent o f earlier studies. 5. C onsideration Study size and follow -up are large enough to y ield p recise estim ates o f risk an d ensure adequate statistical power. Response Consideration satisfied. For all cancer sites combined, there were 124 deaths in the exposed cohort, and 283 in the cohort o f gas supply workers. No site-specific cancers were examined in this paper. This document is a draftfor review purposes only and does not constitute Agency policy. B-12 DRAFT--DO NOT CITE OR QUOTE 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. A m J Epidemiol, 1995, 1442:1165-1175. The authors discuss the potential role o f other occupational exposures (i.e., dim ethyl sulfate, solvents, and benzene), smoking, and suitability o f the com parison cohort o f gas supply workers. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. Serum and adipose tissues were used to estimate TCDD exposure in 190 workers. A one-com partm ent first-order kinetic model was used to estimate exposure at end o f exposure for these workers. Regression m ethods were then used to estim ates TCDD exposures for all workers. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria not satisfied. Exposure was based on half-life estimates from individuals w ith repeated serum measures. Other dioxin-like com pounds were considered w ith the TOTTEQ exposure metric. No consideration, however, was given to latency or lagged exposures. Conclusion The exposure data used w ithin this study are well-suited to a dose-response analysis given the associations observed, the characterization o f exposure using serum, and quality of ascertainm ent o f cancer outcomes. However, subsequent methods have been applied to the cohort to derive different exposures to TCDD using area under the curve approaches, w hich updates the analysis herein. Therefore, subsequent studies (i.e., B echer et al., 1998) will supersede this evaluation. 1 2 3 Table B-10. Flesch-Janys et al., 1998-- All cancer sites com bined, site4 specific analysis 5 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. M ortality follow-up was extended until the end o f 1992, an increase in 3 years from previous analyses o f the cohort. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Exposure was well characterized using sera data. W hile serum samples provided only from a subsample of surviving workers, these levels were consistent w ith expected levels in different production departments. The authors examined other potential occupational coexposures (e.g., p-hexachlorocyclohexane) and indirectly exam ined the potential effect o f smoking on the associations that were detected. 3. C onsideration Study dem onstrates an association b etw een TC D D and adverse h ealth effect w ith evidence o f an exposure-response relationship. This document is a draftfor review purposes only and does not constitute Agency policy. B-13 DRAFT--DO NOT CITE OR QUOTE Response 4. Consideration Response 5. C onsideration Response Consideration satisfied. A dose-response relationship across quartiles o f TCDD was observed w ith cancer m ortality based on the SM R statistic (SM Rs = 1.24, 1.34, 1.34, 1.73), and a linear test fo r trend w as statistically significant (p = 0.01). Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Consideration satisfied. The exposure measure was an integrated TCDD concentration over time estimate that back-calculated TCDD exposures to the end of the employment. Categorical and continuous TCDD exposures were examined in relation to the health outcome. These efforts im prove the exposure assessm ent o f earlier studies. Study size and follow -up are large enough to yield precise estim ates o f risk and ensure adequate statistical power. Consideration satisfied. F or all cancer sites combined, there were 124 cancer deaths. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Environ Health Perspect, 1998, 106(2):655-662. The authors address uncertainties in the estimation o f exposure, describe the potential for confounding from P-2,4,5-T, hexachlorocyclohexane, and cigarette smoking. In fact, they showed that blood levels o f TCDD were not associated w ith smoking in a subsample suggesting little bias from lack o f sm oking data. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. Serum samples, taken from 190 workers were used to derive TCDD levels for the entire cohort. M ethods used to estim ate exposure took into account elim ination of TCDD during em ploym ent periods w hen exposure took place, and the methods o f the area under the curve was used as it takes into account variations in concentration over time, and reflects cumulative exposure. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria satisfied. Exposure estim ated based on half-lives observed in individuals w ith repeated samples. Area under the curve approach was used w hich is an im provement from past characterizations o f exposure in this cohort. Conclusion 1 2 3 4 5 The study provides data suitable for dose-response modeling. D erivation o f exposure was done using current understanding o f elim ination o f TCDD. Estimates o f risks were derived from external com parisons to the general population that are unlikely to be biased by healthy w orker effect, but risks generated using internal cohort comparisons would be preferable. B echer et al., (1998) assessed this same data taking cancer latency into account, therefore Flesch-Janys et al., (1998) will not be further considered for dose-response modeling. This document is a draftfor review purposes only and does not constitute Agency policy. B-14 DRAFT--DO NOT CITE OR QUOTE 1 Table B-11. Becher et al., 1998-- All cancer sites com bined 2 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. M edical records used to identify deaths over the period 1952-1992. The follow-up interval was lengthy. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Risks adjusted for exposures to TEQ, p-hexachlorbenzene, and em ploym ent characteristics. Smoking was shown to be sim ilar to the com parison cohort of gas workers. 3. C onsideration Study dem onstrates an association b etw een TC D D and adverse h ealth effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. A variety o f exposure measures for both TCDD and TEQs found positive associations w ith cancer mortality. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. The exposure measure was an integrated TCDD concentration over time estimate that back-calculated TCDD exposures to the end of the employment. Categorical and continuous TCDD exposures were exam ined in relation to the health outcome. Different models explored the shape of the dose-response curve. These efforts improve the exposure assessm ent o f earlier studies. 5. C onsideration Study size and follow -up are large enough to y ield p recise estim ates o f risk an d ensure adequate statistical power. Response Consideration satisfied. F or all cancer sites combined, there were 124 cancer deaths. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Environ Health Perspect, 1998, 106(2):663-670. The authors discuss uncertainties associated w ith their use o f exposure metrics, inability to evaluate effects for PCDD/Fs other than dioxin due to high correlations w ith p-HCH, and inability to characterize risks associated w ith exposures in children. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. The authors derived a measure o f cumulative dose as a time-dependent variable ("area under curve") using serum m easures available in a sample o f 275 workers. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria satisfied. TCDD levels estimates through back-extrapolation o f serum levels based on half-life estimates obtained from previous studies. Latency was considered, and a variety of exposure metrics including nonlinear relationships were evaluated. This document is a draftfor review purposes only and does not constitute Agency policy. B-15 DRAFT--DO NOT CITE OR QUOTE Conclusion In this paper, a variety o f exposure metrics were found to be positively associated w ith cancer mortality. The additional lifetime risk o f cancer corresponded to a daily intake o f 1pg ranged betw een .01 and 0.001. This study w as m odeled in the 2003 R eassessm ent and is considered for further dose-response evaluations herein. 1 B.1.4. The Seveso Cohort Studies 2 3 Table B-12. Bertazzi et al., 2001-- All cancer sites com bined, site-specific 4 analyses 5 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. M ortality appears to be well captured from the vital statistics registries in the region (99% complete). Vital status was ascertained using similar methods for both the exposed and reference populations. Both cancer and noncancer mortality outcomes were evaluated. Ideally, would have evaluated incident rather than decedent outcomes for cancer. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Individual-level data on potential confounders (i.e., age, calendar period, and gender) were adjusted for. Inform ation from other independent surveys suggests similarity betw een sm oking behaviors across the regions. Com parison o f cancer mortality rates before the time o f the accident betw een the regions also revealed no differences. 3. C onsideration Study dem onstrates an association b etw een TC D D and adverse h ealth effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied (for all cancers combined). No statistically significant excesses noted in Zone A, or Zone B relative to reference area. Evidence of an exposure-response relationship was detected for lymphatic and hematopoietic tissues by num ber o f years since first exposure. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration not satisfied. Subjects were assigned to one o f the zones (A, B, R, or reference) based on official residence on the day o f the accident or at entry into the area. Exposure m isclassification is likely and lack o f individual-level data precludes an exam ination o f this source of error. 5. C onsideration Study size and follow -up are large enough to y ield p recise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. In total, 27, and 222, cancer deaths were found among residents of Zones A, and B, respectively. This allowed exam ined o f gender-specific effects. 1. C riteria Study is published in the peer-reviewed scientific literature and has an appropriate discussion o f the strengths and limitations. This document is a draftfor review purposes only and does not constitute Agency policy. B-16 DRAFT--DO NOT CITE OR QUOTE Response 2. Criteria Response 3. C riteria Response C riteria satisfied. A m J Epidem iol, 2001 Ju n 1; 153(11):1031--1044. A uthors discuss completeness of mortality ascertainment, diagnostic accuracy o f death certificates particularly w ith respect to diabetes, limited available o f blood dioxin measures that did not permit estim ation o f TCDD dose on an individual-level basis. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria not satisfied. Individual-level exposure data are unavailable. Exposure based on place of residence at time o f the explosion. Soil sam pling perform ed indicated considerable variability in TCDD levels w ithin each region. In addition, place o f residency at tim e of explosion does not ensure individuals were at their home around the time of the accident. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria not satisfied. A n ecological measure o f exposure (region o f residency at time of accident) was used to categorize individuals according to their possible exposure. Latencies were considered. W hile such an approach has value for identifying wherever excesses occurred am ong highly exposed populations, it is not precise enough to conduct a quantitative dose-response analysis. Conclusion The lack of individual-level exposure data precludes quantitative dose-response modeling using these data. 1 2 3 Table B-13. Pesatori et al., 2003-- All cancer sites com bined, site-specific 4 analyses 5 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. M ortality was ascertained from 1977-1996, and, as reported in other related manuscripts, appears to be well captured from the vital statistics registries in the region (99% complete). Cancer incidence data was available from 1977-1991. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Individual-level data on potential confounders (i.e., age, calendar period, and gender) were adjusted for. 3. C onsideration Study dem onstrates an association b etw een TC D D and adverse h ealth effect w ith evidence o f an exposure-response relationship. Response Consideration not satisfied. A lthough risk o f all cancer mortality was not associated w ith zone of residence, increased risk o f cancer incidence was observed in Zone A. Among men, excess lymphatic and hem atopoietic cancer incidence was observed in Zone A (primarily to non-H odgkin's lym phom a). Soft tissues sarcom a cancer incidence w as also associated w ith residence in Zone R among males, but not the more highly exposed zones (A and B). Among females living in Zones A and B, higher rates were observed for multiple m yelom a (RR = 4.9, 95% CI = 1.5-16.1), cancer of the vagina (RR = 5.5, 95% CI = 1.3-23.8), and cancer o f the biliary tract (RR = 3.0, 95% CI = 1.1-8.2). This document is a draftfor review purposes only and does not constitute Agency policy. B-17 DRAFT--DO NOT CITE OR QUOTE 4. Consideration Response 5. C onsideration Response Exposure assessm ent methodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Consideration not satisfied. Subjects were assigned to one o f the zones (A, B, R, or reference) based on official residence on the day o f the accident or at entry into the area. Exposure m isclassification is likely and lack o f individual-level data precludes an exam ination o f this source of error. Study size and follow-up are large enough to yield precise estimates o f risk and ensure adequate statistical power. Consideration satisfied for some endpoints, although several o f the cancer specific mortality results am ong w om en were based on very small num ber o f deaths (i.e., <5). 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Occup Env M ed, 1998; 55:126-131. Authors discuss limitations such as residency-based exposure assignment, absence of smoking, differential and death certification in exposed versus nonexposed areas. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria not satisfied. Individual-level exposure data are unavailable. Exposure based on place of residence at time o f the explosion. Soil sam pling perform ed indicated considerable variability in TCDD levels w ithin each region. In addition, place o f residency at tim e of explosion does not ensure individuals were at their home around the time of the accident. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria not satisfied. A n ecological measure o f exposure (region o f residency at time of accident) was used to categorize individuals according to their possible exposure. Latencies were considered. W hile such an approach has value for identifying wherever excesses occurred among highly exposed populations, it is not precise enough to conduct a quantitative dose-response analysis. Conclusion No dose-response patterns evident in the study, and the study lacked quantifiable measures of TCDD at an individual-level basis. The data are not w ell suited for dose-response analysis. 1 2 3 Table B-14. Consonni et al., 2008-- A ll cancer sites com bined, site-specific 4 analyses 5 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. M ortality appears to be well captured from the vital statistics registries in the region (99% complete). Both cancer and noncancer mortality evaluated, although diagnostic accuracy o f death certificates is likely low. Ideally, would have evaluated incident rather than decedent outcomes for cancer. This document is a draftfor review purposes only and does not constitute Agency policy. B-18 DRAFT--DO NOT CITE OR QUOTE 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Individual-level data on potential confounders (i.e., age, calendar period, and gender) were adjusted for. Comparison o f cancer mortality rates before the time of the accident betw een the regions also revealed no differences. Inform ation from other independent surveys suggests similarity betw een smoking behaviors across the regions. 3. C onsideration Study dem onstrates an association b etw een TC D D and adverse h ealth effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied for some outcomes. For all cancer sites combined, no evidence of dose-response was observed relative to general population across Zones A, B and R. Only statistically significant excess found in Zone A was for chronic rheumatic disease but based on only three deaths. H igher cancer excesses were found in Zone A after a latency period was incorporated; however, no dose-response relationship observed w ith this latency period. Evidence o f an exposure-response relationship was detected for lymphatic and hematopoietic tissues by zone o f residence. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration not satisfied. Subjects were assigned to one o f the zones (A, B, R, or reference) based on official residence on the day o f the accident or at entry into the area. Exposure m isclassification is likely and lack o f individual-level data precludes an exam ination o f this source of error. 5. C onsideration Study size and follow -up are large enough to y ield p recise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. In total, 42, 244, and 1,848 cancer deaths were found am ong residents of Zones A, B, and R respectively. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Am J Epidemiol, 2008, 167:847-858. Authors discuss potential for selection bias, lim itation o f residential based measure of exposure, similarities o f mortality ascertainm ent in exposed and referent populations, and multiple testing. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria not satisfied. Individual-level exposure data are unavailable. Exposure based on place of residence at time o f the explosion. Soil sam pling perform ed indicated considerable variability in TCDD levels w ithin each region. In addition, place o f residency at tim e of explosion does not ensure individuals were at their home around the time o f the accident. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria not satisfied. A n ecological measure o f exposure (region o f residency at time of accident) was used to categorize individuals according to their possible exposure. Latencies were considered. W hile such an approach has value for identifying wherever excesses occurred am ong highly exposed populations, it is not precise enough to conduct a quantitative dose-response analysis. This document is a draftfor review purposes only and does not constitute Agency policy. B-19 DRAFT--DO NOT CITE OR QUOTE Conclusion The lack of individual-level exposure data precludes quantitative dose-response modeling using these data. 1 2 3 Table B-15. Baccarelli et al., 2006-- Site-specific analysis 4 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. Polym erase chain reaction (PCR) m ethods were used to describe outcom e m easures. The prevalence o f t( 14; 18) w as estim ated as those individuals having a t(14; 18) positive blood sample divided by the t(14; 18) frequency (num ber o f copies per million lymphocytes). 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Questionnaire data were used to collect inform ation on cigarette smoking. Other potential confounders (age, smoking status, and duration o f smoking). In addition, both exposure and outcome were objectively and accurately measured. 3. C onsideration Study dem onstrates an association b etw een T C D D an d adverse health effect w ith evidence o f an exposure-response relationship. Response C onsideration w as not satisfied. A ssociations w ere detected betw een the frequency o f t( 14; 18) and plasm a TCDD levels as w ell as zone o f residence at the time o f the explosion. No association w as detected fo r these exposure m easures and prevalence o f t( 14; 18). A doseresponse trend was detected for TCDD and the m ean number o f t(14;18) translocations/106 lym phocytes, h ow ever the relevance o f t( 14; 18) in lym phocytes to non-H odgkin's lym phom a is uncertain. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. The authors highlight that exposure metrics represent both past and current body burdens. They employ several different exposure metrics of TCDD: place of residence (Zone A, B, R or reference), categorical serum m easures, a linear term, log (base 10) transform ed TCDD, and individuals w ith chloracne diagnosed after the accident. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk an d ensure adequate statistical power. Response Consideration satisfied. Analyses are made using 72 highly exposed, and 72 low exposed individuals. 1. C riteria Response 2. Criteria Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Carcinogenesis, 2006, 27(10):2001-2007. The authors discuss the lim itation o f using t( 14; 18) translocations as an outcom e m easure, and the uncertain role it plays in the developm ent o f non -H o d g k in 's lym phom a. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. This document is a draftfor review purposes only and does not constitute Agency policy. B-20 DRAFT--DO NOT CITE OR QUOTE Response 3. C riteria Response Criteria satisfied. A total o f 144 subjects w ere included in the study. This included 72 subjects who had low exposures, and 72 who had high exposures based on serum concentrations. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria satisfied. A variety o f measures were em ployed including current TCDD levels, as well as surrogates o f exposure at the time o f the accident. Conclusion W hile an association w as observed w ith the frequency o f t(14; 18) translocation, it is uncertain w hether this translates into an increased risk o f non -H o d g k in 's lym phom a. G iven the speculative nature of this endpoint and lack o f dem onstrated adverse effect, dose-response analyses for this outcome were not conducted. 1 2 3 Table B-16. W arner et al., 2002-- Breast cancer incidence 4 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. Diagnoses o f incident breast cancer were based on interview and inform ation from m edical records appears thorough. O f the 15 cases o f b reast cancer, 13 w ere confirm ed by pathology and the rem aining 2 by surgery report only. Three cases o f breast cancer were excluded w hich represents a large proportion o f the total cases identified. This would reduce sample size and could result in bias if the exclusion was association w ith TCDD exposure. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Inform ation was collected on an extensive series o f risk factors by using an interviewer adm inistered questionnaire. Participation rates for the survey were fairly good (80%). 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. Lim ited evidence (not statistically significant) of a dose-response w hen TCDD was analyzed as a categorical variable; only one breast cancer case was in the referent exposure category. In the analysis o f TC D D as a continuous m easure (log10TC D D ), the hazard ratio associated w ith a 10-fold increase in TCDD serum levels was 2.1 (95% CI: 1.0-4.6). 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. Different exposure metrics were considered in these analyses (categorical, continuous, measures on a log-scale). Exposure data are of high quality as they are based on serum samples taken among w om en near the time o f the accident. As such, exposure assignment is not dependent on as many assum ption as used in occupational cohorts were back-extrapolation for many years had to be performed. This document is a draftfor review purposes only and does not constitute Agency policy. B-21 DRAFT--DO NOT CITE OR QUOTE 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration somewhat satisfied. Inadequate follow-up for cancer limited the num ber of cases available. Sample size also lim ited the conclusions draw from the categorical analysis based on very few cases for some exposure categories. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Paper published in Environ H ealth Perspect, 2002 Jul, 110(7):625-628. A m ajor lim itation o f the study is the sm all num ber o f incident cases o f b reast cancer (n = 15), im portant strengths o f the study include characterization o f TCDD using serum collected near the time o f the accident. Exposure must be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. Serum was used to estim ate TCDD levels in 981 o f 1271 eligible wom en who had lived in either o f the two contam inated sites in 1976. D ata represent an objective measure o f TCDD near the time of the exposure. Data obtained near the time o f exposure w hich minimized the potential for exposure misclassification. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate w indow(s) o f exposure examined. Criteria satisfied. Exposure characterized using serum measures obtained close to the time of the accident. Conclusion W hile characterization o f exposure and availability o f other risk factor data at an individual-level b asis are im portant strengths o f this study, sm all sam ple size (n = 15 cases) based on inadequate follow-up is a key limitation. Quantitative dose-response analyses were conducted using this study, but continued follow-up of the study population or consideration of all cancer outcomes would be valuable. 1 2 3 B.1.5. The Chapaevsk Study 4 5 Table B-17. Revich et al., 2001-- All cancer sites com bined, and site-specific 6 analyses 7 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration cannot be evaluated. Insufficient details are provided in the paper to gauge the com pleteness and coverage o f the cancer registry and mortality data. H ealth outcom es were studied on the basis o f inform ation in the official m edical statistics. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration not satisfied. Given that this is an ecological study, bias may be present. This document is a draftfor review purposes only and does not constitute Agency policy. B-22 DRAFT--DO NOT CITE OR QUOTE 3. C onsideration Study dem onstrates an association betw een TC D D and adverse h ealth effect w ith evidence o f an exposure-response relationship. Response Consideration cannot be evaluated. Dose-response was not evaluated as exposure was based on residency in the region vs. no residency. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration not satisfied. No individual-level exposure estimates were used. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. A total o f 476 cancer deaths were observed among males, and 376 cancer deaths observed among females. The precision o f the SMRs is dem onstrated w ith fairly narrow confidence intervals for m any causes o f death. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria not satisfied. Published in Chemosphere, 2001, 43(4-7):951-966. Authors do not address the completeness o f the mortality follow-up, and whether there are differences in death registrations betw een regions. The authors do acknowledge, however, that new investigations being undertaken would characterize exposure using serum-based measures. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria not satisfied. It is a cross-sectional study that compares mortality rates betw een regions. No individual-level exposure data available. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria not satisfied. No individual-level exposure estim ates were used in the study. Conclusion These cancer data are cross-sectional in nature and not appropriate for a dose-response analysis. 1 2 3 B.1.6. The Air Force Health ("Ranch H ands") Study 4 5 Table B-18. A khtar et al., 2004-- All cancer sites com bined and site-specific 6 analyses 7 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. Cancer incidence and mortality based on inform ation from repeated m edical examinations, medical records and death certificate. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. This document is a draftfor review purposes only and does not constitute Agency policy. B-23 DRAFT--DO NOT CITE OR QUOTE Response Consideration not satisfied. The risk estimates were adjusted for a num ber o f factors m easured on an individual level including smoking. However, analyses are unable to distinguish betw een exposure to TCDD and 2,4-D as both were used in equal parts in the form ulation o f Agent Orange. 3. C onsideration Study dem onstrates an association b etw een T C D D an d adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. There is evidence o f a dose-response for all cancers and for some site-specific cancers (i.e., m alignant m elanoma, and prostate cancer). 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. H igh quality exposure data for m ost veterans was collected, so extrapolation to other members o f the cohort was not required. The serum dioxin m easurements also correlated w ell w ith reported skin exposure to herbicide in Vietnam, but collection o f the samples 25 years later required back-extrapolation. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk an d ensure adequate statistical power. Response Consideration satisfied. In total, 117 incidence cancers identified in the Ranch H ands cohort. For those sites w ith a dose-response association, malignant m elanom a and prostate cancer, there w ere 16 and 34 incident cases, respectively. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Published in J Occup Environ M ed, 2004, 46(2):123-136. Authors highlight that this is only cancer incidence study in US veterans, and the lengthy interval of follow-up (35-40 years)-- both im portant strengths o f the study. They addressed potential bias from healthy-worker effect, and uncertainties surrounding the estim ation o f TCDD exposure (extrapolation 30 years after exposure), as well as exposure to other chemical exposures. Study uses incident outcomes for cancer. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. Individual exposure estimates are based on measurements o f dioxin serum lipid concentrations. They were available for 1,009 R anch H ands and 1,429 in the com parison cohort. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria satisfied. TCDD exposures at the end o f duty were estimated by back-extrapolating 1987 serum values. Conclusion The m ajor lim itation o f the study is the inability to isolate effects o f TCDD from other chemicals used in the form ulation o f the herbicides. This limitation precludes dose-response modeling o f the TCDD and cancer outcom es data. 1 2 3 This document is a draftfor review purposes only and does not constitute Agency policy. B-24 DRAFT--DO NOT CITE OR QUOTE 1 Table B-19. M ichalek and Pavuk, 2008-- All cancer sites com bined 2 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. Cancer incidence was ascertained through the use of medical records. D eath certificate were used to identify some malignancies. Little data is provided on the num ber o f individuals lost to follow-up, however the same mechanisms o f case ascertainment were applied to both the com parison and R anch Hand cohorts. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration not satisfied. Inform ation collected from repeated physical examinations allowed for the adjustm ent o f risk factors such as smoking. Agent Orange was a 50% mixture o f 2,4-D and TCDD; therefore, potential for confounding by other coexposures is likely. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied for some comparisons. Statistically significant associations were noted w ith cancer incidence and TCDD w hen analyses were restricted to workers who served at m ost two years in Southeast A sia and those who sprayed more than 30 days before 1967. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. Initial TCDD dose were estimated at the end o f the tour o f duty for the Ranch Hands. Individual-level serum dioxin measurements correlated well with correlated w ith days o f spraying and calendar period of service, but collection o f the samples roughly 20 years later required back-extrapolation. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. A total o f 347 incident cases o f cancer were used in the analyses. For stratified analyses, statistical pow er is more limited. For example, only 67 incident cancer in the subset o f workers who spent less than 2 years in Southeast Asia, and sprayed for at least 30 days before 1967. 1. C riteria Response 2. Criteria Response 3. C riteria Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied J Occup Environ M ed 2008; 50:330-340. The authors discuss issues related to exposure misclassification error, and suggest approaches for improving characterization of days o f spraying. Congener specific data were unavailable, thereby not allowing for congener specific risks or adjustments to be made. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. TCDD data was available for 986 veterans in the Ranch Hand cohort, and 1,597 m em bers of the com parison cohort. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. This document is a draftfor review purposes only and does not constitute Agency policy. B-25 DRAFT--DO NOT CITE OR QUOTE Response Conclusion Criteria satisfied. TCDD exposures at the end of duty were estim ated by back-extrapolating 1987 serum values. R anch H and veterans were exposed to other contam inants in the herbicides that were mixed, thereby m aking it difficult to determine independent effects o f TCDD on cancer. In particular, 2,4-D has been shown to be associated w ith some cancers, notable cancer o f the prostate. This limitation precludes dose-response modeling of TCDD and cancer using data from this cohort. 1 2 B.1.7. O ther Studies of Potential Relevance to Dose-Response M odeling 3 4 Table B-20. `t M annetje et al., 2005-- A ll cancer sites com bined, site specific 5 analyses 6 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response 2. Consideration Consideration satisfied. National records for death registrations through the New Zealand H ealth Inform ation Service (NZHIS). Subjects not registered as having died during the study period were confirmed to be actually alive and resident in New Zealand using the New Zealand Electoral Roll, drivers' license, and social security records. R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response 3. C onsideration Consideration not satisfied. Seventeen percent o f workers were lost to follow up but it is unclear if bias resulted. The dichotom ous exposure m easure was based on exposure to TCDD, chlorinated dioxins and phenoxy herbicides, so confounding is a possibility by these coexposures. Study demonstrates an association betw een TCDD and adverse health effect w ith evidence of an exposure-response relationship. Response 4. Consideration Response 5. C onsideration Response Consideration satisfied. Dose-response evidence for duration o f employment and elevated mortality noted only in synthesis workers. Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Exposure m easures were lim ited to duration of em ploym ent and exposed/unexposed. Study size and follow-up are large enough to yield precise estim ates of risk and ensure adequate statistical power. Consideration satisfied. For all cancer sites combined, there were 43 cancer deaths am ong the production workers, and 35 such deaths among the sprayers. Site-specific cancer analyses are lim ited by small sample sizes. 1. C riteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria not satisfied Occup Env Med, 2005; 62:34-40. A high percentage of the cohort was lost to follow-up (17%). The authors fail to m ention this im portant lim itation in this paper. This document is a draftfor review purposes only and does not constitute Agency policy. B-26 DRAFT--DO NOT CITE OR QUOTE 2. Criteria Response 3. C riteria Response Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria not satisfied. This study used duration o f exposure, at an individual level, as a surrogate measure o f TCDD. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Criteria not satisfied. Exposure was defined according to duration, and not concentrations of TCDD. Latency intervals were not evaluated. Conclusion Overall, quantitative exposure data are lacking for TCDD and limited dose-response relationships were observed across duration o f exposure categories. Furthermore, confounding by coexposures is a possibility. Taken together, these data are not suitable for inclusion in a dose-response analysis 1 2 3 Table B-21. M cBride et al., 2009b-- All cancer sites com bined, site-specific 4 analysis 5 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. The New Zealand H ealth Inform ation Service M ortality Collection and the R egistrar-G eneral's Index to D eaths. A dditional searches w ere b ased o n the last known address from the work record; the electoral roll and the habitation index; the telephone book; the internet; and Terranet property inform ation database. A n additional search was carried out through the Births, Deaths, and M arriages office o f the New Zealand Departm ent of Internal Affairs. Lastly, autom ated personnel and pension records were also used to locate past New Plym outh workers and identify some deaths. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration not satisfied. Considerable am ount o f workers were lost to follow up (22%), but it is unclear if bias resulted. The dichotom ous exposure m easure was based on exposure to TCDD, chlorinated dioxins and phenoxy herbicides, so confounding is a possibility by these coexposures. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration not satisfied. There was no exam ination of dose-response effects. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. Dichotomous exposure (exposed/unexposed) and duration of employment were examined from job exposure classification assessed via occupational history records industrial hygienists/factory personnel knowledge and questionnaires. Authors discuss limitations in the assignment o f exposure among cohort members. This document is a draftfor review purposes only and does not constitute Agency policy. B-27 DRAFT--DO NOT CITE OR QUOTE 5. C onsideration Response Study size and follow-up are large enough to yield precise estimates o f risk and ensure adequate statistical power. C onsideration not satisfied. A low num ber o f deaths (n = 76) may have lim ited ability to detect effects small in magnitude and exposure-response relationships. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion of the strengths and limitations. Criteria satisfied. Published in Occup M edicine, 2009; 59(4):255-263. The authors highlight cohort lost to follow-up, the limited size o f the cohort, differences in cohort definitions between sprayers and producers, and the potential for other exposures during employment at the plant. Exposure must be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria not satisfied. TCDD exposures were not quantified. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate w indow(s) o f exposure examined. Criteria not satisfied. Effective dose could not be estim ated given the lack o f individual-level exposure data. Conclusion The study lacks the quantification o f exposures at an individual level, precluding doseresponse analysis. This study is not considered further in the dose-response m odeling analysis. 1 2 3 Table B-22. M cBride et al., 2009a-- All cancer sites com bined, site-specific 4 analysis 5 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. The New Zealand H ealth Inform ation Service M ortality Collection and the R egistrar-G eneral's Index to D eaths w ere used to identify deaths. A dditional searches were based on the last know n address from the work record; the electoral roll and the habitation index; the telephone book; the internet; and several other public databases in New Zealand. A n additional search was carried out through the Births, Deaths, and Marriages office o f the New Zealand Departm ent o f Internal Affairs. Lastly, autom ated personnel and pension records were also used to locate past New Plym outh workers and identify some deaths. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. W orkers lost to follow-up were an unlikely source o f bias especially for internal analyses. Confounding by other coexposures (e.g., 2,4,6-TCP) unlikely to have resulted in bias, due to presumed poor correlation w ith TCDD. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. This document is a draftfor review purposes only and does not constitute Agency policy. B-28 DRAFT--DO NOT CITE OR QUOTE Response Consideration not satisfied. The linear test for trend for TCDD exposure was not statistically significant for all cancer sites (combined), as well as lung cancer mortality. Dose-response relationships were not apparent across quartiles o f TCDD exposure for all cancer sites com bined, digestive cancers, lung cancer, soft tissue sarcom as o r n o n-H odgkin's Lym phom a. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. Cumulative exposure to TCDD as a tim e-dependent metric was estimated for each w orker from serum samples, but the authors did not examine a continuous measure o f TCDD exposure (lagged or unlagged). 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion of the strengths and limitations. Criteria satisfied. Published in J Occup Environ M ed 51:1049-1056. This paper discussed the 22% o f the cohort lost to follow-up, differences in cohort definitions betw een sprayers and producers, and the potential for other exposures during em ploym ent at the plant. Exposure must be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. Serum measures available for 346 workers were used to derive TCDD exposures for the entire cohort using the area under the curve approach. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate w indow(s) o f exposure examined. Criteria satisfied. Effective dose could be estimated from serum-derived cumulative exposure estimates. Conclusion Given that no dose-response associations were found, the data are not suited to dose-response analysis. 1 2 3 Table B-23. H ooiveld et al., 1998-- All cancer sites com bined, site-specific 4 analysis 5 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. Outcom es were mortality. Few deaths expected to be m issed since only 5% o f the cohort was lost to follow-up or had emigrated. This document is a draftfor review purposes only and does not constitute Agency policy. B-29 DRAFT--DO NOT CITE OR QUOTE 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration not satisfied. Although dioxin-like com pounds (PCDDs, PCDFs, and PCBs) were m easured in the serum samples, these were not incorporated into the analysis. Therefore, confounding cannot be ruled out as an explanation o f the reported association. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. A dose-response pattern was observed for internal cohort comparison for all cancer mortality, w ith RRs o f 5.0 and 5.6 for the m edium and high exposure, respectively. Dose-response patterns evident for lung cancer as well. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. Detailed occupational histories to assign dichotomous exposures (exposed/unexposed) based on m axim um exposure levels. Although serum data also collected for TCDD and other coexposures (PCDDs, PCDFs, and PCBs), study only presents data for TCDD exposure. TCDD exposures at time of maximum exposure were extrapolated from measured serum. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration not satisfied for internal cohort comparisons in either m en or women. Among men, only 7 cancer deaths were observed am ong those in the unexposed part o f the cohort, and 51 am ong exposed w orkers. F o r external cohort com parisons, a total o f 20 deaths w ere observed. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion of the strengths and limitations. Criteria satisfied. Am J Epidemiol, 1998, 147:891-901. The authors address potential limitations of estimating TCDD exposure from a subsample of surviving workers, lack of smoking data, the healthy w orker effect, and relevance o f other occupational exposures. Exposure must be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. Serum samples were obtained from 94 o f 144 subjects who were asked to participate in serum m easurem ent study. O f these, a further 44 excluded due to absence due to holiday or work (n = 22), and nonexposed workers excluded because m atching exposed w orker not participating (n = 20). TCDD levels w ere extrapolated to the tim e o f m axim um exposure. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate w indow(s) o f exposure examined. Criteria not satisfied. Exposures assigned based on levels at m axim um exposure. Assignm ent o f exposure based on nonrepresentative sample o f 50 survivors am ong the occupational cohort. Conclusion The small num ber o f identified cancer deaths, limitations in terms o f the exposure assignment (based on nonrepresentative sample, and m axim um exposure level) and concern over potential confounding by coexposures preclude using these data for a dose-response analysis. This document is a draftfor review purposes only and does not constitute Agency policy. B-30 DRAFT--DO NOT CITE OR QUOTE 1 B.2. EV ALUA TIO N OF N O N C A N C ER STUDIES 2 B.2.1. NIO SH Cohort 3 4 Table B-24. Steenland et al., 1999-- M ortality (noncancer) 5 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. The study evaluated mortality from all cancer sites (combined). As described in the paper, the sources o f vital status and cause of death information were received from the Social Security death files, the National D eath Index, and the Internal Revenue Service. Vital status was know n for 99.4% o f the cohort members, cause o f death information is available for 98% o f the decedents. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration not satisfied. External comparisons for all-cause and cardiovascular mortality do not appear to be affected by the "healthy w orker effect" as sim ilar patterns were observed w ith internal cohort comparisons. Nonetheless, internal cohort com parisons are unable to adjust for many o f the individual-level risk factors for cardiovascular disease. 3. C onsideration Study dem onstrates an association betw een TC D D an d adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. A dose-response relationship was observed w ith ischemic heart disease (linear test for trend p = 0.05), and w ith TCDD on a log-transform ed scale the p-value was <0.001. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. The study conducted detailed sensitivity analyses and evaluated different assumptions regarding latency, log-transformed TCDD exposures, and half-life values for TCDD. Associations were stronger for log-transformed values, and latency intervals o f 15 years. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. This is the largest o f the occupational cohorts w ith exposures to TCDD. The cohort consisted o f 5,132 male workers and a total o f 456 deaths from ischemic heart disease. This perm its characterization o f risk for all cancer sites (combined). 1. C riteria Response 2. Criteria Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied Journal o f the National Cancer Institute, 1999, 91(9):779-786. The authors discussed the potential for bias from smoking, and other occupational exposures for which data for both were lacking at an individual basis. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. This document is a draftfor review purposes only and does not constitute Agency policy. B-31 DRAFT--DO NOT CITE OR QUOTE Response 3. C riteria Response Criteria not satisfied. Exposure scores assigned at an individual level based on job-exposure matrix (JEM). The JEM was based on estimated factor o f contact w ith TCDD in each job, level o f TCCD contam ination of materials at each plant over time, and proportion of day w orker could be in contact w ith materials. These factors were multiplied together to derive a daily exposure score, w hich was accum ulated over the working history o f each w orker to obtain a cumulative measure of TCDD. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. The follow -up o f the cohort extended from 1942 until the end o f 1993. Greater than 25 years o f follow-up have accrued in cohort allowing for latency to be examined. Different assumptions on the half-life o f TCDD were evaluated and produced sim ilar results. Latency intervals were incorporated, w ith strongest associations noted no lag. Suggests mechanisms occur at the same tim e as exposure. However, noncancer mortality is not a viable endpoint to consider for further dose-response analysis. Conclusion TCDD exposures were quantified in this study, and a dose-response relationship was observed w ith ischemic heart disease mortality. The sample size was sufficient, and the follow-up interval was lengthy. However, no individual-level data were available for cardiovascular conditions, and the inability to adjust for these exposures introduces considerable uncertainty into the risk estimates. Furthermore, noncancer mortality is not considered a viable endpoint for dose-response analysis. 1 2 3 Table B-25. Collins et al., 2009-- M ortality (noncancer) 4 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. Vital status complete for all but two workers. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. No inform ation collected on smoking status, but no excess in lung cancer or nonm alignant respiratory diseases noted. Analyses took into account potential for exposure to pentachlorophenol. External cohort comparisons should be interpreted cautiously due to healthy w orker effect, but internal cohort com parisons should not be influence by this bias. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration not satisfied. No statistically significant mortality excess for any noncancer mortality outcome evaluated. This included ischemic heart disease, stroke, nonm alignant respiratory disease, ulcers, cirrhosis, and external causes o f death (accidents). M odeling of continuous measure o f TCDD was not related to diabetes, ischemic heart disease, or nonm alignant respiratory mortality. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. This document is a draftfor review purposes only and does not constitute Agency policy. B-32 DRAFT--DO NOT CITE OR QUOTE Response 5. C onsideration Response Consideration satisfied. The authors used these serum from 280 form er TCP workers to estimate historical exposure levels of TCDD, furans, and polychlorinated biphenyls for all 1,615 workers. Exposure assessm ent included detailed w ork history, industrial hygiene monitoring, and the presence o f chloracne cases among groups o f workers. This data was integrated into a 1-compartment, first-order pharmacokinetic to determine the average TCDD dose associated w ith jobs in each group, after accounting for the presence of background exposures estim ated from the residual serum TCDD concentration in the sampled individuals. The authors did not evaluate departures from linearity, or examine skewness at higher exposures. No presentation of exposure levels was provided. Study size and follow-up are large enough to yield precise estim ates of risk and ensure adequate statistical power. Consideration satisfied. A total o f 662 deaths were observed. O f these, 218 were from ischem ic heart disease, and 16 from diabetes (two outcom es fo r w h ich associations have b een noted elsewhere). 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Published in Am J Epidemiol, 2009, 170(4):501-506. The authors discuss potential for exposure m isclassification, large size o f the cohort, lengthy follow-up interval, and large number of workers who provided serum from w hich TCDD exposures were estim ated. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. This study has the greatest num ber o f serum samples obtained from a specific plant. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. N oncancer mortality is not a viable endpoint to consider for further dose-response analysis. Conclusions No dose-response associations were noted for noncancer mortality outcomes. The data are, therefore, not suited for dose-response modeling. 1 2 3 B.2.2. BASF Cohort 4 5 Table B-26. Ott and Zober, 1996-- M ortality (noncancer) 6 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. M ortality ascertainm ent appeared to be fairly complete. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. This document is a draftfor review purposes only and does not constitute Agency policy. B-33 DRAFT--DO NOT CITE OR QUOTE Response 3. C onsideration Response 4. Consideration Response 5. C onsideration Response Consideration satisfied. Inform ation was collected on smoking status, body mass index, and other occupational exposures, however a large portion of the cohort was firefighters who may have been exposed to other occupational carcinogens. However, the recruitment of survivors may results in under-ascertainm ent o f mortality. Study dem onstrates an association betw een TCDD and adverse health effect w ith evidence of an exposure-response relationship. Consideration not satisfied. For external cohort comparisons across the three TCDD exposure categories, there was no dose-response pattern observed for any o f the noncancer causes of death. Cox regression risk estim ates for all cause or circulatory disease mortality w hen TCDD was m odeled as a continuous variable were not statistically significant. Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Consideration satisfied. Cumulative measure of TCDD expressed was derived from serum measures. Exposure was also estimated by chloracne status o f the cohort members. The authors have not addressed the potential im plication o f deriving TCDD exposure estimates for the whole cohort using sera data that were available for only about half of the cohort. Study size and follow -up are large enough to yield precise estim ates o f risk and ensure adequate statistical power. Consideration satisfied. For all causes o f death, there were 92 deaths, while 37 circulatory deaths. M any o f the cause-specific death had less than 5 deaths in the upper exposure category. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Occup Environ M ed, 1996, 53:606-612. A large component o f the cohort was assem bled by actively seeking out workers who were alive in the m id 1980s. As a result, it is likely a num ber o f deaths were missed. This is supported by m uch low er SM Rs in this com ponent of the cohort published in earlier studies o f the cohort. This underascertainm ent of mortality results in biased SMR statistics (underestimated). The authors do highlight the value of the serum based measures to estimate TCDD exposure Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. Serum samples, taken in 1989, were available for 138 surviving workers out of 254 and allowed for cumulative TCDD levels to be estimated using regression techniques in the rem ainder o f the cohort. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. Exposure assignment took into the affect that body mass index had on TCDD half-lives. TCDD levels estimates through back-extrapolation o f serum levels based on half-life estimates obtained from previous studies. Latency was considered w ith stronger association observed in external comparisons incorporating a latency o f 20 years. The follow up o f the cohort was lengthy (>50 years). However, noncancer mortality is not a viable endpoint to consider for further dose-response analysis. This document is a draftfor review purposes only and does not constitute Agency policy. B-34 DRAFT--DO NOT CITE OR QUOTE Conclusion No associations noted w ith any noncancer deaths. External com parisons should be treated cautiously especially for cardiovascular mortality w hich is recognized to often be biased by the healthy-worker effect. In the absence o f any outcome w ith an association w ith TCDD exposure, dose-response analyses o f these data were not undertaken. 1 B.2.3. H am burg Cohort 2 3 Table B-27. Flesch-Janys et al., 1995; Flesch-Janys et al., 1996 erratum -- 4 M ortality (noncancer) 5 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. M edical records used to identify deaths over the period 1952-1992. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Similarity in smoking rates betw een control cohort and the exposed workers was similar based on independent surveys. Occupational exposures to benzene, and dimethyl sulfate were unlikely to bias dose-response pattern observed as these exposures occurred in production departments w ith low to medium levels of TCDD exposure. 3. C onsideration Study dem onstrates an association b etw een TC D D and adverse h ealth effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. Dose-response relationship observed for all-cause mortality, cardiovascular mortality, and ischemic heart disease mortality across 6 exposure categories, w ith the cohort o f gas supply workers used as the referent. The linear tests for trend for these three outcom es w ere all statistically significant (p < 0.05). 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. The exposure measures was an integrated TCDD concentration over time estimate that back-calculated TCDD exposures to the end of the employment. Categorical and continuous TCDD exposures were examined in relation to the health outcome. These efforts improve the exposure assessm ent o f earlier studies. 5. C onsideration Study size and follow -up are large enough to y ield p recise estim ates o f risk an d ensure adequate statistical power. Response Consideration satisfied. For all causes o f death combined, there were 414 deaths in the exposed cohort, and 943 in the cohort o f gas supply workers. A total o f 157 and 76 deaths from cardiovascular disease, and ischemic heart disease were noted. The corresponding num ber in the cohort o f gas supply workers was 459, and 205, respectively. 1. C riteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. A m J Epidemiol, 1995, 1442:1165-1175. The authors discuss the potential role o f other occupational exposures (i.e., dim ethyl sulfate, solvents, benzene), smoking, and suitability o f the comparison cohort o f gas supply workers. This document is a draftfor review purposes only and does not constitute Agency policy. B-35 DRAFT--DO NOT CITE OR QUOTE 2. Criteria Response 3. C riteria Response Conclusion Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. Serum and adipose tissues were used to estimate TCDD exposure in 190 workers. A one-com partm ent first-order kinetic model was used to estimate exposure at end o f exposure for these workers. Regression m ethods were then used to estim ates TCDD exposures for all workers. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. Exposure based on half-life estimates from individuals w ith repeated serum measures. Other dioxin-like compounds were considered w ith the TOTTEQ exposure metric. N oncancer mortality, however, is not a viable endpoint to consider for further doseresponse analysis. Although, the exposure data used w ithin this study are well-suited to a dose-response analysis for all-cause and cardiovascular mortality given the associations observed, use o f noncancer m ortality endpoint is not am enable fo r further dose-response analysis. 1 B.2.4. The Seveso W om en's H ealth Study 2 3 Table B-28. Eskenazi et al., 2002a-- M enstrual cycle characteristics 4 1. C onsideration M ethods used to ascertain health outcomes identified were unbiased, highly sensitive, and specific. Response Consideration satisfied. Inform ation was also obtained from medical records for all obstetric and gynecologic conditions. Inform ation on menstrual cycles was obtained from questionnaires. W om en were asked about length o f cycles, regularity, how many days flow lasted, and heaviness o f m enstrual flow (scanty, moderate, or heavy). M easurem ent error is likely for the subjective nature o f self-reported menstrual parameters but specificity and sensitivity is difficult to ascertain due to lack o f validation data for these measures. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Detailed risk factor information was collected from questionnaire, allowing for the potential confounding influence o f many risk factors to be controlled for. The length o f cycle study findings may have been affected by the presence o f a few outliers. 3. C onsideration Study demonstrates an association betw een TCDD and adverse health effect w ith evidence of an exposure-response relationship. Response Consideration satisfied. A positive dose-response relationship was found w ith TCDD among w om en who were prem enarcheal at time o f the explosion and longer m enstrual cycle. Increased TCDD resulted in a reduced odds of scanty m enstrual flow. No association was noted w ith these two outcomes among postm enarcheal women. A decreased risk o f irregular cycles was observed w ith higher TCDD levels. This document is a draftfor review purposes only and does not constitute Agency policy. B-36 DRAFT--DO NOT CITE OR QUOTE 4. Consideration Response 5. C onsideration Response Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Criteria satisfied. Serum concentrations o f TCDD offer improved exposure assessment, although delineating the critical exposure window is challenging given the nature o f the very high initial exposure. Study size and follow-up are large enough to yield precise estim ates o f risk and ensure adequate statistical power. Consideration satisfied. Cohort was large enough as analyses were conducted on 301 women. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion of the strengths and limitations. Criteria satisfied. Am J Epidem iol, 2002; 156(4) 383-392. Lim itations included an inability to assess affects on m enstrual cycle at time body burdens were the highest (at time o f the accident). Also, TCDD was estimated for 1976, not concurrent w ith their cycles in the previous year, and a large num ber o f w om en were excluded due to intrauterine device or oral contraceptive use. Strengths included population-based nature o f study, w ith characterization of exposure using serum, and levels of other polychlorinated dibenzo-p-dioxins and dibenzofurans were at background levels. Findings for length o f m enstrual cycle may be unduly influenced by the presence o f some outliers. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. The study population was based on 301 w om en as those who were over the age o f 44 were excluded, as well as w om en w ith surgical o f natural menopause, w om en w ith T u rn er's syndrom e, those w ho had b ee n pregnant o r b reastfed in the p ast year, an d those w ho had used an intrauterine device or oral contraceptives. For 272 women, TCDD levels were based on serum data provided in 1976; TCDD levels were back-extrapolated to 1976 levels for the other 29 women. The effective dose and oral exposure can be reasonably estimated and the measures o f exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) o f exposure examined. Response had to be a nonfatal endpoint. Criteria satisfied. Ideally, TCDD exposures would be concurrent w ith reporting o f cycle characteristics. Herein, TCDD exposures were based on levels in 1976; however, given the long half-life of TCDD and the same follow-up interval for all women, TCDD exposures in 1976 should correlate well w ith levels near the time o f interview. Further, the critical window of exposure can be estimated for the wom en that were premenarcheal at the time of the accident (13 years). Conclusion This study meets all o f the criteria and considerations for further dose-response analysis. The determ ination o f the relevant time interval over w hich TCDD dose should be considered is uncertain . 1 2 3 4 5 6 This document is a draftfor review purposes only and does not constitute Agency policy. B-37 DRAFT--DO NOT CITE OR QUOTE 1 Table B-29. Eskenazi et al., 2002b-- Endom etriosis 2 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration not satisfied. Results o f a pilot study showed that ultrasounds had excellent specificity and sensitivity for ovarian endometriosis. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response C onsideration n ot satisfied. M ore th an h a lf o f the w om en w ere classified as `u n certain ' w ith respect to endom etriosis disease status. 3. C onsideration Study dem onstrates an association b etw een T C D D an d adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration not satisfied. W hile an increased risk o f endometriosis was observed across the 3 TCDD categories, these risks were not statistically significant relative to the lowest exposure category. The test fo r tren d b ased o n a continuous m easure (log10T C D D ) w as also n ot statistically significant. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Criteria satisfied. Serum concentrations o f TCDD offer im proved exposure assessment, although delineating the critical exposure window is challenging given the nature of the very high initial exposure. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk an d ensure adequate statistical power. Response C onsideration n ot satisfied. O nly a total o f 19 cases o f endom etriosis w ere identified. 1. C riteria Response 2. Criteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion of the strengths and limitations. C riteria satisfied. E nviron H ealth P erspect 2002; 110(7) 6 2 9 -6 3 4 . A u th o r's highlight th at this is the first study to examine the relationship betw een TCDD and endometriosis, and the availability o f sera data to estimate TCDD levels. Lim itations included the small num ber o f w om en w ith endometriosis, and inability to confirm disease status using laparoscopy. Finally, young w om en may have been underrepresented due to cultural difficulties in exam ining w om en who had never been sexually active. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. Eligible study subjects were wom en betw een 1 m onth and 40 years o f age at tim e o f accident. These analyses excluded virgins, those w ith T u rn er's syndrom e, and w om en who refused the examination o f ultrasound. Serum data were available for the 601 participants on w hich the analyses are based. O f these, 559 had serum m easures taken in 1976/77, 25 betw een 1978 and 1981, and 17 w om en in 1996. This document is a draftfor review purposes only and does not constitute Agency policy. B-38 DRAFT--DO NOT CITE OR QUOTE 3. C riteria Response The effective dose and oral exposure can be reasonably estimated and the measures o f exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) o f exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. TCDD exposure was estimated at the time o f "conception attempt" using serum measures, w ith extrapolation from 1976 levels using half-life assumptions. It is difficult to identify the relevant time interval over w hich TCDD dose should be considered for dose-response analysis. The critical window o f exposure is unknown. Conclusion The lack o f a statistically significant association coupled w ith a large num ber of w om en for w hich endom etriosis disease status was "uncertain", precludes the use o f these data to conduct doseresponse analysis. 1 2 3 Table B-30. Eskenazi et al., 2003-- Birth outcom es 4 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration not satisfied. Outcomes were identified through self-reported questionnaires. W om en were found to over-report birth weight, and have a tendency to underreport birth defects in children. As a large num ber o f w om en in Seveso underwent voluntary abortion in the first year after the explosion, an awareness bias may have contributed to differential reporting o f pregnancy histories. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration not satisfied. See above. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration not satisfied. There was no association betw een spontaneous abortions and log10TC D D , o r w ith births sm all fo r gestational age. A n inverse association w ith b irth w eight was noted in first eight years following the accident as were the num ber o f births small for gestational age; however, none achieved statistical significance a tp < 0.05. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Criteria satisfied. Serum concentrations o f TCDD offer im proved exposure assessment, although delineating the critical exposure window is challenging given the nature o f the very high initial exposure. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. For spontaneous abortions there were 769 pregnancies. Fetal growth and gestational age analysis was carried out on 608 singleton births that occurred post explosion. 1. C riteria Study is published in the peer-reviewed scientific literature and has an appropriate discussion of the strengths and limitations. This document is a draftfor review purposes only and does not constitute Agency policy. B-39 DRAFT--DO NOT CITE OR QUOTE Response 2. Criteria Response 3. C riteria Response Criteria satisfied. Environ Health Perspect, 2003, 111(7):947-953. The authors highlight potential limitation o f reliance on self-reported data to ascertain pregnancy outcomes. They also address the relevance o f paternal exposures to TCDD on the developing fetus-- such exposure data were not considered in this study. Exposure must be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. A total o f 745 w om en in the SWHS had reported getting pregnant, o f these 510 w om en were pregnant after the explosion (888 pregnancies). Analyses of spontaneous abortions based on 476 w om en (excludes those w ith voluntary abortion, ectopic pregnancy, or m olar pregnancy). TC D D m easured fo r 413 w om en in 1976/77, 12 w o m en b etw een 1978 and 1981, and 1996 fo r 19 w om en. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate w indow(s) o f exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. TCDD exposures were extrapolated to 1976 values. However, it is difficult to identify the relevant time interval over w hich TCDD dose should be considered for dose-response analysis. Conclusion The findings of the study are som ewhat lim ited due to the reliance on self-reported inform ation for pregnancy outcomes, and lack of paternal exposures. The findings were not statistically significant. Considered together, quantitative dose-response analyses for this study population were not undertaken. 1 2 3 Table B-31. W arner et al., 2004-- A ge at m enarche 4 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. In this study age at menarche was based on retrospective recall 5 to 19 years before the interview . P revious w o rk suggests m oderate to h igh correlations b etw een actual and recalled menarche, misclassification of outcome would bias risk estimates towards the null (assuming nondifferential misclassification). 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. D ata collected from self-reported questionnaires allow for the potential confounding influence o f many risk factors to be taken into account. Some m isclassification o f outcome may bias risk estim ates towards the null. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration not satisfied. There was no association betw een TCDD levels and the age at menarche w ith either the continuous or categorical measures o f TCDD. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. This document is a draftfor review purposes only and does not constitute Agency policy. B-40 DRAFT--DO NOT CITE OR QUOTE Response 5. C onsideration Response Criteria satisfied. Serum concentrations o f TCDD offer im proved exposure assessment, although delineating the critical exposure window is challenging given the nature o f the very high initial exposure. Study size and follow-up are large enough to yield precise estimates o f risk and ensure adequate statistical power. Consideration satisfied. Cohort was large enough as analyses were perform ed using 282 w om en who were premenarcheal at the time o f the explosion. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion of the strengths and limitations. Criteria satisfied. Environ Health Perspect, 2004, 112:1289-1292. Authors discuss use of pooled serum from residents o f the unexposed zone, and that those in lowest exposure group had high exposures relative w ith contemporary levels for the area. Strengths o f study include use of serum to estimate TCDD exposure. Exposure must be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. The SWHS included w om en betw een 1 m onth and 40 years of age at time of accident w ho attem pted to get pregnant after the explosion (n = 463). This study is restricted to those who were prem enarcheal at the time o f the explosion (n = 282). Serum w as collected for these wom en, prim arily in 1976-1977 (n = 257), betw een 1978 and 1981 for 23, and in 1996-1997 for the 2 remaining women. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate w indow(s) o f exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. TCDD exposures in 1976 were estimated by extrapolation serum levels obtained after this date using the Filser model. Both categorical and continuous measures of exposure were modeled. In utero measures o f exposure are likely most relevant exposure based on findings from anim al studies. Conclusion No association betw een TCDD levels and age at menarche was found. There may be some m isclassification o f age at menarche based on self-report, and biologically, the most relevant dose as suggested by anim al studies occurs in utero. Additionally, it is difficult to identify the relevant time interval over w hich TCDD dose should be considered for dose-response analysis. For these reasons, these data are not suited to a dose-response analysis. 1 2 3 Table B-32. Eskenazi et al., 2005-- A ge at m enopause 4 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. Outcome measures were obtained based on self-reported data collected from questionnaires. Studies have shown that self-reports of age at menopause are reported w ith accuracy and reliability, and among w om en w ith surgical menopause, the self reported age correlated well w ith that on the medical records. This document is a draftfor review purposes only and does not constitute Agency policy. B-41 DRAFT--DO NOT CITE OR QUOTE 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. D ata obtained from the questionnaire allow for the potential confounding influence o f several potential confounders to be controlled for. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration not satisfied. Although risks o f earlier menopause increased in the first four quintiles, w ith a statistically significant trend, no increased risk was noted in the highest exposure category (hazard ratio = 1.0 relative to low est exposure group). Study authors suggest this is due to the "inverted U " dose response often seen w ith horm onally active com pounds. A dditionally, no statistically significant association w as noted w ith log10TC D D for the individual quintiles. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Criteria satisfied. Serum concentrations o f TCDD offer im proved exposure assessment, although delineating the critical exposure window is challenging given the nature o f the very high initial exposure. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. The study included 616 women. O f these, 260 were premenopausal, 169 classified as natural m enopause, 83 as surgical m enopause, 24 as im pending m enopause, 33 as prem enopausal, and 58 in a n "other" category. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion of the strengths and limitations. Criteria satisfied. E nviron H ealth Perspect, 113:858-862 (2005). A uthors highlight this is first study to look at relationship betw een dioxin and age at menopause. Other limitations of the study include lowest exposure group (< 20.4 ppt) includes exposures level that are far higher than background, and age at menopause was based on retrospective recall. Strength of study is ability to characterize TCDD using serum measures. Exposure must be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. C riteria satisfied. The Seveso W o m en 's H ealth Study collected serum sam ple w h ich allow ed TC D D exposures to be characterized. Those w om en (n = 616) w ho had not reached natural menopause at the time of the accident were included in the study. Serum measures collected in 1976/77 were available for 564 women, for 28 women, sera was collected betw een 1978 and 1981, while for 24 women, sera was collected in 1996/97. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate w indow(s) o f exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. TCDD levels were estimated at the time o f the explosion using available inform ation on TCDD half-life. However, it is difficult to identify the relevant time interval over w hich TCDD dose should be considered for dose-response analysis. The critical window of exposure can be estim ated but is large and highly uncertain. This document is a draftfor review purposes only and does not constitute Agency policy. B-42 DRAFT--DO NOT CITE OR QUOTE Conclusion The findings do not provide strong support for a dose-response relationship. As such, they are not well suited to a quantitative dose-response analysis. 1 2 3 Table B-33. W arner et al., 2007-- O varian function 4 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. Ovarian cyst analysis based on w om en who underwent ultrasound (n = 310). O varian follicle analysis based on self-report on m enstrual cycle and done in w om en in preovulatory cycle (n = 96) at time o f ultrasound. Horm onal analysis based on w om en in last 14 days o f cycle (n = 129). 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. D ata collected from self-reported questionnaires allow for the potential confounding influence o f many risk factors to be taken into account. Some m isclassification o f outcome based on self-reports o f menstrual cycle may bias risk estimates towards the null. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration not satisfied. There was no association betw een serum TCDD levels and the num ber or size of ovarian follicles. TCDD was also not associated w it the odds o f ovulation. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Criteria satisfied. Serum concentrations o f TCDD offer im proved exposure assessment, although delineating the critical exposure window is challenging given the nature o f the very high initial exposure. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. Cohort was large enough as analyses were perform ed using 129 w om en for ovulation outcome, and horm one analyses based on 87 w om en in luteal, and 55 in m idluteal phases. 1. C riteria Response 2. Criteria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion of the strengths and limitations. Criteria satisfied. Environ Health Perspect, 2007,115:336-340. A n important limitation cited by the authors was that w om en may not have been exposed at critical period (prenatally). Phases o f the cycle may also have been m isclassified as this was based on self-reported data. Strength, first study to have exam ined ovarian function and TCDD exposures. Exposure must be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. The SWHS included w om en betw een 1 m onth and 40 years of age at time of accident who were betw een 20-40 years o f age and not using oral contraceptives at follow-up (n = 363).O f these, serum w as collected fo r 330 w om en betw een 1976 and 1977, betw een 1978 and 1982 for 25 women, and betw een 1996 and 1997 for 8 women. This document is a draftfor review purposes only and does not constitute Agency policy. B-43 DRAFT--DO NOT CITE OR QUOTE 3. C riteria Response The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate w indow(s) o f exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. The w om en may not have been exposed at critical period (prenatally). Conclusion No association betw een TCDD levels and ovarian function was found. There may be some m isclassification o f period o f the cycle based on self-report, and biologically, the most relevant dose as suggested by animal studies occurs in utero. For these reasons, these data are not suited to a dose-response analysis. 1 2 3 Table B-34. Eskenazi et al., 2007-- U terine leiom yom a 4 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. Outcomes were determined using two definitions: current fibroids, or past diagnosis o f fibroids. For past diagnosis o f fibroids, self-reported data and medical records were used to determine w hether w om en were previously diagnosed w ith fibroids, these were confirmed w ith medical records. A total o f 25 w om en indicated they had never been diagnosed w ith fibroids. M edical records indicate a past diagnosis for these women, and they were classified as such. For current fibroids, this was determined at the time o f the interview for 634 w om en using transvaginal ultrasound examinations. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. In the SWHS questionnaires were administered to the participants and detailed data for reproductive characteristics, smoking, body mass index, and alcohol use were collected so risks could readily be adjusted for these covariates. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied, but inversely. A n inverse dose-response pattern w ith the percentage of wom en diagnosed (current and past history-- combined) w ith fibroids across 3 categories of exposure. Namely, the percentages o f wom en w ith fibroids in the <20, 20.1-75.0, and >75.0 ppt categories were 41.1%, 26.8%, and 20.0%, respectively. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. A variety o f different exposure metrics were considered including linear, categorical, splines, and logJ0TCD D . 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. A total o f 251 w om en were found to have fibroids, and there were 62, 110, and 79 w om en w ith fibroids diagnosed in the 3 TCDD exposure categories. 1. C riteria Study is published in the peer-reviewed scientific literature and has an appropriate discussion of the strengths and limitations. This document is a draftfor review purposes only and does not constitute Agency policy. B-44 DRAFT--DO NOT CITE OR QUOTE Response 2. Criteria Response 3. C riteria Response Criteria satisfied. Am J Epidemiol, 2007, 166:79-87. In this study, the authors found an inverse association betw een TCDD and uterine leiom yom a risk. The authors highlighted strengths of the study that included the longitudinal design, serum measures taken at an individual-level basis and most taken w ithin 2 years o f the accident, ability to include outcomes am ong those who did not take an ultrasound by using an adapted statistical approach. A n im portant lim itation that was the differences in risk by the stage o f developm ent could not be assessed as all wom en were exposed postnatally, and only 4 cases were observed among those who were premenarcheal at the time of exposure. Exposure must be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. C riteria satisfied. F inal sam ple consisted o f 956 w om en in the Seveso W o m en 's H ealth Study w ithout a history of fibroids. For 872 o f these women, serum was collected in 1976 and 1977. For 56 women, TCDD was measured in w om en betw een 1978 and 1981, and for 28 women the serum was collected in 1996. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate w indow(s) o f exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. TCDD exposures were back extrapolated to expected levels in 1976 (at the time o f the accident). However, it is difficult to identify the relevant time interval over w hich TCDD dose should be considered for dose-response analysis. The critical window of exposure is unknown. Conclusion The data suggest an inverse (protective) effect betw een fibroids and exposure to TCDD. As such, these data are not suited to further dose-response analyses. 1 2 3 B.2.5. O ther Seveso Noncancer Studies 4 5 Table B-35. M ocarelli et al., 2008-- Sem en quality 6 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. Serum levels o f TCDD were measured on an individual basis for men in exposed areas; pooled samples from m en in uncontam inated areas were measured to assess background TCDD exposure levels. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. W hile compliance rates may have introduced some possible bias, this does not seem likely as different effects noted betw een the 22-31 and 32-39 year old age groups. Information collected for other risks factors, w hich have been used as adjustment factors in the models. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. This document is a draftfor review purposes only and does not constitute Agency policy. B-45 DRAFT--DO NOT CITE OR QUOTE Response Consideration satisfied. Figure 3 suggests dose-response relationship among those aged 1-9 at the time o f the accident for sperm concentration and motility. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. Serum concentrations o f TCDD offer improved exposure assessment, although delineating the critical exposure window is challenging. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. A nalyses are based on 135 m ales exposed to TCDD. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion of the strengths and limitations. C riteria satisfied. E nvironm ental H ealth Perspective s, 2008, 116(1):70-77. T he authors describe strengths associated w ith characterization of exposure (using serum samples), and representativeness o f study population. Lim itation o f study includes low compliance (but high for sem en sample studies), namely, 60% am ong a group o f healthy men. The compliance rate was higher among exposed group (69%). Exposure must be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. C riteria satisfied. Involved m ales, < 16 years old at tim e o f accident. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate w indow(s) o f exposure examined. Response has to be a nonfatal endpoint. Criteria satisfied. TCDD exposures were based on serum samples. Serum samples were draw n (in 1997/1998) from participants w hose 1976 sam ples w ere above 15 ppt. P ooled samples obtained in 1997/98 were used to describe background TCDD levels in uncontaminated areas. The associated betw een TCDD exposure and semen quality was found statistically significant for the boys w ith 1 and 9 years o f age at the time of the accident. This provides a critical window o f exposure to estimate TCDD concentration. Conclusion Health outcomes are exposures are w ell characterized using serum data. However, the m en exposed betw een the ages o f 1 and 9 to elevated TCDD levels had reduced sem en quality 22 years later. It is difficult to discern w hether this effect is a consequence o f the initial high exposure between 1 and 9 years of age or a function o f the cumulative exposure for this entire exposure window beginning at the early age. Nonetheless, quantitative dose-response analyses for this outcome were conducted. 1 2 3 Table B-36. M ocarelli et al., 2000-- Sex ratio 4 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. Birth records examined for those who lived in parents who lived in the area and who provided serum samples. This document is a draftfor review purposes only and does not constitute Agency policy. B-46 DRAFT--DO NOT CITE OR QUOTE 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. Paternal TCDD exposures were associated w ith an increased probability o f fem ale births (p = 0.008). 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. Serum samples were used to estim ate m aternal and paternal TCDD levels. No discussion o f exposure levels in reference population. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. Statistically significant findings achieved. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion of the strengths and limitations. Criteria not satisfied. The Lancet, 2000, 355:1858-1863. There is no discussion on the strengths and lim itations o f this study. Exposure must be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. Serum levels of TCDD were obtained from parents using samples provided in 1976/77. Serum m easures available for 296 mothers and 239 fathers. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate w indow(s) o f exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. Serum based measures o f TCDD were obtained shortly after the accident. TCDD levels were also extrapolated to the time o f conception. However, it is difficult to identify the relevant time interval over w hich TCDD dose should be considered for dose-response analysis. The critical window o f exposure is unknown. Conclusion The data from this study demonstrate a positive dose-response relationship w ith paternal TCDD levels at the time of the accident and increased likelihood for female births. However, It is difficult to identify the relevant time interval over w hich TCDD dose should be considered; specifically, it is difficult to discern w hether this effect is a consequence o f the initial high exposure during childhood or a function o f the cumulative exposure for this entire exposure window beginning at the early age. Using the initial exposures in a dose-response model would yield LOAELs that are too high to be relevant to factor into the RfD calculation. Dose-response analysis for this outcome is, therefore, was not conducted. 1 2 3 4 5 This document is a draftfor review purposes only and does not constitute Agency policy. B-47 DRAFT--DO NOT CITE OR QUOTE 1 Table B-37. Baccarelli et al., 2008--Neonatal thyroid function 2 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. M easures o f b-TSH are taken using a standardized protocol 72 hours after birth. These b-TSH measures are taken on all newborns born in the region o f Lombardy of w hich Seveso if a part of. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied for com ponent o f the study based on plasm a dioxin measures. For the com parisons involving place o f residence at the tim e o f the accident, exposure m isclassification is likely given variability in soil TCDD exposure levels w ithin these areas. 3. C onsideration Study dem onstrates an association b etw een T C D D an d adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. M ean neonatal b-TSH was 0.98pU/m l [0.90-1.08] in the reference area, 1.35pU/ml [1.22-1.49] in zone B, and 1.66pU/ml [1.19-2.31] in zone A (p < 0.001). The plotted frequency distributions have similar shapes, but have shifted to the right for areas of higher exposures. Neonatal b-TSH was correlated w ith current m aternal plasm a TCDD (P-0.47, p < 0.001) in the 51 newborns for w hich individual m aternal serum TCDD values w ere available. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. TEQs were measured among the 38 w om en for w hich serum samples were available and were defined for a mixture o f dioxin-like compounds. M aternal m ean total TEQs (PCDDs, PCDFs, coplanar PCBs, and noncoplanar PCBs) was 41.8 ppt. Two measures of exposure included place o f residence at time o f accident and plasm a samples obtained from mothers at the time o f delivery. Similarities in positive dose-response relationships give stronger weight to the findings. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk an d ensure adequate statistical power. Response C onsideration satisfied fo r exposure m etric that w as b ased on `place o f resid en ce'. F o r plasm a based estimate o f m aternal TCDD there w ere only 51 m other-child pairs. Only seven children in total were found to have b-TSH levels in excess o f 5 uU/ml; this implies lim ited statistical power involving this health outcome. 1. C riteria Response 2. Criteria Study is published in the peer-review ed scientific literature and has an appropriate discussion of the strengths and limitations. Criteria satisfied. PLOS M edicine 2008; 5(7)1133-1142. The authors discuss the strength o f the study related to characterization o f exposure using serum sampling, and ability to adjust for factors related to b-TSH or TCDD levels (gender, birth weight, birth order, m aternal age, hospital and type o f delivery). They also highlight that a lim itation o f study was that the influence of m other-child dioxin transfer through colostrum could not be assessed because no information on breastfeeding before b-TSH m easurem ent was available. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. This document is a draftfor review purposes only and does not constitute Agency policy. B-48 DRAFT--DO NOT CITE OR QUOTE Response 3. C riteria Response Criteria satisfied. In the population-based study, eligible w om en who resided in zones A and B at the time o f the accident (n = 1,772) w ere m atched to nonexposed women. In the study based on plasm a dioxin m easurem ents, participants w ere the 51 children b o rn to 38 w o m en fro m zones A, B, R, or a reference zone for w hich plasm a dioxin measurements were available. The effective dose and oral exposure can be reasonably estimated and the measures o f exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) o f exposure examined. Response has to be a nonfatal endpoint. Criteria satisfied. M aternal TCDD levels were estimated at the time o f delivery based on plasm a samples, and the critical window o f exposure can be defined as the 9 m onth gestation period. Conclusion The data provide an opportunity for quantitative dose-response analyses. 1 2 3 Table B-38. A laluusua et al., 2004-- Oral hygiene 4 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. Ascertainm ent o f dental health was done blind to place o f residence, used standard protocol for caries developed by the WHO, and the clinical examination supplemented by radiographic examination. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Additional risk factor inform ation was collected on questionnaires. These factors were considered as adjustm ent factors. Findings potentially susceptible to participation biases. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. Increased prevalence o f developmental enamel effects found w ith increased TCDD serum measures. Namely, prevalence in unexposed region was 26%, w hereas in the low, m iddle, and high TCCD groups the prevalence w as 10, 40, and 60%, respectively. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. TCDD exposure level based on serum lipids. No discussion of exposure levels in reference population. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Criteria satisfied. Despite small numbers, statistically significant findings were achieved. 1. C riteria Study is published in the peer-reviewed scientific literature and has an appropriate discussion of the strengths and limitations. This document is a draftfor review purposes only and does not constitute Agency policy. B-49 DRAFT--DO NOT CITE OR QUOTE Response 2. Criteria Response 3. C riteria Response Criteria satisfied. Environm ental H ealth Perspectives, 2004, 112(13)1313-1318. Authors m ention two im portant strength o f the study: characterization o f TCDD exposure using serum collected shortly after the time o f the accident, and the fact that developmental defects are perm anent in nature. Therefore, they represent a health outcome can evaluated years later. Little discussion was made o f the im pact o f differential compliance rates betw een the exposed (74%) and nonexposed (58%) groups. Authors m ention two im portant strength o f the study: characterization o f TCDD exposure using serum collected shortly after the time o f the accident, and the fact that developm ental defects are perm anent in nature. Therefore, they represent a health outcome can evaluated years later. Little discussion was made o f the impact of differential com pliance rates betw een the exposed (74%) and nonexposed (58%) groups. Exposure must be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. Serum levels o f TCDD could be estimated for children in exposed areas. No serum levels were available for reference group o f children, and assum ption o f zero exposure was made. This seems reasonable. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate w indow(s) o f exposure examined. Response has to be a nonfatal endpoint. Criteria satisfied. It is difficult to discern w hether this effect is a consequence o f the initial high exposure during childhood or a function o f the cumulative exposure o f the entire exposure window beginning at early age. However, assumptions can be made regarding the critical window o f exposure and the relevant dose can be calculated. Conclusion The considerations for conducting a dose-response analysis have been satisfied w ith the study population o f only those subjects who lived in the ABR zone at the time of the accident; exposure data are unavailable for those in the referent area. W hile is difficult to identify the relevant time interval over w hich TCDD dose should be considered, quantitative dose-response analysis for this outcome was conducted. 1 2 3 Table B-39. Bertazzi et al., 2001-- M ortality (noncancer) 4 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied for some causes o f death, but not others. M ortality appears to be well captured from the vital statistics registries in the region (99% complete). Some health outcom es (e.g., diabetes) are subject to m isclassification using death certificate data. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Although individual-level data for individual risk factors are not available, the potential for confounding is likely minimal. For e.g., independent surveys suggests similarity betw een smoking behaviors across the regions. Exposure m isclassification based on place o f residency likely to bias risk estim ates towards the null. 3. C onsideration Study dem onstrates an association b etw een TC D D and adverse h ealth effect w ith evidence o f an exposure-response relationship. This document is a draftfor review purposes only and does not constitute Agency policy. B-50 DRAFT--DO NOT CITE OR QUOTE Response Consideration not satisfied for m ost causes o f death. A n exception was the dose-response relationship was observed for chronic obstructive pulmonary disease across Zones A, and B. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration not satisfied. Exposure classification was based on the address o f the residence on the date of the accident or w hen the person first entered the area. Although TCDD blood levels were also measured, these were not examined w ith respect to health outcomes. The lack o f individual-level data also precluded an exam ination o f these uncertainties. 5. C onsideration Study size and follow -up are large enough to y ield p recise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. A total o f 494 noncancer deaths were found among residents of Zones A, and B, respectively. This allowed exam ined o f gender-specific effects. 1. C riteria Response 2. Criteria Response 3. Critieria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Am J Epidemiol, 2001, 153:1031-1044. Authors discuss lack of individual-level exposure data and other risk factors (e.g., smoking), difficulties in extrapolating to background levels, diagnostic accuracy o f using death certificates. Strengths included similarities betw een exposed and comparison population for several risk factors, com pleteness o f follow-up, and consistent m ethods to identify mortality outcomes in the exposed and comparison populations. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria not satisfied. Individual-level exposure data are unavailable. Exposure based on place of residence at time o f the explosion. Soil sam pling perform ed indicated considerable variability in TCDD levels w ithin each region. In addition, place o f residency at tim e of explosion does not ensure individuals were at their home around the time of the accident. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. A n ecological measure o f exposure (region o f residency at time of accident) was used to categorize individuals according to their possible exposure. Latencies were considered. W hile such an approach has value for identifying whether excesses occurred among highly exposed populations, it is not precise enough to conduct a quantitative doseresponse analysis. Furtherm ore, noncancer mortality is not a viable endpoint to consider for further dose-response analysis. Conclusion 1 2 3 4 5 Study is not suitable for dose-response analysis due to m ortality as endpoint and lack o f individual-level exposure data. This document is a draftfor review purposes only and does not constitute Agency policy. B-51 DRAFT--DO NOT CITE OR QUOTE 1 Table B-40. Consonni et al., 2008--Mortality (noncancer) 2 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied for some causes o f death, but not others. M ortality appears to be well captured from the vital statistics registries in the region (99% complete). Some health outcom es (e.g., diabetes) are subject to m isclassification using death certificate data. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. Although individual-level data for individual risk factors are not available, the potential for confounding is likely minimal. For e.g., inform ation from other independent surveys suggests similarity betw een smoking behaviors across the regions. Exposure m isclassification based on place o f residency is likely to bias risk estim ates towards the null. 3. C onsideration Study dem onstrates an association b etw een TC D D and adverse h ealth effect w ith evidence o f an exposure-response relationship. Response Consideration not satisfied. Statistically significant association noted in m ost highly exposed area for chronic rheumatic disease and chronic obstructive pulmonary disease. Dose-response pattern noted across Zones A, B and R for circulatory disease mortality 5-9 years after the accident. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration not satisfied. Lack o f individual-level data precludes an examination o f these uncertainties. 5. C onsideration Study size and follow -up are large enough to y ield p recise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied for some causes o f death but not others. For example, only three deaths from diabetes occurred among residents o f Zone A. The lim itation related to statistical pow er is exacerbated for stratified analyses carried out by num ber o f years since the accident. 1. C riteria Response 2. Criteria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Am J Epidemiol, 2008, 167:847-858. Authors discuss potential for selection bias, lim itation o f residential based measure of exposure, similarities o f mortality ascertainm ent in exposed and referent populations, and multiple testing. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria not satisfied. Individual-level exposure data are unavailable. Exposure based on place of residence at time o f the explosion. Soil sam pling perform ed indicated considerable variability in TCDD levels w ithin each region. In addition, place o f residency at tim e of explosion does not ensure individuals were at their home around the time of the accident. This document is a draftfor review purposes only and does not constitute Agency policy. B-52 DRAFT--DO NOT CITE OR QUOTE 3. C riteria Response The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. A n ecological measure o f exposure (region o f residency at time of accident) was used to categorize individuals according to their possible exposure. Latencies were considered. W hile such an approach has value for identifying whether excesses occurred among highly exposed populations, it is not precise enough to conduct a quantitative dose-response analysis. Furthermore, noncancer mortality is not a viable endpoint to consider for further dose-response analysis. Conclusion Study is not suitable further dose-response evaluation due to noncancer morality endpoint. 1 2 3 Table B-41. Baccarelli et al., 2005-- Chloracne 4 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. Chloracne cases identified using standardized criteria. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. 3. C onsideration Study dem onstrates an association b etw een TC D D and adverse h ealth effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. Plasm a TCDD was associated w ith an increased risk o f chloracne. The odds ratios increased in a dose-response pattern across zone o f residence. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. Authors discussed implications o f differential elim ination rates by age and body growth. 5. C onsideration Study size and follow -up are large enough to y ield p recise estim ates o f risk and ensure adequate statistical power. Response C onsideration satisfied. A total o f 101 chloracne cases w ere identified, and 211 controls w ere selected. Statistically significant findings were observed in several comparisons. 1. C riteria Response 2. Criteria Study is published in the peer-reviewed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. British Journal o f Dermatology, 2005, 152, 459-465. The authors detail the lim ited statistical pow er they had available in the study. They also highlight a strength o f the study that included uniqueness o f age and sex distribution o f chloracne cases, characterization o f TCDD that could be done using sera samples, and availability o f both clinical and epidem iological data. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. This document is a draftfor review purposes only and does not constitute Agency policy. B-53 DRAFT--DO NOT CITE OR QUOTE Response 3. C riteria Response Criteria satisfied. TCDD was estimated in both chloracne cases and control using serum measures. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Response has to be a nonfatal endpoint. Criteria satisfied. Serum based m easures of TCDD were obtained shortly after the accident. Chloracne is thought to be caused by the initial high exposure. Conclusion Exposure to TCDD at sufficiently high levels is recognized to cause chloracne. This study provides lim ited relevance to dose-response modeling of TCDD as exposure levels typically observed in the general population are m uch lower. 1 2 3 Table B-42. Baccarelli et al, 2002 and 2004-- Im m unological effects 4 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. Com m on m ethods were used to describe blood levels o f plasm a im m unoglobulins (IgA, IgG, and IgM ) and com plem ent com ponents (C3 and C4). 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. B oth exposure and outcome were objectively and accurately measured. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse h ealth effect w ith evidence o f an exposure-response relationship. Response Consideration satisfied. Plasm a IgG levels were inversely related w ith TCDD. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. B oth categorical (quintiles) and continuous measures of TCDD were exam ined in the dose-response analysis. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. Analyses are made using 72 highly exposed, and 72 low exposed individuals. 1. C riteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Toxicology letters, 2004, 149:287-293 and Environ H ealth Perspect, 2002, 110(12):1169-1173. The authors highlight that few studies have looked at immunological effects o f TCDD in humans, that the current study was able to exclude those w ith concurrent medical conditions, and the ability to characterize exposure using serum measures. Lim itations addressed were the uncertainty about the clinical relevance of the dose-response pattern found, and the relatively small size o f the study population. This document is a draftfor review purposes only and does not constitute Agency policy. B-54 DRAFT--DO NOT CITE OR QUOTE 2. Criteria Response 3. C riteria Response Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. A total o f 120 subjects were included in the study. This included 62 randomly selected from the high exposed zone, and 58 selected from the reference area. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. Dose-response relationships were examined using current TCDD levels. However, it is difficult to identify the relevant time interval over w hich TCDD dose should be considered for dose-response analysis. Conclusion An inverse dose-response association between IgG and TCDD was observed, however, because the relationship can not be described in term s of clinical relevance w ith respect to a specific health outcome, it is our view that these data are not suited to dose-response modeling. 1 2 3 B.2.6. Chapaevsk Study 4 5 Table B-43. Revich et al., 2001-- M ortality (noncancer) and reproductive 6 health 7 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration cannot be evaluated. Insufficient details are provided in the paper to gauge the com pleteness and coverage o f the cancer registry and mortality data. H ealth outcom es were studied on the basis o f inform ation in the official m edical statistics 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration not satisfied. It is an ecological study. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse h ealth effect w ith evidence o f an exposure-response relationship. Response Consideration cannot be evaluated. Dose-response was not evaluated as exposure was based on residency in the region vs. no residency. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration not satisfied. No individual-level exposure estimates were used. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. Population-based data over several years were used to make ecological com parisons. This document is a draftfor review purposes only and does not constitute Agency policy. B-55 DRAFT--DO NOT CITE OR QUOTE 1. C riteria Response 2. Criteria Response 3. C riteria Response Conclusion Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. Published in Chemosphere, 2001, 43(4-7):951-966. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria not satisfied. It is a cross-sectional study that compares mortality rates betw een regions. No individual-level exposure data available. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. No exposure estimates were used in the study. These cancer data are cross-sectional in nature and not appropriate for a dose-response analysis. 1 B.2.7. Air Force Health ("Ranch H ands") Study 2 3 Table B-44. M ichalek and Pavuk, 2008-- Diabetes 4 1. C onsideration M ethods used to ascertain h ealth outcom es identified w ere unbiased, highly sensitive, and specific. Response 2. Consideration Consideration satisfied. Prevalent diabetes identified from medical records from repeated medical check-ups. Preferred m ethod o f ascertaining outcome relative to use o f death certificates. Risk estimates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response 3. C onsideration Consideration not satisfied. Adjustm ent was made for a num ber of risk factors related to diabetes (e.g., BM I, family history, smoking). However, A gent Orange was a 50% mixture of 2,4-D and TCDD; therefore, potential for confounding by other coexposures is likely. Study dem onstrates an association betw een TCDD and adverse health effect w ith evidence of an exposure-response relationship. Response C onsideration satisfied. T he R R fo r an increase in 10 units w as 1.29 ( p < 0.001), and the risks across the background, low and high exposure categories, relative to the unexposed were 0.86, 1.45, and 1.68. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. Initial TCDD dose were estim ated at the end o f the tour o f duty for the Ranch Hands. Individual-level serum dioxin measurements correlated well w ith correlated w ith days o f spraying and calendar period of service, but collection o f the samples roughly 20 years later required back-extrapolation. 5. C onsideration Study size and follow -up are large enough to y ield p recise estim ates o f risk an d ensure adequate statistical power. This document is a draftfor review purposes only and does not constitute Agency policy. B-56 DRAFT--DO NOT CITE OR QUOTE Response Consideration satisfied. There were a total o f 439 cases o f diabetes identified. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. Criteria satisfied. J Occup Environ M edicine, 2008, 50:330-340. The authors address strengths and limitations related to the accuracy of the one-compartment pharmacokinetic model, im pact o f the covariate time spent in Southeast Asia, and potential exposure misclassification on days sprayed. Exposure m ust be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria not satisfied. TCDD estimates were derived using serum samples. However, Ranch H and veterans were exposed to other compounds in the herbicides, such as 2,4-D. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Response has to be a nonfatal endpoint. Criteria satisfied. TCDD levels at the end o f service were estimated. Extrapolation was done using a half-life o f 7.6 years. Exposures were grouped into comparison, background, low and high. This allows for a shape o f the dose-response curve to be evaluated. A continuous m easure o f TC D D w as also exam ined (log10TCD D ). Conclusion Ranch Hand veterans were exposed to other contaminants in the herbicides that were mixed, thereby m aking it difficult to determine independent effects o f TCDD on diabetes. In our view, this limitation precludes dose-response modeling of TCDD and diabetes using data from this cohort. 1 2 3 B.2.8. O ther Noncancer Studies of Dioxin 4 5 Table B-45. M cBride et al., 2009a-- M ortality (noncancer) 6 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. The New Zealand H ealth Inform ation Service M ortality Collection and the R egistrar-G eneral's Index to D eaths w ere used to identify deaths. A dditional searches were based on the last know n address from the work record; the electoral roll and the habitation index; the telephone book; the internet; and Terranet property inform ation database. A n additional search was carried out through the Births, Deaths, and M arriages office o f the New Zealand Departm ent o f Internal Affairs. Lastly, automated personnel and pension records were also used to locate past New Plym outh workers and identify some deaths. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration satisfied. W orkers lost to follow-up were an unlikely source o f bias especially for internal analyses. Confounding by other coexposures (e.g., 2,4,6-TCP) unlikely to have resulted in bias, due to presumed poor correlation w ith TCDD. This document is a draftfor review purposes only and does not constitute Agency policy. B-57 DRAFT--DO NOT CITE OR QUOTE 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration not satisfied. There was no cause of death among those considered for w hich a dose-response trend was observed across four exposure categories o f TCDD. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. Dichotomous exposure (exposed/unexposed) and duration of employment were examined from job exposure classification assessed via occupational history records industrial hygienists/factory personnel knowledge and questionnaires. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration not satisfied. 1. C riteria Response 2. Criteria Response 3. C riteria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion of the strengths and limitations. Criteria satisfied. Published in J Occup Environ Med, 2009, 51:1049-1056. The other studies in the cohort highlight the 22% o f the cohort lost to follow-up, the lim ited size o f the cohort tissue sarcomas, differences in cohort definitions betw een sprayers and producers, and the potential for other exposures during em ploym ent at the plant. Exposure must be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria satisfied. Serum measures available for 346 workers were used to derive TCDD exposures for the entire cohort using the area under the curve approach. The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate w indow(s) o f exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. Dichotomous exposure assessm ent did not allow individual estimates of dose to be developed. However, noncancer mortality is not a viable endpoint to consider for further dose-response analysis. Conclusion 1 2 A considerable portion o f the cohort was lost to follow-up, and no dose-response associations noted. As a result, the data are not suited to dose-response analysis. This document is a draftfor review purposes only and does not constitute Agency policy. B-58 DRAFT--DO NOT CITE OR QUOTE 1 Table B-46. M cBride et al., 2009b-- M ortality (noncancer) 2 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration satisfied. The New Zealand H ealth Inform ation Service M ortality Collection and the R egistrar-G eneral's Index to D eaths w ere used to identify deaths. A dditional searches were based on the last know n address from the work record; the electoral roll and the habitation index; the telephone book; the internet; and Terranet property inform ation database. A n additional search was carried out through the Births, Deaths, and M arriages office o f the New Zealand Departm ent o f Internal Affairs. Lastly, automated personnel and pension records were also used to locate past New Plym outh workers and identify some deaths. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration not satisfied. Considerable am ount o f workers were lost to follow up (22%), but it is unclear if bias resulted. The dichotom ous exposure m easure was based on exposure to TCDD, chlorinated dioxins and phenoxy herbicides, so confounding is a possibility by these coexposures. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration not satisfied. Because no individual exposure estimates were available for these analyses, dose-response could not be evaluated. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration satisfied. Consideration satisfied. Dichotomous exposure (exposed/unexposed) and duration o f employment were examined from job exposure classification assessed via occupational history records industrial hygienists/factory personnel knowledge and questionnaires. Authors discuss limitations in the assignment of exposure among cohort members. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. 1. C riteria Response 2. Criteria Response Study is published in the peer-reviewed scientific literature and has an appropriate discussion of the strengths and limitations. Criteria satisfied. Published in Occup M edicine, 2009, 59(4):255-263. The authors highlight cohort lost to follow-up, the limited size o f the cohort, differences in cohort definitions between sprayers and producers, and the potential for other exposures during employment at the plant. Exposure must be prim arily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria not satisfied. Exposures were not quantified. The dichotomous exposure measure was based on exposure to TCDD, chlorinated dioxins and phenoxy herbicides. This document is a draftfor review purposes only and does not constitute Agency policy. B-59 DRAFT--DO NOT CITE OR QUOTE 3. C ritiera Response The effective dose and oral exposure can be reasonably estimated and the measures of exposure are consistent w ith the current biological understanding of dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate w indow(s) o f exposure examined. Response has to be a nonfatal endpoint. Effective dose could not be estim ated given the lack o f individual-level exposure data. N oncancer mortality is not a viable endpoint to consider for further dose-response analysis. Conclusion The study lacks the quantification o f exposures at an individual level, and a considerable portion o f the cohort was lost to follow-up. As a result, the data are not suited to dose-response analysis. 1 2 3 Table B-47. Ryan et al., 2002-- Sex ratio 4 1. C onsideration M ethods used to ascertain health outcom es identified w ere unbiased, highly sensitive, and specific. Response Consideration not satisfied. Company records were used to identify births, the date o f birth, and the sex o f the child. No information was provided on the expected completeness of identifying births in this manner. M oreover, the study was expanded to include workers who heard about the study in a public forum. Therefore, the study could be influenced by participation bias. 2. Consideration R isk estim ates are not susceptible to biases from confounding exposures or from study design or statistical analysis. Response Consideration not satisfied. See above. 3. C onsideration Study dem onstrates an association betw een TC D D and adverse health effect w ith evidence o f an exposure-response relationship. Response Consideration not satisfied. The study com pared birth ratios am ong m en and w om en employed at the plant to the general population. No categories of exposure were examined. 4. Consideration Exposure assessm ent m ethodology is clear and adequately characterizes individual-level exposures. The limitations and uncertainties in the exposure assessm ent are considered. Response Consideration not satisfied. This is not relevant as no analyses were done in relation to exposure levels. 5. C onsideration Study size and follow -up are large enough to y ield precise estim ates o f risk and ensure adequate statistical power. Response Consideration satisfied. For the categories o f exposure used (yes/no), and the stratified analyses by sex and subcohort, the study allows for the birth ratios to be estimated w ith sufficient precision. 1. C riteria Study is published in the peer-review ed scientific literature and has an appropriate discussion o f the strengths and limitations. This document is a draftfor review purposes only and does not constitute Agency policy. B-60 DRAFT--DO NOT CITE OR QUOTE Response 2. Criteria Response 3. C riteria Response Criteria not satisfied. Published in Environ Health Perspect, 2002, 110(11):A699-A701. The authors discussed the limitations of using serum collected many years after they stopped working to estimate TCDD exposures w hen the preferred metric would be TCDD levels at the tim e o f conception. They did not address issues about the representativeness of the study participants to the entire cohort o f workers, nor did they address the limitation o f not being able to conduct dose-response analyses using individual-level TCDD data. Exposure m ust be primarily TCDD and is properly quantified so that dose-response relationships can be assessed. Criteria not satisfied. W hile serum m easures were available for 84 o f the 198 participants of the study, birth ratios were com pared betw een the cohort o f 2,4,5-T and 2,4,5-trichlorphgenol w orkers relative to the city o f Ufa. There was no attem pt to derive birth ratios in relation to exposure levels. The serum data were only used to demonstrate that these workers, on average, had TCDD levels 30 times higher than U fa residents. The effective dose and oral exposure can be reasonably estimated and the measures o f exposure are consistent w ith the current biological understanding o f dose. The reported dose is consistent w ith a toxicologically relevant dose. Latency and appropriate window(s) of exposure examined. Response has to be a nonfatal endpoint. Criteria not satisfied. TCDD exposures were based on serum measures taken in some cases many years after children were born; no attem pt was made to back-extrapolate to the time of conception. Conclusion The data are not suitable for dose-response modeling. Risk estimates have not been derived in relation to TCDD exposure levels. There exist uncertainties about the representativeness o f the participants in relation to the cohort as a whole, and insufficient details are provided to evaluate the extent in w hich all births were identified. W hile these data should not be used for quantitative dose-response m odeling, the m uch low er M /F birth ratio am ong exposed fathers is consistent w ith the finding by M ocarelli et al, and lends support to those findings. 1 2 3 B.3. REFERENCES 4 Akhtar, FZ; Garabrant, DH; Ketchum , NS; et al. (2004) C ancer in US A ir Force veterans o f the V ietnam War. J 5 Occup Environ M ed 46(2):123-136. 6 A laluusua, S; Calderara, P; G erthoux, PM ; et al. (2004) D evelopm ental dental aberrations after the d ioxin accident 7 in Seveso. Environ H ealth Perspect 112(13):1313-1318. Aylward, LL; Brunet, RC; Starr, TB; et al. (2005a) Exposure reconstruction for the TCD D -exposed N IO SH cohort using a concentration- and age-dependent model o f elimination. Risk Anal 25(4):945-956. Aylward, LL; Brunet, RC; Carrier, G; et al. (2005b) C oncentration-dependent TCDD elim ination kinetics in humans: toxicokinetic modeling for m oderately to highly exposed adults from Seveso, Italy, and Vienna, Austria, and im pact on dose estim ates for the NIOSH cohort. J Expo Anal Environ Epidem iol 15(1):51-65. 8 Baccarelli, A; M ocarelli, P; Patterson, DG, Jr.; et al. (2002) Im m unologic effects o f dioxin: new results from Seveso 9 and comparison w ith other studies. Environ H ealth Perspect 110(12):1169-1173. 1 0 Baccarelli, A; Pesatori, AC; M asten, SA; et al. (2004) A ryl-hydrocarbon receptor-dependent pathw ay and toxic 11 effects o f TCDD in humans: a population-based study in Seveso, Italy. Toxicol Lett 149(1-3):287-293. This document is a draftfor review purposes only and does not constitute Agency policy. B-61 DRAFT--DO NOT CITE OR QUOTE 1 Baccarelli, A; Pesatori, AC; Consonni, D; et al. (2005) H ealth status and plasm a dioxin levels in chloracne cases 2 20 years after the Seveso, Italy accident. Br J Dermatol 152(3):459-465. 3 Baccarelli, A; Hirt, C; Pesatori, AC; et al. (2006) t(14;18) translocations in lym phocytes o f healthy dioxin-exposed 4 individuals from Seveso, Italy. Carcinogenesis 27(10):2001-2007. 5 Baccarelli, A; Giacomini, SM; Corbetta, C; et al. (2008) N eonatal thyroid function in Seveso 25 years after m aternal 6 exposure to dioxin. PLoS M ed 5(7):1133-1142. 7 Becher, H; Steindorf, K; Flesch-Janys, D. (1998) Quantitative cancer risk assessm ent for dioxins using an 8 occupational cohort. Environ H ealth Perspect 106(Suppl 2):663-670. 9 Bertazzi, PA; Consonni, D; Bachetti, S; et al. (2001) H ealth effects o f dioxin exposure: a 20-year m ortality study. 10 Am J Epidem iol 153(11):1031-1044. 11 Cheng, H; Aylward, L; Beall, C; et al. (2006) TCDD exposure-response analysis and risk assessment. R isk Anal 12 26:1059-1071. 13 Collins, JJ; Bodner, K; Aylward, LL; et al. (2009) M ortality rates am ong trichlorophenol workers w ith exposure to 14 2,3,7,8-tetrachlorodibenzo-p-dioxin. Am J Epidem iol 170(4):501-506. 15 Consonni, D; Pesatori, AC; Zocchetti, C; et al. (2008) M ortality in a population exposed to dioxin after the Seveso, 16 Italy, accident in 1976: 25 years o f follow-up. Am J Epidemiol 167(7):847-858. 17 Eskenazi, B; W arner, M; M ocarelli, P; et al. (2002a) Serum dioxin concentrations and m enstrual cycle 18 characteristics. Am J Epidemiol 156(4):383-392. 19 Eskenazi, B; M ocarelli, P; W arner, M ; et al. (2002b) Serum dioxin concentrations and endom etriosis: a cohort study 2 0 in Seveso, Italy. Environ H ealth Perspect 110(7):629-634. 21 Eskenazi, B; M ocarelli, P; W arner, M ; et al. (2003) M aternal serum dioxin levels and birth outcom es in w om en o f 2 2 Seveso, Italy. Environ H ealth Perspect, 111(7), 947-953. 2 3 Eskenazi, B; W arner, M; M arks, AR; et al. (2005) Serum dioxin concentrations and age at menopause. Environ 2 4 Health Perspect 113(7):858-862. 2 5 E skenazi, B; W arner, M ; Sam uels, S; et al. (2007) Serum d ioxin concentrations and risk o f uterine leiom yom a in the 2 6 Seveso W omen's H ealth Study. A m J Epidem iol 166(1):79-87. 2 7 Fingerhut, M A; Halperin, W E; M arlow , DA; et al. (1991) C ancer mortality in w orkers exposed to 2 8 2,3,7,8-tetrachlorodibenzo-p-dioxin. N Engl J M ed 324(4):212-218. 2 9 Flesch-Janys, D; Berger, J; Gum, P; et al. (1995) Exposure to polychlorinated dioxins and furans (PCDD/F) and 3 0 mortality in a cohort o f workers from a herbicide-producing plant in Hamburg, Federal Republic of Germany. Am J 31 Epidemiol 142(11):1165-1175. 3 2 Flesch-Janys, D; Becher, H; Gum , P; et al. (1996) E lim ination o f polychlorinated dibenzo-p-dioxins and 33 dibenzofurans in occupationally exposed persons. J Tox Environ H ealth 47(4):363-378. 3 4 Flesch-Janys, D; Steindorf, K; Gum , P; et al. (1998) Estim ation o f the cum ulated exposure to polychlorinated 3 5 dibenzo-p-dioxins/furans and standardized mortality ratio analysis o f cancer mortality by dose in an occupationally 3 6 exposed cohort. Environ H ealth Perspect 106(Suppl 2):655-662. 3 7 Hooiveld, M; H eederik, DJ; Kogevinas, M; et al. (1998) Second follow -up o f a D utch cohort occupationally 3 8 exposed to phenoxy herbicides, chlorophenols, and contaminants. Am J Epidemiol 147(9):891-901. This document is a draftfor review purposes only and does not constitute Agency policy. B-62 DRAFT--DO NOT CITE OR QUOTE 1 M anz, A; Berger, J; Dwyer, JH; et al. (1991) Cancer m ortality am ong w orkers in chem ical plant contam inated w ith 2 dioxin. Lancet 338(8773):959-964. 3 M cBride, DI; Collins, JJ; Hum phry, NF; et al. (2009a) M ortality in w orkers exposed to 4 2,3,7,8-tetrachlorodibenzo-p-dioxin at a trichlorophenol plant in New Zealand. J Occup Environ M ed 5 51(9):1049-1056. 6 M cBride, DI; Burns, CJ; Herbison, GP; et al. (2009b) M ortality in em ployees at a New Zealand agrochem ical 7 m anufacturing site. Occup M ed (Oxford, England) 59(4):255-263. 8 M ichalek, JE; Pavuk, M. (2008) Diabetes and cancer in veterans of Operation Ranch Hand after adjustment for 9 calendar period, days o f spraying, and time spent in Southeast Asia. J Occup Environ M ed 50(3):330-340. 10 M ocarelli, P; Gerthoux, PM ; Ferrari, E; et al. (2000) Paternal concentrations o f dioxin and sex ratio o f offspring. 11 Lancet 355(9218):1858-1863. 12 M ocarelli, P; Gerthoux, PM ; Patterson, DG, Jr.; et al. (2008) D ioxin exposure, from infancy through puberty, 13 produces endocrine disruption and affects hum an semen quality. Environ H ealth Perspect 116(1):70-77. 14 Ott, M G; Zober, A. (1996) Cause specific mortality and cancer incidence am ong em ployees exposed to 15 2,3,7,8-TCDD after a 1953 reactor accident. Occup Environ M ed 53(9):606-612. 1 6 Pesatori, AC; C onsonni, D; B achetti, S; et al. (2003) Short- and long-term m orbidity and m ortality in the population 17 exposed to dioxin after the " Seveso accident" . Ind H ealth 4 1 (3 ):1 2 7 -138. 18 Revich, B; Aksel, E; U shakova, T; et al. (2001) D ioxin exposure and public health in Chapaevsk, Russia. 19 Chemosphere 43(4-7):951-966. 2 0 Ryan, JJ; Amirova, Z; Carrier, G. (2002) Sex ratios o f children o f R ussian pesticide producers exposed to dioxin. 21 Environ Health Perspect, 110(11):A699-701. 2 2 Steenland, K; Piacitelli, L; Deddens, J; et al. (1999) Cancer, heart disease, and diabetes in w orkers exposed to 23 2,3,7,8-tetrachlorodibenzo-p-dioxin. J Natl Cancer I 91(9):779-786. 2 4 Steenland, K; Deddens, J; Piacitelli, L. (2001) R isk assessm ent for 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) 2 5 based on an epidem iologic study. A m J Epidem iol 154(5):451-458. 2 6 't M annetje, A; M cL ean, D; C heng, S; et al. (2005) M ortality in N ew Z ealand w orkers exposed to phenoxy 2 7 herbicides and dioxins. Occup Environ M ed 62(1):34-40. 2 8 W arner, M; Eskenazi, B; M ocarelli, P; et al. (2002) Serum dioxin concentrations and breast cancer risk in the 2 9 Seveso W om en's H ealth Study. E nviron H ealth P erspect 110(7):625-628. 3 0 W arner, M ; Sam uels, S; M ocarelli, P; et al. (2004) Serum d ioxin concentrations and age at m enarche. E nviron 31 Health Perspect 112(13):1289-1292. 3 2 W arner, M; Eskenazi, B; Olive, DL; et al. (2007) Serum dioxin concentrations and quality o f ovarian function in 33 wom en o f Seveso. Environ H ealth Perspect 115(3):336-340. 3 4 Zober, A; M esserer, P; Huber, P. (1990) Thirty-four-year mortality follow-up o f BASF employees exposed to 3 5 2,3,7,8-TCDD after the 1953 accident. Int A rch Occup Environ H ealth 62(2):139-157. This document is a draftfor review purposes only and does not constitute Agency policy. B-63 DRAFT--DO NOT CITE OR QUOTE [T his p ag e in ten tio n ally left blank.] DRAFT D O N O T C IT E O R Q U O T E M ay 2010 E xternal R ev iew D raft APPENDIX C Kinetic Modeling N O T IC E T H IS D O C U M E N T IS A N E X T E R N A L R E V IE W D R A F T . It has n o t been form ally released by the U .S. E n v iro n m en tal P ro tectio n A g en cy and should n o t at th is stage b e co n stru ed to rep resen t A gency policy. It is b ein g circu lated fo r co m m en t on its technical accuracy and policy im plications. N ational C en ter fo r E n v iro n m en tal A ssessm ent O ffice o f R esearch and D evelopm ent U .S. E n v iro n m en tal P ro tectio n A g en cy C incinnati, O H CONTENTS--APPENDIX C: Kinetic Modeling A P P E N D I X C . k i n e t i c M o d e l i n g .....................................................................................................................................C -1 C .1. L IT E R A T U R E S E A R C H S T R A T E G Y A N D R E S U L T S -- ID E N T IF Y IN G R E C E N T PU B L IC A T IO N S F O R U PD A T IN G TC D D T O X IC O K IN E T IC M O D E L I N P U T P A R A M E T E R S ................................................................................................................ C -1 C .1 .1 . D a t a B a s e s S e a r c h e d ............................................................................................................................C -1 C .1 .2 . L i t e r a t u r e S e a r c h S tr a t e g y a n d A p p r o a c h ............................................................................. C - 2 C .1 .2 .1 . C h e m i c a l S e a r c h T e r m s -- D I A L O G S e a r c h ................................................. C - 2 C .1 .2 .2 . P r i m a r y S e a r c h T e r m s ( S p e c i e s ) -- D I A L O G S e a r c h ............................. C - 2 C .1 .2 .3 . S e c o n d a r y S e a r c h T e r m s ( T o x i c o l o g y ) -- D I A L O G S e a r c h ...............C -3 C .1 .3 . C i t a t i o n S c r e e n i n g P r o c e d u r e s a n d R e s u l t s ........................................................................... C -3 C .1.4. R eferen ces S elected fo r M o re D etailed R ev iew fo r U p d atin g th e P B P K M o d e l s ........................................................................................................................................................... C - 6 C .1 .5 . B r i e f D e s c r i p t i o n s o f D I A L O G B i b l i o g r a p h i c D a t a B a s e s S e a r c h e d ..................C - 8 C .2 . T O X I C O K I N E T I C M O D E L I N G C O D E ..............................................................................................C -1 1 C .2 .1 . H u m a n S ta n d a r d M o d e l ................................................................................................................... C -1 1 C .2 .1 .1 . M o d e l C o d e ..................................................................................................................... C -1 1 C .2 .1 .2 . I n p u t F i l e ...........................................................................................................................C - 1 9 C .2 .2 . H u m a n G e s t a t i o n a l M o d e l ............................................................................................................. C - 2 0 C .2 .2 .1 . M o d e l C o d e ..................................................................................................................... C - 2 0 C .2 . 2 . 2 . I n p u t F i l e ...........................................................................................................................C -3 1 C .2 .3 . R a t S ta n d a r d M o d e l ............................................................................................................................C - 3 2 C .2 .3 .1 . M o d e l C o d e ..................................................................................................................... C - 3 2 C .2 .3 .2 . I n p u t F i l e s ........................................................................................................................ C - 4 0 C .2 .4 . R a t G e s t a t i o n a l M o d e l ......................................................................................................................C - 5 5 C .2 .4 .1 . M o d e l C o d e ..................................................................................................................... C - 5 5 C .2 .4 .2 . I n p u t F i l e s ........................................................................................................................ C - 6 5 C .2 .5 . M o u s e S ta n d a r d M o d e l ................................................................................................................... C - 7 3 C .2 .5 .1 . M o d e l C o d e ..................................................................................................................... C - 7 3 C .2 . 5 . 2 . I n p u t F i l e s ........................................................................................................................ C -8 1 C .2 .6 . M o u s e G e s t a t i o n a l M o d e l ............................................................................................................. C - 8 5 C .2 .6 .1 . M o d e l C o d e ..................................................................................................................... C - 8 5 C .2 .6 .2 . I n p u t F i l e s ........................................................................................................................ C - 9 6 C .3. T O X IC O K IN E T IC M O D E L IN G R E S U L T S F O R K E Y A N IM A L B IO A S S A Y S T U D I E S ....................................................................................................................................................................C - 9 8 C .3 .1 . N o n g e s t a t i o n a l S t u d i e s .................................................................................................................... C - 9 8 C .3 .1 .1 . C a n t o n i e t a l. ( 1 9 8 1 ) ................................................................................................. C - 9 8 C .3 .1 .2 . C h u e t a l. ( 2 0 0 7 ) ....................................................................................................... C - 1 0 0 C .3 .1 .3 . C r o f t o n e t a l. ( 2 0 0 5 ) ................................................................................................C - 1 0 2 C .3 .1 .4 . D e l l a P o r t a e t a l. ( 2 0 0 1 ) ( f e m a l e ) ..................................................................C - 1 0 5 C .3 .1 .5 . D e l l a P o r t a e t a l. (2 0 0 1 ) ( m a l e ) ....................................................................... C - 1 0 7 C .3 .1 .6 . F a t t o r e e t a l. ( 2 0 0 0 ) ................................................................................................. C - 1 0 8 C .3 .1 .7 . F r a n c e t a l. ( 2 0 0 1 ) S p r a g u e D a w l e y R a t s .................................................C - 1 1 0 C .3 .1 .8 . F r a n c e t a l. ( 2 0 0 1 ) L o n g - E v a n s R a t s .......................................................... C - 1 1 1 C .3 .1 .9 . F r a n c e t a l. (2 0 0 1 ) H a n s W i s t a r R a t s .......................................................... C - 1 1 3 This document is a draftfor review purposes only and does not constitute Agency policy. C-ii DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) C .4. C .5. C .3 .1 .1 0 . H a s s o u n e t a l. ( 2 0 0 0 ) .............................................................................................C - 1 1 5 C .3 .1 .1 1 . H u t t e t a l. ( 2 0 0 8 ) .......................................................................................................C - 1 1 7 C .3 .1 .1 2 . K i t c h i n a n d W o o d s ( 1 9 7 9 ) ................................................................................. C - 1 1 8 C .3 .1 .1 3 . K o c i b a e t a l. ( 1 9 7 6 ) ..................................................................................................C - 1 2 1 C .3 .1 .1 4 . K o c i b a e t a l. ( 1 9 7 8 ) F e m a l e ............................................................................... C - 1 2 3 C .3 .1 .1 5 . K o c i b a e t a l. ( 1 9 7 8 ) M a l e ................................................................................... C - 1 2 5 C .3 .1 .1 6 . L a t c h o u m y c a n d a n e a n d M a t h u r ( 2 0 0 2 ) .....................................................C - 1 2 7 C .3 .1 .1 7 . L i e t a l. ( 1 9 9 7 ) .............................................................................................................C - 1 2 8 C .3 .1 .1 8 . M u r r a y e t a l. ( 1 9 7 9 ) A d u l t P o r t i o n ...............................................................C - 1 3 2 C .3 .1 .1 9 . N T P ( 1 9 8 2 ) -- F e m a l e R a t s , C h r o n i c ............................................................ C - 1 3 3 C .3 .1 .2 0 . N T P ( 1 9 8 2 ) -- M a l e R a t s , C h r o n i c ................................................................ C - 1 3 5 C .3 .1 .2 1 . N T P ( 1 9 8 2 ) -- F e m a l e M i c e , C h r o n i c ........................................................ C - 1 3 7 C .3 .1 .2 2 . N T P ( 1 9 8 2 ) -- M a l e M i c e , C h r o n i c .............................................................. C - 1 3 8 C .3 .1 .2 3 . N T P ( 2 0 0 6 ) 1 4 W e e k s ..........................................................................................C - 1 4 0 C .3 .1 .2 4 . N T P ( 2 0 0 6 ) 3 1 W e e k s ...........................................................................................C - 1 4 2 C .3 .1 .2 5 . N T P ( 2 0 0 6 ) 5 3 W e e k s ...........................................................................................C - 1 4 4 C .3 .1 .2 6 . N T P ( 2 0 0 6 ) 2 Y e a r s ................................................................................................ C - 1 4 6 C .3 .1 .2 7 . S e w a ll e t a l. ( 1 9 9 5 ) ..................................................................................................C - 1 4 8 C .3 .1 .2 8 . S h i e t a l. ( 2 0 0 7 ) A d u l t P o r t i o n .........................................................................C - 1 5 0 C .3 .1 .2 9 . S m i a lo w ic z e t a l. ( 2 0 0 8 ) ......................................................................................C - 1 5 1 C .3 .1 .3 0 . T o t h e t a l., 1 Y e a r ( 1 9 7 9 ) ................................................................................... C - 1 5 3 C .3 .1 .3 1 . V a n B i r g e l e n e t a l. ( 1 9 9 5 ) ..................................................................................C - 1 5 5 C .3 .1 .3 2 . V a n d e n H e u v e l e t a l. ( 1 9 9 4 ) ...........................................................................C - 1 5 7 C .3 .1 .3 3 . W h i t e e t a l. ( 1 9 8 6 ) ................................................................................................. C - 1 6 0 C .3 .2 . G e s t a t i o n a l S t u d i e s .......................................................................................................................... C - 1 6 3 C .3 .2 .1 . B e l l e t a l. ( 2 0 0 7 ) .......................................................................................................C - 1 6 3 C .3 .2 .2 . H a a v i s t o e t a l. ( 2 0 0 6 ) .............................................................................................C - 1 6 4 C .3 .2 .3 . H o j o e t a l. ( 2 0 0 2 ) .................................................................................................... C - 1 6 6 C .3 .2 .4 . I k e d a e t a l. ( 2 0 0 5 ) ...................................................................................................C - 1 6 7 C .3 .2 .5 . K a t t a i n e n e t a l. ( 2 0 0 1 ) ............................................................................................ C - 1 6 8 C .3 .2 .6 . K e l l e r e t a l. ( 2 0 0 7 ) ................................................................................................... C - 1 7 0 C .3 .2 .7 . L i e t a l. ( 2 0 0 6 ) 3 - D a y .............................................................................................C - 1 7 1 C .3 .2 .8 . M a r k o w s k i e t a l. ( 2 0 0 1 ) ........................................................................................C - 1 7 3 C .3 .2 .9 . M i e t i n n e n e t a l. ( 2 0 0 6 ) ...........................................................................................C - 1 7 4 C .3 .2 .1 0 . N o h a r a e t a l. ( 2 0 0 0 ) ................................................................................................ C - 1 7 6 C .3 .2 .1 1 . O h s a k o e t a l. ( 2 0 0 1 ) ................................................................................................ C - 1 7 7 C .3 .2 .1 2 . S c h a n t z e t a l. ( 1 9 9 6 ) a n d A m i n e t a l. ( 2 0 0 0 ) .......................................... C - 1 7 9 C .3 .2 .1 3 . S e o e t a l. ( 1 9 9 5 ) ......................................................................................................... C - 1 8 0 R E S P O N S E S U R F A C E T A B L E S .........................................................................................................C - 1 8 3 C .4 .1 . N o n g e s t a t i o n a l L i f e t i m e ...............................................................................................................C - 1 8 4 C .4 .2 . N o n g e s t a t i o n a l 5 - Y e a r A v e r a g e ..............................................................................................C - 1 9 2 C .4 .3 . G e s t a t i o n a l .............................................................................................................................................C - 1 9 8 R E F E R E N C E S ....................................................................................................................................................C - 2 0 4 This document is a draftfor review purposes only and does not constitute Agency policy. C-iii DRAFT--DO NOT CITE OR QUOTE 1 APPENDIX C. KINETIC M ODELING 2 3 4 C.1. LITERA TU RE SEARCH STRATEG Y AND RESULTS-- IDEN TIFY IN G RECENT 5 PUBLICATIONS FOR UPDATING TCDD TOXICOKINETIC M ODEL INPUT 6 PARAMETERS 7 T he purpose o f th is literature search w as to identify recen t publications th at address the 8 in p u t p aram eters fo r th e p hysiologically b ased pharm aco k in etic (P B P K ) m odels A ylw ard and 9 colleagues (described in articles p u blished in 2005 and 2009) and E m o n d and colleagues 10 (d escrib ed in articles p u b lish ed in 2 004, 2 005, an d 2006). T h is literatu re search w as p art o f th e 11 U .S . E n v iro n m e n ta l P r o te c tio n A g e n c y ( E P A ) 's p re p a ra tio n o f a re s p o n s e to th e N a tio n a l 12 A c ad em y o f S c ie n c e s' re v ie w (H ealth Risks from D ioxin and Related Compounds: Evaluation o f 13 the EPA R eassessm ent, N A S , 2 0 0 6 ]) o f E P A Exposure a n d H um an H ealth R eassessm ent o f 14 2.3.7.8- Tetrachlorodibenzo-p-Dioxin (TCDD) and R elated Compounds (U .S . E P A , 2 0 0 3 ), h e re in 15 c a lle d th e " 2 0 0 3 R e a sse ssm e n t." E n g lish -o n ly re fe re n c e s fro m 2 0 0 3 to M a y 2 0 0 9 w e re se a rc h e d 16 u sin g b ib lio g rap h ic d ata b ases relev an t to h ealth effects and to x ico lo g y o f 17 2 .3 .7 .8 - tetrach lo ro d ib en zo -p -d io x in (T C D D ). T he search fo cu sed on to x ico k in etic d ata th at 18 co u ld b e u se d to u p d a te th e d y n am ic d isp o sitio n o f 2 ,3 ,7 ,8 -T C D D in m ice, rats, g u in e a pigs, 19 m o n k ey s, and hum ans. 20 In th e prim ary search, E P A iden tified 775 distinct citations b ased on th e literature search 21 criteria d escribed below . E P A also p erform ed an in d ep en d en t supplem ental search to avoid 22 m issin g key studies. E P A id en tified 28 p ap ers fo r fu rth er analysis th at ap p eared on first rev iew 23 to rep o rt data to up d ate th e in p u t p aram eters o f th e A ylw ard and E m o n d P B P K m odels; 24 co n sid eratio n s fo r selec tio n are d esc rib ed in S ectio n C .1.3. 25 26 C.1.1. Data Bases Searched 27 E P A u sed th e follo w in g D IA L O G b ib lio g rap h ic data bases in th e prim ary search. B rief 28 d esc rip tio n s o f th e D IA L O G d ata b a se s search ed are p ro v id ed in S ectio n C .1.5. 29 3 0 1. F ile 6: N T I S 31 2. F ile 41: P o llu tio n A b stracts 32 3. F ile 55: B iosis 33 4. F ile 153: IP A T oxicology This document is a draftfor review purposes only and does not constitute Agency policy. C -1 D R A F T -- D O N O T C IT E O R Q U O T E 1 5. F ile 155: M ed L in e 2 6. F ile 156: T o x F ile 3 7. F ile 157: B iosis T oxicology 4 8. F ile 159: C a n c e r L it 5 9. F ile 336: R T E C S 6 7 T he P U B M E D data b ase w as used fo r the supplem ental search. 8 9 C.1.2. Literature Search Strategy and Approach 10 T h e p rim ary search u sed a tiered k ey -w o rd ap p ro ach , as d o cu m en ted b elo w . T he 11 p rin c ip a l s e a rc h te rm w a s th e C h e m ic a l A b s tra c t S e rv ic e R e g is try N u m b e r (C A S R N ) o r s p e c ific 12 ch em ical nam e, 2 ,3 ,7 ,8 -tetrach lo ro d ib en zo -p -d io x in o r 2 ,3 ,7 ,8 -T C D D . T he n ex t tie r o f search 13 te rm s w a s sp e c ie s, a n d fin a lly to x ic o k in e tic k e y w o rd s , as lis te d b e lo w . T h e p e rio d o f th e se a rc h 14 w as 2003 th ro u g h M ay 2009, and articles w ere lim ited to E n g lish language. 15 T h e su p p le m e n ta l P U B M E D se a rc h w a s lim ite d to th e m o st re c e n t fiv e y e a rs (2 0 0 4 to 16 p resen t) and u sed fo u r co m b in atio n s o f k ey w ords: 17 18 T C D D + p h arm a c o k in e tic + h u m an s, 19 T C D D + to x ico k in etic + h u m an s, 20 T C D D + pharm aco k in etic + anim als, and 21 T C D D + to x ico k in etic + anim als. 22 23 C .1.2.1. Chemical Search Terms--DIALOG Search 24 C A SR N : 1746-01-6 25 2 ,3 ,7 ,8 -tetrach lo ro d ib en zo -p -dioxin 26 2,3,7 ,8 -T C D D 27 2 8 C .1.2.2. Primary Search Terms (Species)--DIALOG Search 29 G uinea pig(s) 30 H um an(s) 31 M o n k ey (s) 32 M ouse This document is a draftfor review purposes only and does not constitute Agency policy. C-2 DRAFT--DO NOT CITE OR QUOTE 1 M ice 2 R odent(s) 3 R at(s) 4 5 C .1.2.3. Secondary Search Terms (Toxicology)--DIALOG Search 6 * = truncated 7 1w = term s are w ith in 1 w o rd o f each o th er and in th e o rd er specified (see search term 32) 8 1. A b s o r* 16. E lim in* 32. M echanism (1w ) 2. A D M E 17. E xcret* a c tio n 3. A ryl h y drocarbon 18. E p id em io lo g * 33. M etabo* re c e p to r 19. F eces 34. O ral* 4. A hR 20. Feed* 35. P450 5. B io av ail* 6. B ilia r* 21. F irst order kinetics 22. Food* 36. P artitio n coefficient 37. PB PK 7. B iotransform * 8. C y to c h ro m e 23. G astro* 24. G avage* 38. P harm acodynam ic* 39. P harm acokinetic* 9. C Y P* 10. C Y P 1A 1 11. C Y P 1A 2 12. D iet, dietary, diets 13. D isp o sit* 14. D istrib* 25. H alf-life 26. Induct* 27. Ingest* 28. In silico 29. K inetic* 30. L iver 40. P h y siologically based 41. pharm acokinetic 42. P ro tein bind* 43. T oxicokinetic* 44. U rin* 15. D rink* 31. Lym ph* 1 2 ADM E = absorption, distribution, metabolism, elimination; AhR = aryl hydrocarbon receptor; CYP = cytochrome 3 P450. 4 5 6 C.1.3. Citation Screening Procedures and Results 7 Initial D IA L O G searches resu lted in a very larg e n u m b er o f citation hits. T herefore, 8 som e title and key w ord restrictions w ere applied iteratively to screen out less relev an t citations 9 (e.g., re q u irin g som e search term s in title, req u irin g 2 ,3 ,7 ,8 -T C D D ra th e r th an ju s t T C D D ). 10 T hen, u sin g referen ce m an ag em en t softw are, p o o led in fo rm atio n o b tain ed fro m th e v ario u s 11 D I A L O G d a ta b a s e s w a s s c re e n e d to re m o v e d u p lic a te s . C ita tio n s th e n w e r e n u m b e re d This document is a draftfor review purposes only and does not constitute Agency policy. C-3 DRAFT--DO NOT CITE OR QUOTE 1 sequentially (as a u n iq u e identifier). In form ation retrieved included th e follo w in g (w hen 2 available): author(s), p u b licatio n year, title, source d o cu m en t nam e, volum e, and p ag e num bers. 3 T he D IA L O G search and d uplicate rem oval p ro ced u re p ro d u ced 775 u n iq u e citations. In 4 th e n ex t step, all 775 citatio n s w ere screen ed fo r p o ten tial ap p licab ility to u p d atin g p aram eters in 5 th e A ylw ard and E m o n d P B P K m odels. O f these 775 citations, 26 w ere selected fo r m ore 6 d etailed rev iew to determ ine th eir p o ten tial applicability, and full p u b licatio n s w ere retrieved. 7 T w o citations w ere added from th e supplem ental search, giving a total o f 28 articles id entified 8 fo r fu rth er review . 9 B ib lio g rap h ic in fo rm atio n fo r th e 28 articles selected fo r full rev iew is p ro v id ed in th e 10 referen ce list at th e en d o f th is section. T ab le C-1 su m m arizes th e m o d el in p u t p aram eters 11 p o te n tia lly a d d r e s s e d b y th e s e le c te d a rtic le s. 12 D u rin g 2003 to M ay 2 009, th e au th o rs o f th e tw o k in etic m o d els u n d er co n sid eratio n 13 p u b lis h e d se v e ra l a rtic le s. F o r th e E m o n d m o d e l, w h ic h w a s firs t p u b lis h e d in 2 0 0 4 (E m o n d 14 et al., 2 0 0 4 ), tw o su b se q u e n t p a p e rs h av e b e e n p u b lish ed (E m o n d et al., 2 0 0 5 , 2 0 0 6 ). T h e 15 A y lw a rd m o d el, w h ic h o rig in a te d fro m th e 1995 p a p e rs b y C a rrie r et al. (1 9 9 5 a , b ), w a s la te r 16 u p d a te d b y th e sam e g ro u p (A y lw a rd et al., 2 0 0 5 a, b). T h e m a jo r ch a n g e im p le m e n ted in th e last 17 tw o p ap ers w as th e d escrip tio n o f a d eso rp tio n p ro cess in th e d ig estiv e tract. T h e tran sfer rate 18 d esc rib ed is slow , b u t fo r a lo w b o d y b u rd e n o f T C D D , th is p ro c ess re m a in s sig n ifican t. T his 19 co n cep t w as rep o rted in 2 0 0 2 b y M o ser and M cL a ch lan (2002). T h e m ajo r m o d ificatio n s 20 expected to up d ate th e E m o n d m odel are (1) consideration o f the desorption process in the 21 g astrointestinal tract and (2) rearran g em en t o f th e elim ination constant, w h ich w ill have a 22 n eg lig ib le im p act on th e sim ulation. T hese changes are m o tiv ated by p lau sib le observations 23 rep o rted in th e literature. 24 B ecau se o f th e body b u rd en found in hum ans and th e im p o rtan ce o f selecting an 25 ap p ro p riate dose m etric in h u m an risk assessm ent, th e p h y sio lo g ical m odel is an im p o rtan t tool 26 fo r asse ssin g th e k in e tic s fo llo w in g ex p o su re to T C D D (K im et al., 2 0 03). B a se d on th e 27 literature iden tified in th is search, th e m ajor co n tributions th at should b e review ed w ith resp ect to 28 th e A ylw ard and E m o n d k inetic m odels are n o t m odes o f action or pharm aco k in etic m echanism s, 29 b u t rath er inform ation fo r v erify in g or im proving th e accuracy o f som e m odel param eters. 30 P h arm aco k in etics ty p ically refers to fo u r d istin ct steps in clu d in g absorption, distribution, 31 m etab o lism , and ex cretio n . P h y sio lo g ically -b ased m o d els co n sid er each step. In th e m o d el each This document is a draftfor review purposes only and does not constitute Agency policy. C -4 D R A F T -- D O N O T C IT E O R Q U O T E 1 step is p aram eterized to reflect b etter p red ictio n s o f th e real observations. O ccasionally, 2 rev iew in g th ese m o d els is essential to determ ine if any k ey p ro cesses or p aram eters m ig h t be 3 described w ith b etter accuracy. T his perspective underlies the review o f the literature described 4 here. T he rev iew ind icates T C D D disposition has b eco m e reco g n ized as relatively significant 5 since th e publication o f th e E m o n d and A ylw ard m odels. T he literatu re th at provides 6 in fo rm atio n related to im p ro v in g th ese m odels, how ever, is lim ited. F o r th e b en efit o f this 7 exercise, E P A selected th e literature th at w o u ld likely co ntribute significantly to m odel response, 8 o r to clarify o r confirm d ifferen t k ey issu es d riv in g th e m odel results. R eg ard in g th e tw o T C D D 9 m odels, th e tw o m ajo r issues th at should b e evaluated w ith resp ect to th e recen t literature 10 id en tified are th e elim in atio n p ro file and th e in d u ctio n o f C Y P 1A 2. 11 R e v ie w in g th e e lim in a tio n v a r ia tio n in d iffe re n t s p e c ie s a n d te s tin g v a r ia b le e lim in a tio n 12 w ith a d ata set ap p ears to b e ap p ro p riate. T he literatu re rep o rts th a t v ario u s facto rs m ig h t 13 in flu e n c e e lim in a tio n ra te . R e c e n t p u b lic a tio n s re p o rt th e in flu e n c e o f d iv e rs e p re d ic to rs su c h 14 age, b o d y fat, o r sm o k in g h a b it o n th e e lim in a tio n h alf-life (M ilb rath et al., 2 0 0 9 ; K e rg e r e t al., 15 2 0 0 6 , 2 0 0 7 ). D e te rm in in g w h e th e r u sin g th e M ilb ra th et al. in fo rm a tio n w o u ld h e lp a c c o u n t fo r 16 in trasp ecies v ariab ility in elim in atio n rate in th e E m o n d an d A y lw ard k in etic m o d els w o u ld be 17 u sefu l. In 2 0 0 6 , E m o n d et al. re v ie w e d th e in flu e n c e o f b o d y fa t m a ss an d C Y P 1 A 2 in d u c tio n on 18 th e p h a rm a c o k in e tic s o f T C D D . T h ese tw o fa cto rs a p p e a r to c o n trib u te sig n ifican tly to 19 elim in atio n and th e ir in flu en ces seem to b e d riv en b y T C D D b o d y b u rd en . M u llero v a and 20 K op eck y (2007) discussed th e influence o f adipose tissue and th e "y o -y o " effects on v arious 21 d iseases th at m ig h t b e in flu en ced by p ersisten t org an ic p o llu tan t distribution. O ne group 22 explored the im p o rtan ce o f variab le elim ination and com pared these p redictions to first-o rd er 23 elim ination u sin g th e A ylw ard and E m o n d m odels and supported th ese approaches fo r risk 24 asse ssm e n t (H e in zl et al., 2 0 0 7 ). T w o g ro u p s o f a u th o rs c o n sid ered a o n e-c o m p a rtm e n t m o d el to 25 d eriv e th e e lim in a tio n h alf-life (A y lw a rd e t al., 2 0 0 9 ; N a d a l et al., 2 0 0 8 ). C o m p arin g th e 26 h alf-life they obtained u sing th is approach fo r a range o f body b u rd en to th e v ariab le elim ination 27 h alf-life w o u ld be interesting. 28 T he second im p o rtan t m echanism driv in g th e d istrib u tio n and elim in atio n o f T C D D is the 29 in d u c tio n o f C Y P 1 A 2 , id e n tifie d as th e m a jo r lig an d p ro te in in liv e r (D ilib erto et al., 1997). F o r 30 th at process, authors suggested different aspects th at should be investigated, including the 31 im p o rtan ce o f th e d o se m etrics in th e targ e t tissu e and th e in d u cib le level o f C Y P 1 A 2 (W ilk es This document is a draftfor review purposes only and does not constitute Agency policy. C -5 D R A F T -- D O N O T C IT E O R Q U O T E 1 et al., 2 0 0 8 ; S task al et al., 2 0 0 5 ). O th e r p ap e rs ad d ress th e in tra sp e c ie s v a ria b ility o f leth al 2 p o ten cy in m atu re sp ecies v ersu s th e d ev e lo p in g fe tu s (K ra n sle r et al., 2 0 0 7 ; K o rk a la in e n e t al., 3 2004). Still o th ers p o in t o u t p ro n o u n ced d ifferen ces am o n g sp ecies (n am ely , g u in ea pigs, 4 ham sters, m ice, and rats) (B ohonow ych and D enison, 2007), as observed in studies o f long-term 5 effects o f lo w T C D D dose in liv er and in studies com paring hep atic accum ulation and clearance 6 o f T C D D (K o re n a g a e t al., 20 0 7 ; B o v e rh o f et al., 2 005). T h e in te rsp e c ie s v a ria tio n o f th e 7 b in d in g affinity co n stan t o f A h R also has b een rep o rted (C o n n o r and A ylw ard, 2006; N o h ara 8 et al., 2 0 0 6 ). 9 T he articles iden tified in this literatu re rev iew should b e adequate to up d ate th e A ylw ard 10 and E m o n d m o d els, w h ich n eed to b e ev alu ated acco rd in g to th e sam e stru ctu re o f co m p artm en ts 11 d e s c rib e d in th e lite ra tu re b y th e tw o m o d e l a u th o rs. 12 13 C.1.4. References Selected for M ore Detailed Review for Updating the PB PK M odels A y lw a rd , L L ; B ru n et, R C ; C arrier, G ; e t al. (2 0 0 4 ). C o n c e n tra tio n -d e p e n d e n t T C D D elim in a tio n kinetics in hum ans: to x icokinetic m odeling for m oderately to highly exposed adults from Seveso, Italy , and V ien n a, A ustria, an d im p act on d o se estim ates fo r th e N IO S H cohort. J E x p o A nal E n v i r o n E p i d e m i o l 1 5 ( 1 ) :5 1 --6 5 . A y lw a rd , L L ; B ru n et, R C ; S tarr, T B ; et al. (2 0 0 5 ). E x p o su re re c o n stru c tio n fo r th e T C D D ex posed N IO S H co h o rt u sin g a co n cen tratio n - and ag e-d ep en d en t m odel o f elim ination. R isk A n a l 2 5 ( 4 ) : 9 4 5 --9 5 6 . A y lw a rd , L L ; B o d n er, K M ; C o llin s, JJ; et al. (2 0 0 9 ). T C D D e x p o su re e stim a tio n fo r w o rk e rs at a N ew Z ealan d 2 ,4 ,5 -T m an u factu rin g facility b ased on serum sam pling data. J E x p o Sci E n v iro n E p id em io l. doi: 1 0 .1 0 3 8 /jes.2 0 0 9 .3 1 . B ohonow ych, JE; D enison, M S. (2007). P ersisten t b in d in g o f lig an d s to th e aryl h ydrocarbon receptor. T oxicol Sci 98(1):99-109. B o v e rh o f, D R ; B u rg o o n , L D ; T ash iro , C ; e t al. (2 0 0 5 ). T em p o ra l an d d o se -d e p e n d e n t h e p a tic gene ex p ressio n p attern s in m ice p ro v id e n ew in sig h ts in to T C D D -m ed iated hepatotoxicity. T o x i c o l S c i 8 5 ( 2 ) : 1 0 4 8 --1 0 6 3 . C onnor, K T ; A ylw ard, L L . (2006). H u m an resp o n se to dioxin: aryl h y d ro carb o n recep to r (A hR ) m o lecu lar structure, function, and d ose-response data fo r enzym e induction in dicate an im paired h u m a n A h R . J T o x i c o l E n v i r o n H e a l t h B 9 ( 2 ) : 1 4 7 --1 7 1 . H einzl, H ; M ittlback, M ; E dler, L. (2007). O n th e tran slatio n o f u n certain ty from to x ico k in etic t o t o x i c o d y n a m i c m o d e l s - t h e T C D D e x a m p l e . C h e m o s p h e r e 6 7 ( 9 ) : S 3 6 5 --S 3 7 4 . This document is a draftfor review purposes only and does not constitute Agency policy. C-6 DRAFT--DO NOT CITE OR QUOTE Irigaray, P; M ejean, L ; L aurent, F. (2005). B eh av io u r o f dioxin in pig adipocytes. F o o d C hem T oxicol 4 3 (3 ):4 5 7 -4 6 0 . K e rg er, B D ; L eu n g , H W ; S cott, P ; et al. (2 0 0 6 ). A g e - an d c o n c e n tra tio n -d e p e n d e n t elim in a tio n h alf-life o f 2 ,3 ,7 ,8 -tetrach lo ro d ib en zo -p -d io x in in S eveso children. E n v iro n H ealth P ersp ect 1 1 4 (1 0 ):1 5 9 6 -1602. K e rg er, B D ; L eu n g , H W ; S cott, P K ; et al. (2 0 0 7 ). R e fin e m e n ts o n th e a g e -d e p e n d e n t h a lf-life m odel fo r estim ating child body b u rd en s o f polych lo ro d ib en zo d io x in s and dibenzofurans. C hem osphere 6 7 (9 ):S 2 7 2 -S 2 7 8 . K im , A H ; K o h n , M C ; N y sk a , A ; et al. (2 0 0 3 ). A re a u n d e r th e cu rv e as a d o se m e tric fo r p ro m o tio n al resp o n ses fo llo w in g 2 ,3 ,7 ,8 -tetrach lo ro d ib en zo -p -d io x in exposure. T o xicol A ppl P harm acol 1 9 1 (1 ):1 2 -21. K o re n a g a , T; F u k u sa to , T; O h ta, M ; et al. (2 0 0 7 ). L o n g -te rm effec ts o f su b c u ta n e o u sly in je c te d 2 .3 .7 .8 - tetrach lo ro d ib en zo -p -d io x in on th e liv er o f rhesus m onkeys. C hem o sp h ere 67(9):S 3 9 9 -S 4 0 4 . K o rk alain en , M ; T uom isto, J; P o h jan v irta, R. (2004). P rim ary structure and in d u cib ility by 2 .3 .7 .8 - tetrach lo ro d ib en zo -p -d io x in (T C D D ) o f aryl h y d ro carb o n recep to r rep resso r in a T C D D sen sitiv e and a T C D D -resistan t rat strain. B io ch em B io p h y s R es C o m m u n icatio n s 3 1 5 (1 ):1 2 3 -131. K ransler, K M ; M cG arrig le, B P; O lson, JR . (2007). C o m p arativ e developm ental to x icity o f 2 .3 .7 .8 - tetrach lo ro d ib en zo -p -d io x in in th e h am ster, rat and g u in ea pig. T o x ico lo g y 2 2 9 (3 ):2 1 4 -2 2 5 . M a ru y a m a , W ; Y o sh id a, K ; T an ak a, T; e t al. (2 0 0 2 ). D e te rm in a tio n o f tissu e -b lo o d p a rtitio n co efficien ts fo r a p h y sio lo g ical m odel fo r hum ans, and estim atio n o f dioxin co n cen tratio n in tissues. C h em osphere 4 6 (7 ):9 7 5 -9 8 5 . M a ru y a m a , W ; Y o sh id a, K ; T an ak a, T; et al. (2 0 0 3 ). S im u latio n o f d io x in a c c u m u la tio n in h u m an tissu es and analysis o f rep ro d u ctiv e risk. C h em o sp h ere 53(4):301-313. M aruyam a, W ; A oki, Y. (2006). E stim ated can cer risk o f dioxins to hum ans usin g a b ioassay and p h ysiologically b ased pharm aco k in etic m odel. T oxicol A ppl P harm acol 2 1 4 (2 ):1 8 8 -1 9 8 . M ilb ra th , M O ; W e n g er, Y ; C h an g , C -W ; et al. (2 0 0 9 ). A p p a re n t H a lf-L iv e s o f D io x in s, F u ran s, and P o ly c h lo rin ated B ip h en y ls as a F u n ctio n o f A ge, B o d y F at, S m o k in g Status, and B re a st F eeding. E nviron H ealth P ersp ect 117(3 ):4 1 7 -4 2 5 . M oser, G A ; M cL achlan, M S. (2002). M o d elin g d igestive tract ab sorption and desorption o f lip o p h ilic o rg an ic co n tam in an ts in hum ans. E n v iro n Sci T echnol 3 6 (1 5 ):3 3 1 8 -25. M u llero v a, D ; K o p eck y , J. (2007). W h ite ad ip o se tissue: sto rag e and effecto r site fo r en vironm ental pollutants. P hysiol R es 5 6 (4 ):3 7 5 -3 8 1 . This document is a draftfor review purposes only and does not constitute Agency policy. C-7 DRAFT--DO NOT CITE OR QUOTE N a d a l, M ; P e re llo , G ; S ch u h m ach e r, M ; et al. (2 0 0 8 ). C o n c e n tra tio n s o f P C D D /P C D F s in plasm a o f subjects living in th e vicinity o f a hazardous w aste incinerator: F ollow -up and m odeling validation. C hem osphere 7 3 (6 ):9 0 1 -9 0 6 . N o h a ra , K ; A o , K ; M iy a m o to , Y ; et al. (2 0 0 6 ). C o m p a riso n o f th e 2 ,3 ,7 ,8 -te tra c h lo ro d ib e n z o p -d io x in (T C D D )-in d u ced C Y P 1A 1 g en e ex p ressio n p ro file in ly m p h o cy tes from m ice, rats, and hum ans: M o st p o ten t in d u ctio n in hum ans. T ox ico lo g y 2 2 5 (2 -3 ):2 0 4 -2 1 3 . O lsm an , H ; E n g w a ll, M ; K a m m a n n , U ; et al. (2 0 0 7 ). R e la tiv e d iffe re n c e s in ary l h y d ro c a rb o n recep to r-m ed iated response fo r 18 poly b ro m in ated and m ixed halo g en ated dibenzo-p-dioxins and -fu ran s in cell lin es fro m fo u r d ifferen t species. E n v iro n T oxicol C hem 2 6 (1 1 ):2 4 4 8 -2 4 5 4 . S ag h ir, SA ; L e b o fsk y , M ; P in so n , D M ; e t al. (2 0 0 5 ). V a lid a tio n o f H a b e r's R u le (d o se X tim e= co n stan t) in rats and m ice fo r m o n o ch lo ro acetic acid and 2 ,3 ,7 ,8 -tetrach lo ro d ib en zo p -d io x in u n d er co n d itio n s o f k in etic steady state. T o x ico lo g y 2 1 5 (1 -2 ):4 8 -5 6 . S ch ecter, A ; P av u k , M ; P o p k e, O ; et al. (2 0 0 3 ). D io x in , d ib en zo fu ran , an d c o p la n a r P C B L e v e ls in L ao tian b lo o d and m ilk from A g en t O ran g e-sp ray ed and n onsprayed areas, 2001. J T oxicol E nviron H ealth A 6 6 (2 1 ):2 0 6 7 -2 0 7 5 . S task al, D F ; D ilib e rto , JJ; D e v ito , M J; et al. (2 0 0 5 ). In h ib itio n o f h u m a n an d ra t C Y P 1 A 2 b y T C D D and d ioxin-like chem icals. T oxicol Sci 8 4 (2 ):2 2 5 -2 3 1 . T o y o sh ib a, H ; W a lk e r, N J; B a ile r, A J; e t al. (2 0 0 4 ). E v a lu a tio n o f to x ic e q u iv ale n cy fa c to rs fo r in d u ctio n o f cytochrom es P 4 5 0 C Y P 1A 1 and C Y P 1A 2 enzym e activity by d io x in -lik e com pounds. T oxicol A ppl P harm acol 1 9 4 (2 ):1 5 6 -1 6 8 . W ilk es, JG ; H a ss, B S ; B u z a tu , D A ; e t al. (2 0 0 8 ) . M o d e lin g an d a ssa y in g d io x in -lik e b io lo g ic al effects fo r both dioxin-like and certain non-d io x in -lik e com pounds. T oxicol Sci 1 0 2 (1 ):187-195. 1 C.1.5. B rief Descriptions of D IA LO G Bibliographic Data Bases Searched 2 T he N atio n al T echnical In form ation S ervice (N T IS ) database com prises sum m aries o f 3 U .S . g o v ern m en t-sp o n so red research , d evelopm ent, and en g in eerin g , plus analyses p repared b y 4 federal ag en cies, th eir co n tracto rs, or g ran tees. It is th e m ean s th ro u g h w h ich u n classified , 5 p u blicly available, u n lim ited d istrib u tio n rep o rts are m ade av ailab le fo r sale from 240 agencies. 6 A d ditionally, som e state and local g o v ern m en t ag en cies co n trib u te su m m aries o f th eir reports to 7 th e database. N T IS also p rovides access to the results o f gov ern m en t-sp o n so red research and 8 d ev e lo p m e n t fro m co u n tries o u tsid e th e U n ited States. O rg an izatio n s th a t cu rren tly co n trib u te to 9 th e N T IS d atabase in clude b u t are n o t lim ited to th e follow ing: th e Japan M inistry o f 10 In tern atio n al T rad e and In d u stry (M IT I); lab o rato ries ad m in istered b y th e U n ited K in g d o m This document is a draftfor review purposes only and does not constitute Agency policy. C-8 DRAFT--DO NOT CITE OR QUOTE 1 D ep artm en t o f Industry; th e G erm an F ederal M in istry o f R esearch and T echnology (B M F T ); and 2 the F ren ch N ational C en ter fo r S cien tific R esearch (C N R S ). 3 P o llu tio n A b stracts provides access to environm ental inform ation th at com bines 4 inform ation on scientific research and g o v ern m en t p olicies in a single reso u rce. T opics o f 5 gro w in g concern are extensively covered from th e standpoints o f atm osphere, em issions, 6 m athem atical m odels, effects on people and anim als, and environm ental action in response to 7 global p o llu tio n issu es. T his datab ase also contains m aterial from co n feren ce p ro ceed in g s and 8 hard-to-find sum m arized docum ents along w ith inform ation from prim ary jo u rn als in the field o f 9 pollution. 10 B IO S IS P rev iew s co n tain s citatio n s fro m B io lo g ical A b stra cts (B A ) and B io lo g ical 11 A b s tra c ts /R e p o r ts , R e v ie w s , a n d M e e tin g s (B A /R R M ) (fo rm e rly B io R e s e a rc h In d e x ), th e 12 m ajo r p u b licatio n s o f B IO S IS . T h ese p u b licatio n s co n stitu te th e m ajo r E n g lish -lan g u ag e 13 s e rv ic e p ro v id in g c o m p re h e n s iv e w o rld w id e c o v e ra g e o f re s e a rc h in th e b io lo g ic a l a n d 14 b io m ed ical sciences. B io lo g ical A b stracts in clu d es ap p ro x im ately 3 5 0 ,0 0 0 acco u n ts o f original 15 re se a rc h y e a rly fro m n e a rly 5 ,0 0 0 p rim a ry jo u rn a l a n d m o n o g ra p h title s. B A /R R M in c lu d e s an 16 ad d itio n al 2 0 0 ,0 0 0 + citatio n s a y ea r fro m m eetin g ab stracts, rev iew s, b o o k s, b o o k ch ap ters, 17 n o tes, letters, and selected reports. 18 IP A T o x ico lo g y p ro v id es fo c u sed to x ic o lo g y in fo rm a tio n on all p h ase s o f th e 19 d ev elo p m en t and u se o f d ru g s and on p ro fessio n al p h arm aceu tical p ractice. T h e scope o f th e 20 datab ase ran g es from th e clinical and practical to th e th eo retical aspects o f to x ico lo g y literature. 21 A u n iq u e featu re o f abstracts rep o rtin g clinical studies is th e in clu sio n o f th e study design, 22 n u m b er o f patien ts, dosage, d o sag e form s, and do sag e schedule. 23 M ed ical L iteratu re, A nalysis, and R etrieval S y stem O n lin e (M E D L IN E ), p ro d u ced by 24 th e U .S. N atio n al L ib rary o f M ed icin e (N L M ), is N L M 's p rem ier b ib lio g rap h ic database. It 25 co n tain s m o re th a n 15 m illio n re fe re n c e s to jo u rn a l artic le s in life scien ces w ith a c o n c en tratio n 26 on biom edicine. T he b ro ad coverage o f th e database includes b asic biom edical research and the 27 clinical sciences since 1950, including nursing, dentistry, veterin ary m edicine, pharm acy, allied 28 health, and pre-clin ical sciences. M E D L IN E also covers life sciences th at are vital to 29 b io m ed ical p ractitio n ers, researchers, and educators, in clu d in g som e aspects o f biology, 30 environm ental science, m arine biology, and p lan t and anim al science, as w ell as b io p h y sics and 31 ch em istry . M E D L IN E is in d e x e d u sin g N L M 's c o n tro lle d v o c a b u la ry , M e d ic a l S u b je ct This document is a draftfor review purposes only and does not constitute Agency policy. C -9 D R A F T -- D O N O T C IT E O R Q U O T E 1 H ead in g s (M eS H ). A pproxim ately 400,000 records are added p er year, o f w h ich m ore than 2 76 p ercen t are in E nglish. M E D L IN E contains A ID S L IN E , H ealthS T A R , T oxline, In P rocess 3 (form erly k n o w n as P re-M E D L IN E ), In D ata R eview , and P O P L IN E . 4 T oxF ile covers th e toxicological, pharm acological, b iochem ical, and physiological 5 effects o f drugs and o th er chem icals. A d v erse drug reactions, chem ically in d u ced diseases, 6 carcinogenesis, m utagenesis, terato g en esis, en vironm ental pollution, w aste disposal, radiation, 7 and food contam ination are typical areas o f coverage. T he databases E nvironm ental M utagen 8 In form ation C en ter (E M IC ), D evelopm ental and R ep ro d u ctiv e T oxicology (D A R T ), and T oxic 9 S ubstances C ontrol A ct T est S ubm issions (T S C A T S ) are in clu d ed in T oxF ile. It is n o t clearly 10 stated w h e th e r th e C h em ical C arcin o g en esis R esearch In fo rm atio n S y stem (C C R IS ), H azard o u s 11 S u b s ta n c e s D a ta B a n k (H S D B ), o r G e n e tic T o x ic o lo g y D a ta B a n k (G E N E -T O X ) a re in c lu d e d in 12 T oxF ile. C o n seq u en tly , a separate, o n -lin e search w a s co n d u cted to en su re th a t th ese d atab ases 13 w e re se a rc h e d . 14 B IO S IS T o x ico lo g y co n tain s citatio n s fro m B A and B A /R R M (fo rm erly B io R esearch 15 In d e x ), th e m a jo r p u b lic a tio n s o f B IO S IS , th a t fo c u s o n to x ic o lo g y a n d re la te d to p ics. 16 R eco rd s are d raw n fro m jo u rn a l articles, co n feren ce p apers, m o n o g rap h s and b o o k chapters, 17 n o tes, letters, and rep o rts, as w ell as o rig in al research . U .S . p aten t reco rd s are also included. 18 C A N C E R L IT is p ro d u c e d b y th e In tern atio n al C a n c e r R e se a rc h D a ta B a n k B ran c h 19 (IC R D B ) o f th e U .S . N atio n al C an cer In stitu te. T h e d atab ase co n sists o f b ib lio g rap h ic reco rd s 20 referencing cancer research publications dating from 1963 to 2002. M o st records contain 21 ab stracts, an d all re co rd s co n tain citatio n in fo rm atio n an d ad d itio n al d escrip tiv e field s su ch as 22 docum ent type and language. B eginning w ith the June 1983 C A N C E R L IT update, records from 23 th e M E D L IN E database dealing w ith can cer to p ics have b een added to C A N C E R L IT . 24 T he R eg istry o f T oxic E ffects o f C hem ical S ubstances (R T E C S ) is a co m p reh en siv e 25 d atab ase o f b asic to x icity in fo rm atio n fo r o v er 150,000 ch em ical su b stan ces in clu d in g 26 p rescrip tio n and n o n -p rescrip tio n drugs, fo o d additives, pesticides, fungicides, herbicides, 27 solvents, diluents, chem ical w astes, reactio n p ro d u cts o f chem ical w aste, and substances u sed in 28 b o th in d u strial and h o u seh o ld situations. R ep o rts o f th e to x ic effects o f each co m p o u n d are 29 cited. In ad d itio n to to x ic effects and g en eral to x ico lo g y review s, d ata on sk in an d /o r eye 30 irritation, m utation, rep ro d u ctiv e consequences and tum origenicity are provided. F ederal 31 stan d ard s and reg u latio n s, N atio n al In stitu te fo r O ccu p atio n al S afety an d H e alth (N IO S H ) This document is a draftfor review purposes only and does not constitute Agency policy. C -10 D R A FT -- D O N O T C IT E O R Q U O T E 1 recom m ended exposure lim its and inform ation on th e activities o f E P A , N IO S H , N ational 2 T o x ico lo g y P ro g ram (N T P ), an d O ccu p atio n al S afety an d H ealth A d m in istratio n (O S H A ) 3 reg ard in g th e substance are also included. T he to x ic effects are lin k ed to literatu re citations from 4 b o th pub lish ed and u n p u b lish ed governm ental reports, and pub lish ed articles from th e scientific 5 literature. T he datab ase co rresp o n d s to th e p rin t v ersio n o f the R T E C S , fo rm erly k n o w n as the 6 T oxic S ubstances L ist, w h ich w as started in 1971. O riginally p repared by th e N IO S H , the 7 R T E C S datab ase is n o w p ro d u ced and d istrib u ted by S ym yx T echnologies, Inc. 8 9 C.2. TO XICO K INETIC M O D ELIN G CODE (EM OND ET AL., 2005) 10 C.2.1. H um an Standard M odel 11 C.2.1.1. M o d e l C o d e 12 P R O G R A M : 'T h re e C o m p a rtm e n t P B P K M o d e l fo r T C D D in H u m an : S ta n d a rd M o d el 13 (N o n -G e sta tio n )' 14 15 !HUM_NON_GEST_ICF_F083109.csl 16 !***************************************************** 17 18 INITIAL INITIALIZATION OF PARAMETERS 19 2 0 !SIMULATION PARAMETERS ==== 21 CONSTANT EXP TIME ON = 0. ! TIME AT WHICH EXPOSURE BEGINS 2 2 (HOUR) 2 3 CONSTANT EXP TIME OFF = 6.132e5 ! TIME AT WHICH EXPOSURE ENDS 2 4 (HOUR) 2 5 CONSTANT DAY CYCLE = 24.0 ! NUMBER OF HOURS BETWEEN DOSES 2 6 (HOUR) 2 7 CONSTANT BCK TIME ON = 6.132e5 ! TIME AT WHICH BACKGROUND 2 8 EXPOSURE BEGINS (HOUR) 2 9 CONSTANT BCK TIME OFF = 6.132e5 ! TIME AT WHICH BACKGROUND 3 0 EXPOSURE ENDS (HOUR) 31 3 2 (EXPOSURE DOSES 33 CONSTANT MSTOTBCKGR = 0.0 ! ORAL BACKGROUND EXPOSURE DOSE 3 4 (NG/KG) 3 5 CONSTANT MSTOT = 1.0E-7 ! ORAL EXPOSURE DOSE (NG/KG) o o o o 3 6 CONSTANT DOSEIV = ! INJECTED DOSE (NG/KG) 3 7 CONSTANT MW = 322.0 ! MOLECULAR WEIGHT (G/MOL) 3 8 MSTOT NM = MSTOT/MW ! CONVERTS THE DOSE TO NMOL/KG 3 9 MSTOT NMBCKGR = MSTOTBCKGR/MW !CONVERTS THE BACKGROUND DOSE TO NMOL/KG 4 0 DOSEIV NM = DOSEIV/MW ! CONVERTS THE INJECTED DOSE TO 41 NMOL/KG 42 4 3 (INITIAL GUESS OF THE FREE CONCENTRATION IN THE LIGAND (COMPARTMENT 4 4 INDICATED> BELOW) ==== 4 5 CONSTANT CFLLI0 = ! LIVER (NMOL/L) This document is a draftfor review purposes only and does not constitute Agency policy. C-11 DRAFT--DO NOT CITE OR QUOTE 1 2 IBINDING CAPACITY (AhR) FOR NON LINEAR BINDING (COMPARTMENT INDICATED 3 BELOW) === 4 CONSTANT LIBMAX = 0.35 I LIVER (NMOL/L) 5 6 I PROTEIN AFFINITY CONSTANTS (1A2 OR AhR, COMPARTMENT INDICATED BELOW) 7 8 CONSTANT KDLI 0.1 ! LIVER (AhR) (NMOL/L) WANG 9 ET AL.. 1997 10 CONSTANT KDLI2 40.0 ! LIVER (1A2) (NMOL/L) EMOND ET 11 AL. 2004 12 13 !EXCRETION AND ABSORPTION CONSTANTS 14 CONSTANT KST = 0.01 ! GASTRIC RATE CONSTANT (HR- 15 1), EMOND ET AL., 2005 16 CONSTANT KABS = 0.06 ! INTESTINAL ABSORPTION CONSTANT 17 (HR-1), EMOND ET AL. 2005 18 19 ELIMINATION CONSTANTS 2 0 CONSTANT CLURI = 4.17D-8 I URINARY CLEARANCE (L/HR), EMOND 21 ET AL., 2005 2 2 CONSTANT KELV = 1.1e-3 I INTERSPECIES VARIABLE 2 3 ELIMINATION CONSTANT (1/HOUR) 24 2 5 !CONSTANT TO DIVIDE THE ABSORPTION INTO LYMPHATIC AND PORTAL FRACTIONS 2 6 CONSTANT A = 0.7 I LYMPHATIC FRACTION, 2 7 WANG ET AL. (1997) 28 2 9 !PARTITION COEFFICIENTS 3 0 CONSTANT PF = 1.0e2 I ADIPOSE TISSUE/BLOOD, 31 WANG ET AL. 1997 3 2 CONSTANT PRE = 1.5 I REST OF THE BODY/BLOOD, 33 WANG ET AL. 1997 3 4 CONSTANT PLI = 6.0 I LIVER/BLOOD, WANG ET 3 5 AL. 1997 36 3 7 (PARAMETERS FOR INDUCTION OF CYP1A2 3 8 CONSTANT PAS INDUC = 1.0 I INCLUDE INDUCTION? (1 = YES, 0 3 9 = NO) 4 0 CONSTANT CYP1A2 1OUTZ = 1.6e3 DEGRADATION CONCENTRATION CONSTANT 41 OF 1A2 (NMOL/L) 4 2 CONSTANT CYP1A2 1A1 = 1.6e3 I BASAL CONCENTRATION OF 1A1 4 3 (NMOL/L) 4 4 CONSTANT CYP1A2 1EC50 = 1.3e2 I DISSOCIATION CONSTANT TCDD-CYP1A2 4 5 (NMOL/L) 4 6 CONSTANT CYP1A2 1A2 = 1.6e3 I BASAL CONCENTRATION OF 1A2 4 7 (NMOL/L) 4 8 CONSTANT CYP1A2 1KOUT = 0.1 I FIRST ORDER RATE OF DEGRADATION 4 9 (H-1) 5 0 CONSTANT CYP1A2 1TAU = 0.25 I HOLDING TIME (H) 51 CONSTANT CYP1A2 1EMAX = 9.3e3 I MAXIMUM INDUCTION OVER BASAL EFFECT 5 2 (UNITLESS) 53 CONSTANT HILL = 0.6 IHILL CONSTANT; COOPERATIVELY LIGANDI 5 4 BINDING :EFFECT CONSTANT (UNITLESS) 55 ! DIFFUSIONAL PERMEABILITY FRACTION 5 6 CONSTANT PAFF = 0.12 I ADIPOSE (UNITLESS) This document is a draftfor review purposes only and does not constitute Agency policy. C-12 DRAFT--DO NOT CITE OR QUOTE 1 CONSTANT PAREF = 0.03 ! REST OF BODY (UNITLESS) 2 CONSTANT PALIF = 0.35 ! LIVER (UNITLESS) 3 4 (TISSUE BLOOD FLOW EXPRESSED AS A FRACTION OF CARDIAC OUTPUT ========= 5 CONSTANT QFF = 0.05 ! ADIPOSE TISSUE BLOOD FLOW FRACTION 6 (UNITLESS), KRISHNAN 2008 7 CONSTANT QLIF = 0.26 ! LIVER (UNITLESS), KRISHNAN 2008 8 9 (COMPARTMENT TISSUE BLOOD EXPRESSED AS A FRACTION OF THE TOTAL 10 COMPARTMENT VOLUME ========= 11 CONSTANT WFB0 = 0.050 ! ADIPOSE TISSUE, WANG ET AL. 1997 12 CONSTANT WREB0 = 0.030 ! REST OF THE BODY, WANG ET AL. 1997 13 CONSTANT WLIB0 = 0.266 ! LIVER, WANG ET AL. 1997 14 15 !EXPOSURE SCENARIO FOR UNIQUE OR REPETITIVE WEEKLY OR MONTHLY EXPOSURE 16 !NUMBER OF EXPOSURES PER WEEK o o 17 CONSTANT WEEK LACK = ! DELAY BEFORE EXPOSURE ENDS 18 (WEEK) 19 CONSTANT WEEK PERIOD = 168 ! NUMBER OF HOURS IN THE WEEK 2 0 (HOURS) 21 CONSTANT WEEK FINISH = 168 ! TIME EXPOSURE ENDS (HOURS) 2 2 !NUMBER OF EXPOSURES PER MONTH 2 3 CONSTANT MONTH_LACK = 0.0 ! DELAY BEFORE EXPOSURE BEGINS 2 4 (MONTH) 25 2 6 !SET FOR BACKGROUND EXPOSURE=========== 2 7 !TIME CONSTANT FOR BACKGROUND EXPOSURE=========== 2 8 CONSTANT Day_LACK_BG = 0.0 ! DELAY BEFORE EXPOSURE BEGINS 2 9 (HOUR) 3 0 CONSTANT Day_PERIOD_BG = 24.0 ! LENGTH OF EXPOSURE (HOUR) 31 3 2 !TIME CONSTANT FOR WEEKLY EXPOSURE 33 CONSTANT WEEK_LACK_BG = 0.0 ! DELAY BEFORE BACKGROUND EXPOSURE 3 4 BEGINS (WEEK) 3 5 CONSTANT WEEK_PERIOD_BG = 168.0 ! NUMBER OF HOURS IN THE WEEK 3 6 (HOURS) 3 7 CONSTANT WEEK FINISH BG = 168.0 ! TIME EXPOSURE ENDS (HOURS) 38 3 9 ! CONSTANT USED IN CARDIAC OUTPUT EQUATION 4 0 CONSTANT QCC = 15.36 ! (L/KG-H), EMOND ET AL. 41 2004 42 4 3 ! COMPARTMENT LIPID EXPRESSED AS THE FRACTION OF TOTAL LIPID 4 4 !Data from Emonds Thesis 2001 4 5 CONSTANT F TOTLIP = 0.8000 ! ADIPOSE TISSUE 4 6 (UNITLESS) 4 7 CONSTANT B TOTLIP = 0.0057 ! BLOOD (UNITLESS) 4 8 CONSTANT RE TOTLIP = 0.0190 ! REST OF THE BODY 4 9 (UNITLESS) 5 0 CONSTANT LI TOTLIP = 0.0670 ! LIVER (UNITLESS) 51 CONSTANT MEANLIPID = 974.0 52 53 END ! END OF THE INITIAL SECTION 54 55 5 6 DYNAMIC ! DYNAMIC SIMULATION SECTION This document is a draftfor review purposes only and does not constitute Agency policy. C-13 DRAFT--DO NOT CITE OR QUOTE 1! 2 ALGORITHM IALG =2 ! GEAR METHOD 3 CINTERVAL CINT = 10.0 ! COMMUNICATION INTERVAL 4 MAXTERVAL MAXT = 1.0e +10 !MAXIMUM INTERVAL CALCULATION 5 MINTERVAL MINT = 1.0E-10 !MINIMUM INTERVAL CALCULATION 6 VARIABLE T = 0.0 7 CONSTANT TIMELIMIT = 1.752e5 !SIMULATION LIMIT TIME (HOUR) 8 CONSTANT Y0 = 0.0 ! AGE (YEARS) AT BEGINNING OF 9 SIMULATION 10 CONSTANT GROWON = 1.0 ! INCLUDE BODY WEIGHT AND HEIGHT 11 GROWTH? (1 = YES, 0 = NO) 12 CINTXY = CINT 13 PFUNC = CINT 14 15 DAY=T/24.0 ! TIME IN DAYS 16 WEEK =T/168.0 ! TIME IN WEEKS 17 MONTH =T/730.0 ! TIME IN MONTHS 18 YEAR=Y0+T/8760.0 ! TIME IN YEARS 19 GYR =Y0 + growon*T/8760.0 ! TIME FOR USE IN GROWTH EQUATION (YEARS) 20 21 DERIVATIVE ! PORTION OF CODE THAT SOLVES DIFFERENTIAL EQUATIONS 22 2 3 ! CHRONIC OR SUBCHRONIC EXPOSURE SCENARIO 2 4 ! NUMBER OF EXPOSURES PER DAY 2 5 DAY_LACK = EXP_TIME_ON DELAY BEFORE EXPOSURE BEGINS (HOURS) 2 6 DAY_PERIOD = DAY_CYCLE EXPOSURE PERIOD (HOURS) 2 7 DAY_FINISH = CINTXY LENGTH OF EXPOSURE (HOURS) 2 8 MONTH_PERIOD = TIMELIMIT EXPOSURE PERIOD (MONTHS) 2 9 MONTH FINISH = EXP TIME OFF LENGTH OF EXPOSURE (MONTHS) 30 31 3 2 ! NUMBER OF EXPOSURES PER DAY AND MONTH 33 DAY_FINISH_BG = CINTXY 3 4 MONTH_LACK_BG = BCK_TIME_ON !DELAY BEFORE BACKGROUD EXPOSURE BEGINS 3 5 (MONTHS) 3 6 MONTH_PERIOD_BG = TIMELIMIT ! BACKGROUND EXPOSURE PERIOD (MONTHS) 3 7 MONTH_FINISH_BG = BCK_TIME_OFF ! LENGTH OF BACKGROUND EXPOSURE (MONTHS) 38 3 9 B = 1.0-A ! FRACTION OF DIOXIN ABSORBED IN THE PORTAL FRACTION OF THE LIVER 40 41 !HUMAN BODY WEIGHT GROWTH EQUATION======== 4 2 ! POLYNOMIAL REGRESSION EXPRESSION WRITTEN 4 3 !APRIL 10 2008, OPTIMIZED WITH DATA OF PELEKIS ET AL. 2001 4 4 ! POLYNOMIAL REGRESSION EXPRESSION WRITTEN WITH 4 5 !HUH AND BOLCH 2003 FOR BMI CALCULATION 46 4 7 ! BODY WEIGHT CALCULATION 4 8 WT0 = (0.0006*GYR**3 - 0.0912*GYR**2 + 4.32*GYR + 3.652) 49 5 0 ! BODY MASS INDEX CALCULATION 51 BH = -2D-5*GYR**4+4.2D-3*GYR**3.0-0.315*GYR**2.0+9.7465*GYR+72.098 52 53 !HEIGHT EQUATION FORMULATED FOR USE FROM 0 TO 70 YEARS 5 4 BHM= (BH/100.0) !HUMAN HEIGHT IN METERS (BHM) 55 HBMI= WT0/(BHM**2.0) ! HUMAN BODY MASS INDEX (BMI) 56 This document is a draftfor review purposes only and does not constitute Agency policy. C-14 DRAFT--DO NOT CITE OR QUOTE 1 ! ADIPOSE TISSUE FRACTION 2 WT0GR= WT0*1.0e3 ! BODY WEIGHT IN GRAMS 3 WF0= -6.36D-20*WT0GR**4.0 +1.12D-14*WT0GR**3.0 -5.8D-10*WT0GR**2.0 +1.2D- 4 5*WT0GR+5.91D-2 5 6 ! LIVER,VOLUME, 7 ! APPROACH BASED ON LUECKE (2007) 8 WLI0= (3.59D-2 -(4.76D-7*WT0GR)+(8.50D-12*WT0GR**2.0)-(5.45D- 9 17*WT0GR**3.0)) 10 11 WRE0 = (0.91 -(WLIB0*WLI0+WFB0*WF0+WLI0+WF0))/(1.0+WREB0) 12 !REST OF THE BODY FRACTION; UPDATED FOR 13 EPA ASSESSMENT 14 QREF = 1.0-(QFF+QLIF) !REST OF BODY BLOOD FLOW 15 QTTQF = QFF+QREF+QLIF ! SUM MUST EQUAL 1 16 17 !COMPARTMENT VOLUME (L OR KG) 18 WF = WF0 * WT0 ADIPOSE 19 WRE = WRE0 * WT0 REST OF THE BODY 2 0 WLI = WLI0 * WT0 LIVER 21 WB=0.075*WT0 ! BLOOD 22 2 3 !COMPARTMENT TISSUE BLOOD (L OR KG) 2 4 WFB = WFB0 * WF ! ADIPOSE 2 5 WREB = WREB0 * WRE ! REST OF THE BODY 2 6 WLIB = WLIB0 * WLI ! LIVER 2 7 !CARDIAC OUTPUT FOR THE GIVEN BODY WEIGHT 2 8 QC= QCC*(WT0**0.75) ! [L BLOOD/HOUR] 29 3 0 QF = QFF*QC ! ADIPOSE TISSUE BLOOD FLOW RATE 31 [L/HR] 3 2 QLI = QLIF*QC ! LIVER TISSUE BLOOD FLOW RATE [L/HR] 33 QRE = QREF*QC !REST OF THE BODY BLOOD FLOW RATE [L/HR] 34 3 5 QTTQ = QF+QRE+QLI ! TOTAL FLOW RATE [L/HR] 36 3 7 (PERMEABILITY ORGAN FLOW [L/HR] 3 8 PAF = PAFF*QF ADIPOSE 3 9 PARE = PAREF*QRE REST OF THE BODY 4 0 PALI = PALIF*QLI LIVER TISSUE 41 4 2 ! ABSORPTION SECTION 4 3 ! INTRAVENOUS 4 4 IV = DOSEIV_NM * WT0 (AMOUNT IN NMOL 4 5 MSTTBCKGR = MSTOT_NMBCKGR *WT0 (AMOUNT IN (NMOL) 4 6 MSTT = MSTOT NM * WT0 (AMOUNT IN NMOL 47 4 8 !REPETITIVE ORAL BACKGROUND EXPOSURE SCENARIOS 4 9 DAY_EXPOSURE_BG = PULSE(DAY_LACK_BG,DAY_PERIOD_BG,DAY_FINISH_BG) 5 0 WEEK_EXPOSURE_BG = PULSE(WEEK_LACK_BG,WEEK_PERIOD_BG,WEEK_FINISH_BG) 51 MONTH_EXPOSURE_BG = PULSE(MONTH_LACK_BG,MONTH_PERIOD_BG,MONTH_FINISH_BG) 52 53 MSTTCH_BG = (DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG)*MSTTBCKGR 54 MSTTFR_BG = MSTTBCKGR/CINT 55 5 6 CYCLE_BG =DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG This document is a draftfor review purposes only and does not constitute Agency policy. C-15 DRAFT--DO NOT CITE OR QUOTE 1 2 3 ! CONDITIONAL ORAL EXPOSURE (BACKGROUND EXPOSURE) 4 IF (MSTTCH_BG.EQ.MSTTBCKGR) THEN 5 ABSMSTT_GB= MSTTFR_BG 6 ELSE 7 ABSMSTT_GB = 0.0 8 END IF 9 10 11 (REPETITIVE ORAL MAIN EXPOSURE SCENARIO 12 DAY_EXPOSURE = PULSE(DAY_LACK,DAY_PERIOD,DAY_FINISH) 13 WEEK_EXPOSURE = PULSE(WEEK_LACK,WEEK_PERIOD,WEEK_FINISH) 14 MONTH_EXPOSURE = PULSE(MONTH_LACK,MONTH_PERIOD,MONTH_FINISH) 15 16 MSTTCH = (DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE)*MSTT 17 CYCLE = DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE 18 MSTTFR=MSTT/CINT 19 2 0 (CONDITIONAL ORAL EXPOSURE 21 IF (MSTTCH.EQ.MSTT) THEN 2 2 ABSMSTT= MSTTFR 2 3 ELSE 2 4 ABSMSTT = 0. 2 5 END IF 26 2 7 CYCLETOT=INTEG(CYCLE,0.0) 28 2 9 ( MASS Balance CHANGE IN THE LUMEN 3 0 RMSTT= -(KST+KABS)*MST+ABSMSTT +ABSMSTT_GB ( RATE OF CHANGE (NMOL/H) 31 MST = INTEG(RMSTT,0.) (AMOUNT REMAINING IN GI TRACT 3 2 (NMOL) 33 3 4 ! ABSORPTION IN LYMPH CIRCULATION 35 LYRMLUM = KABS*MST*A 3 6 LYMLUM = INTEG(LYRMLUM,0.0) 37 3 8 ! ABSORPTION IN PORTAL CIRCULATION 3 9 LIRMLUM = KABS*MST*B 4 0 LIMLUM = INTEG(LIRMLUM,0.0) 41 4 2 ! PERCENT OF DOSE REMAINING IN THE GI TRACT 4 3 PRCT_remain_GIT = 100.0*MST/(MSTT+1E-30) 44 4 5 !IV ABSORTPION SCENARIO --------- 4 6 IVR= IV/PFUNC ! RATE FOR IV INFUSION IN BLOOD 4 7 EXPIV= IVR * (1.0-STEP(PFUNC)) 4 8 IVDOSE = integ(EXPIV,0.0) 49 5 0 !SYSTEMIC BLOOD COMPARTMENT 51 ! MODIFICATION OCT 8 2009 5 2 CB=(QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM)/(QC+CLURI) ! 53 CA = CB CONCENTRATION (NMOL/L) 54 55 !CB=(QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM-RAURI)/QC ! 5 6 ! CA = CB ! CONCENTRATION (NMOL/L) This document is a draftfor review purposes only and does not constitute Agency policy. C-16 DRAFT--DO NOT CITE OR QUOTE 1 2 IURINARY EXCRETION BY KIDNEY 3 ! MODIFICATION OCT 8 2009 4 RAURI = CLURI *CB 5 AURI = INTEG(RAURI,0.0) 6 7 8 ICONCENTRATION UNIT 9 PRCT_B = 100.0*CB/(MSTT+1E-30) I PERCENT OF DOSE 10 CBSNGKGLIADJ = CB*MW/(0.55*B TOTLIP) Iserum concentration in lipid adjust 11 (PG/G LIPID=PPT) 12 CBPPT = CBSNGKGLIADJ 13 CBNGKG = CB*MW 14 15 CBpptRH = CB*MW*10000/(0.55*MEANLIPID) ISERUM CONCENTRATION IN LIPID ADJUST 16 (PG/G LIPID=PPT) 17 18 AUC CBSNGKGLIADJ=INTEG(CBSNGKGLIADJ,0.0) 19 2 0 IADIPOSE TISSUE COMPARTMENT 21 RAFB= QF*(CA-CFB)-PAF*(CFB-CF/PF) !(NMOL/HR) 2 2 AFB = INTEG(RAFB,0.0) !(NMOL) 2 3 CFB = AFB/WFB !(NMOL/KG) 2 4 ITISSUE SUBCOMPARTMENT 2 5 RAF = PAF*(CFB-CF/PF) !(NMOL/HR) 2 6 AF = INTEG(RAF,0.0) !(NMOL) 2 7 CF = AF/WF !(NMOL/KG) 28 2 9 !POST SIMULATION UNIT CONVERSION 3 0 CFTOTAL = (AF + AFB)/(WF + WFB) ! TOTAL CONCENTRATION NMOL/ML 31 PRCT_F = 100.0*CFTOTAL/(MSTT+1E-30) 3 2 CFNGKG =CFTOTAL*MW 33 3 4 IREST OF THE BODY COMPARTMENT===== 3 5 RAREB= QRE*(CA-CREB)-PARE*(CREB-CRE/PRE) !(NMOL/HR) 3 6 AREB = INTEG(RAREB,0.0) !(NMOL) 3 7 CREB = AREB/WREB !(NMOL/KG) 3 8 ITISSUE SUBCOMPARTMENT 3 9 RARE = PARE*(CREB-CRE/PRE) !(NMOL/HR) 4 0 ARE = INTEG(RARE,0.0) !(NMOL) 41 CRE = ARE/WRE !(NMOL/KG) 42 4 3 !POST SIMULATION UNIT CONVERSION 4 4 CRETOTAL = (ARE + AREB)/(WRE + WREB) ! TOTAL CONCENTRATION IN NMOL/ML 4 5 PRCT RE = 100.0*CRETOTAL/(MSTT+1E-30) ! PERCENT OF DOSE 46 4 7 ILIVER COMPARTMENT 4 8 ITISSUE BLOOD SUBCOMPARTMENT 4 9 RALIB = QLI*(CA-CLIB)-PALI*(CLIB-CFLLIR)+LIRMLUM I(NMOL/HR) 5 0 ALIB = INTEG(RALIB,0.0) I(NMOL) 51 CLIB = ALIB/WLIB 5 2 ITISSUE SUBCOMPARTMENT 53 RALI = PALI*(CLIB-CFLLIR)-REXCLI I(NMOL/HR) 5 4 ALI = INTEG(RALI,0.0) I(NMOL) 55 CLI = ALI/WLI I(NMOL/KG) 56 This document is a draftfor review purposes only and does not constitute Agency policy. C-17 DRAFT--DO NOT CITE OR QUOTE 1 2 !FREE TCDD IN LIVER 3 ! MODIFICATION OCTOBER 8 2009 4 CFLLI= IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLIR)) & 5 +((CYP1A2_1O3*CFLLIR/(KDLI2+CFLLIR)*PAS_INDUC)))-CFLLI,CFLLI0) ! 6 CONCENTRATION OF FREE TCDD IN LIVER 7 CFLLIR=DIM(CFLLI,0.0) 8 9 !MODIFIED FROM: 10 !PARAMETER (LIVER_1RMN = 1.0E-30) 11 ! CFLLI= IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLIR & 12 +LIVER_1RMN))+((CYP1A2_1O3*CFLLIR/(KDLI2+CFLLIR & 13 ! +LIVER_1RMN)*PAS_INDUC)))-CFLLI,CFLLI0) 14 ! CFLLIR=DIM(CFLLI,0.0) 15 16 17 CBNDLI= LIBMAX*CFLLIR/(KDLI+CFLLIR) !CONC OF TCDD BOUDN TO AhR 18 19 !CBNDLI= LIBMAX*CFLLIR/(KDLI+CFLLIR+LIVER_1RMN) !CONC BIND 20 21 !POST SIMULATION UNIT CONVERSION 2 2 CLITOTAL = (ALI + ALIB)/(WLI + WLIB) ! TOTAL CONCENTRATION IN NMOL/ML 2 3 PRCT_LI = 100.0*CLITOTAL/(MSTT+1.0E-30) 2 4 rec_occ_AHR= 100.0*CFLLIR/(KDLI+CFLLIR+1.0) ! PERCENT BOUND TO AhR 2 5 OCCUPANCY 2 6 PROT_occ_1A2= 100.0*CFLLIR/(KDLI2+CFLLIR) ! PERCENT BOUND TO 1A2 2 7 OCCUPANCY 2 8 CLINGKG= CLITOTAL*MW ![NG TCDD/KG] 2 9 CBNDLINGKG = CBNDLI*MW 30 31 !FRACTION INCREASE OF INDUCTION OF CYP1A2 3 2 fold_ind=CYP1A2_1OUT/CYP1A2_1A2 33 VARIATIONOFAC =(CYP1A2_1OUT-CYP1A2_1A2)/CYP1A2_1A2 34 3 5 !VARIABLE ELIMINATION BASED ON THE CYP1A2 3 6 KBILE_LI_T = Kelv*VARIATIONOFAC! 37 3 8 REXCLI = KBILE_LI_T*CFLLIR*WLI ! DOSE-DEPENDENT RATE OF BILLIARY EXCRETION 3 9 OF DIOXIN 4 0 EXCLI = INTEG(REXCLI,0.0) !TOTAL AMOUNT OF DIOXIN EXCRETED 41 4 2 !CHEMICAL IN CYP450 (1A2) COMPARTMENT 4 3 !PARAMETER FOR INDUCTION OF CYP1A2 44 4 5 CYP1A2_1KINP = CYP1A2_1KOUT*CYP1A2_1OUTZ ! BASAL RATE OF CYP1A2 PRODUCTION 4 6 SET EQUAL TO BASAL RATE OF DEGRDATION AT STEADY STATE 47 4 8 ! MODIFICATION OCTOBER 8 2009 4 9 CYP1A2_1OUT =INTEG(CYP1A2_1KINP * (1.0 + CYP1A2_1EMAX *(CBNDLI+1.0e-30)**HILL 50 & 51 /(CYP1A2_1EC50**HILL + (CBNDLI+1.0e-30)**HILL)) & 5 2 - CYP1A2_1KOUT*CYP1A2_1OUT, CYP1A2_1OUTZ) ! LEVELS OF CYP1A2 53 ! MODEIFIED FROM: 5 4 !PARAMETER (CYP1A2_1RMN = 1e-30) 55 !CYP1A2_1OUT =INTEG(CYP1A2_1KINP * (1 + CYP1A2_1EMAX *(CBNDLI & 56 ! +CYP1A2_1RMN)**HILL/(CYP1A2_1EC50 + (CBNDLI + CYP1A2_1RMN)**HILL) & This document is a draftfor review purposes only and does not constitute Agency policy. C-18 DRAFT--DO NOT CITE OR QUOTE 1! +CYP1A2_1RMN) - CYP1A2_1KOUT*CYP1A2_1& 2! OUT, CYP1A2_1OUTZ) 3 4 ! EQUATIONS INCORPORATING DELAY OF CYP1A2 PRODUCTION (NOT USED IN 5 SIMULATIONS) 6 CYP1A2_1RO2 = (CYP1A2_1OUT - CYP1A2_1O2)/ CYP1A2_1TAU 7 CYP1A2_1O2 =INTEG(CYP1A2_1RO2, CYP1A2_1A1) 8 CYP1A2_1RO3 = (CYP1A2_1O2 - CYP1A2_1O3)/ CYP1A2_1TAU 9 CYP1A2_1O3 =INTEG(CYP1A2_1RO3, CYP1A2_1A2) 10 11 !CHECK MASS BALANCE 12 BDOSE= LYMLUM+LIMLUM+IVDOSE 13 BMASSE = EXCLI+AURI+AFB+AF+AREB+ARE+ALIB+ALI 14 BDIFF = BDOSE-BMASSE 15 ! BODY BURDEN IN TERMS OF CONCENTRATION (NG/KG) 16 BBNGKG = (AFB+AF+AREB+ARE+ALIB+ALI)*MW/WT0 ! 17 18 !COMMAND END OF THE SIMULATION 19 TERMT (T.GE TIMELIMIT, 'Time limit has been reached 20 21 END ! END OF THE DERIVATIVE SECTION 2 2 END ! END OF THE DYTNAMIC SECTION 2 3 END ! END OF THE PROGRAM 24 25 C.2.1.2. Input File 2 6 % base file name = "TESTJULY2009.m" 2 7 %clear variable 2 8 output clear 2 9 prepare clear year T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 3 0 %output all 31 % PARAMETERS FOR SIMULATION 3 2 CINT = 1 %0.5 33 EXP_TIME_ON = 0. % TIME AT WHICH EXPOSURE BEGINS (HOUR) 3 4 EXP_TIME_OFF = 613200 %324120 % HOUR/YEAR !TIME AT WHICH EXPOSURE 3 5 ENDS (HOUR) 3 6 DAY_CYCLE = 24 NUMBER OF HOURS BETWEEN DOSES (HOUR) 3 7 BCK_TIME_ON = 613200 5 324120 TIME AT WHICH BACKGROUND EXPOSURE 3 8 BEGINS (HOUR) 3 9 BCK_TIME_OFF = 613200 5 324120 TIME AT WHICH BACKGROUND EXPOSURE 4 0 ENDS (HOUR) 41 TIMELIMIT = 613200 5 324120 5 324120 SIMULATION TIME LIMIT 4 2 (HOUR) 4 3 MSTOTBCKGR = 0. ORAL BACKGROUND EXPOSURE DOSE (UG/KG) 44 4 5 % oral dose oral dose oral dose 4 6 MSTOT = 9.97339283634997E-07 ORAL DAILY EXPOSURE DOSE (NG/KG) 4 7 DOSEIV =0 %NG/KG 4 8 % oral dose oral dose oral dose 49 50 MEANLIPID = 730 % 51 PAS INDUC= 1 % INDUCTION INCLUDED? (1=YES, 0=NO) 52 This document is a draftfor review purposes only and does not constitute Agency policy. C-19 DRAFT--DO NOT CITE OR QUOTE 1 C.2.2. H um an G estational M odel 2 C.2.2.1. Model Code 3 P R O G R A M : 'T h re e C o m p a rtm e n t P B P K M o d e l fo r T C D D in H u m a n (G e sta tio n )' 4 5 ! Parameters were change may 16, 2002 6 ! Come from {8MAI_CHR_PRE-EXP_GD} 7 ! Come from {12 Mouse GD}file 8 !******************************************** 9 !{{IMPORTANT-IMPORTANT-IMPORTANT-IMPORTANT}} 10 ! REDUCTION OF MOTHER AND FETUS COMPARTMENT 11 ! 2M_R_TCDD_JULY2002 ////(JULY 18,2002)//// 12 !TCDD_RED_4Species_2003_4 ////(APR 8 ,2003)//// 13 !TCDD_RED_4Species_2003_9 ////(APR 17 ,2003)//// 14 !TCDD_RED_4Species_2003_12 ////(APR 17 ,2003)//// 15 !***************************************************** 16 !APRIL 18 2003 17 !TCDD_4C_4SP_2003 ////(APR 18 ,2003)//// 18 ! was ''Gest 4 species 1.csl'' but update July 2009 19 2 0 !GEST HUM 0 45Y 4 ICF afterKKfix v3 humangestational.csl 21 !HUM_GESTATIONAL_ICF_F083109.csl_ _ 2 2 !HUM_GESTATIONAL_ICF_F100709.csl 23 !***************************************************** 24 2 5 !Legend/Legend/Legend/Legend/Legend/Legend/Legend/Legend/ 2 6 !Legend for this PBPK model 2 7 !Mating: control the tenure of exchange between fetus and 2 8 !Mother and also control imitated tissue growth 2 9 !Control: WTFE, WPLA0, QPLAF 3 0 !(for rat, mouse, human, and monkey) 31 !Control transfer from mother to fetus and fetus to mother by TRANSTIME ON 3 2 !SWITCH_trans = 0 NO TRANSFER _ 33 !SWITCH_trans = 1 TRANSFER OCCURS 3 4 ! These switches are also controlled by mating parameters 35 3 6 INITIAL ! 37 3 8 !SIMULATION PARAMETERS 3 9 CONSTANT PARA_ZERO = 1e-30 4 0 CONSTANT EXP_TIME_ON 0.0 !TIME AT WHICH EXPOSURE BEGINS (HOURS) 41 CONSTANT EXP_TIME_OFF 530.0 !TIME AT WHICH EXPOSURE ENDS (HOURS) 4 2 CONSTANT DAY_CYCLE 24.0 !NUMBER OF HOURS BETWEEN DOSES (HOURS) 4 3 CONSTANT BCK_TIME_ON 0.0 !TIME AT WHICH BACKGROUND EXPOSURE 4 4 BEGINS (HOURS) 4 5 CONSTANT BCK_TIME_OFF 0.0 !TIME AT WHICH BACKGROUND EXPOSURE ENDS 4 6 (HOURS) 4 7 CONSTANT TRANSTIME_ON 0.0 !CONTROL TRANSFER FROM MOTHER TO FETUS 4 8 AT 9 WEEKS OR 1512 HOURS OF GESTATION 49 5 0 ! INTRAVENOUS SEQUENCY 51 CONSTANT IV_LACK = 0.0 5 2 CONSTANT IV_PERIOD = 0.0 53 5 4 !PREGNANCY PARAMETER This document is a draftfor review purposes only and does not constitute Agency policy. C-20 DRAFT--DO NOT CITE OR QUOTE 1 CONSTANT MATTING = 0.0 IBEGINNING OF MATING (HOUR) 2 CONSTANT PFETUS = 4.0 IPARTITION COEFFICIENT 3 CONSTANT CLPLA_FET = 1.0e-3 ICLEARANCE TRANSFER FOR MOTHER TO FETUS 4 (L/HR) 5 6 ICONSTANT EXPOSURE CONTROL 7 IACUTE, SUBCHRONIC, CHRONIC EXPOSURE ===== 8 IOR BACKGROUND EXPOSURE (IN THIS CASE 3 TIMES A DAY)=== 9 CONSTANT MSTOTBCKGR = 0.0 I ORAL BACKGROUND EXPOSURE DOSE (NG/KG) 10 CONSTANT MSTOT = 0.0 I ORAL EXPOSURE DOSE (NG/KG) 11 12 IORAL ABSORPTION 13 I MSTT= MSTOT/1000 *WT0 *1/322*1000 IAMOUNT IN NMOL 14 MSTOT NM = MSTOT/MW ICONVERTS THE DOSE TO NMOL/KG 15 16 IINTRAVENOUS ABSORPTION 17 CONSTANT DOSEIV = 0.0 ! INJECTED DOSE (NG/KG) 18 DOSEIV_NM = DOSEIV/MW ! CONVERTS THE INJECTED DOSE TO NMOL/KG 19 CONSTANT DOSEIVLATE = 0.0 !INJECTED DOSE LATE (UG/KG) 2 0 DOSEIVNMlate = DOSEIVLATE/MW !AMOUNT IN NMOL/G 21 2 2 !INITIAL GUESS OF THE FREE CONCENTRATION IN THE LIGAND (COMPARTMENT 23 INDICATED BELOW)==== 2 4 CONSTANT CFLLI0 = 0.0 !LIVER (NMOL/L) 2 5 CONSTANT CFLPLA0 = 0.0 !PLACENTA (NMOL/L) 26 2 7 IBINDING CAPACITY (AhR) FOR NON LINEAR BINDING (COMPARTMENT INDICATED 2 8 BELOW) (NMOL/L) === 2 9 CONSTANT LIBMAX = 0.35 LIVER (NMOL/L) 3 0 CONSTANT PLABMAX = 0.2 TEMPORARY PARAMETER 31 3 2 IPROTEIN AFFINITY CONSTANTS (1A2 OR AhR, COMPARTMENT INDICATED BELOW) 33 (NMOL/ML)=== 3 4 CONSTANT KDLI = 0.1 ILIVER (AhR) (NMOL/L), WANG ET AL. 1997 3 5 CONSTANT KDLI2 = 40.0 ILIVER (1A2) (NMOL/L), EMOND ET AL. 3 6 2004 3 7 CONSTANT KDPLA = 0.1 IASSUME IDENTICAL TO KDLI (AhR) 38 3 9 IEXCRETION AND ABSORPTION CONSTANT 4 0 CONSTANT KST = 0.01 I GASTRIC RATE CONSTANT (HR-1), EMOND ET 41 AL. 2005 4 2 CONSTANT KABS = 0.06 I INTESTINAL ABSORPTION CONSTANT (HR-1), 4 3 EMOND ET AL. (2005) 44 4 5 !INTERSPECIES ELIMINATION CONSTANT 4 6 !TEST ELIMINATION VARIABLE, EMOND ET AL. 2005 4 7 CONSTANT KELV = 1.1e-3 !4.0D-3 ! INTERSPECIES VARIABLE 4 8 ELIMINATION CONSTANT (1/HOUR) 49 5 0 I ELIMINATION CONSTANTS 51 CONSTANT CLURI = 4.17e-8 I URINARY CLEARANCE (L/HR), EMOND ET AL. 5 2 2005 53 5 4 ! CONSTANT TO DIVIDE THE ABSORPTION INTO LYMPHATIC AND PORTAL FRACTIONS 55 CONSTANT A = 0.7 ! LYMPHATIC FRACTION, WANG ET AL. 1997 56 This document is a draftfor review purposes only and does not constitute Agency policy. C-21 DRAFT--DO NOT CITE OR QUOTE 1 !PARTITION COEFFICIENTS 2 CONSTANT PF = 1.0e2 ! ADIPOSE TISSUE/BLOOD, WANG ET AL. 1997 3 CONSTANT PRE = 1.5 ! REST OF THE BODY/BLOOD, WANG ET AL. 4 1997 5 CONSTANT PLI = 6.0 ! LIVER/BLOOD, WANG ET AL. 1997 6 CONSTANT PPLA = 1.5 ! TEMPORARY PARAMETER NOT CONFIGURED, 7 WANG ET AL. 1997 8 9 !PARAMETER FOR INDUCTION OF CYP 1A2, WANG ET AL. 1997 10 CONSTANT PAS INDUC = 1.0 ! INCLUDE INDUCTION? (1 = YES, 0 = NO) 11 CONSTANT CYP1A2 1OUTZ = 1.6e3 ! DEGRADATION CONCENTRATION CONSTANT OF 12 1A2 (NMOL/L) 13 CONSTANT CYP1A2 1A1 = 1.6e3 ! BASAL CONCENTRATION OF 1A1 (NMOL/L) 14 CONSTANT CYP1A2' 1EC50 = 1.3e2 ! DISSOCIATION CONSTANT TCDD-CYP1A2 15 (NMOL/L) 16 CONSTANT CYP1A2 1A2 = 1.6e3 !BASAL CONCENTRATION OF 1A2 (NMOL/ML) 17 CONSTANT CYP1A2 1KOUT = 0.1 ! FIRST ORDER RATE OF DEGRADATION (H-1) 18 CONSTANT CYP1A2' 1TAU = 0.25 !HOLDING TIME (H) 19 CONSTANT CYP1A2' 1EMAX = 9.3e3 ! MAXIMUM INDUCTION OVER BASAL EFFECT 2 0 (UNITLESS) 21 CONSTANT HILL = 0.6 !HILL CONSTANT; COOPERATIVELY LIGAND 2 2 BINDING EFFECT CONSTANT (UNITLESS) 23 2 4 !DIFFUSIONAL PERMEABILITY FRACTION. WANG ET AL (1997 2 5 CONSTANT PAFF = 0.12 ADIPOSE (UNITLESS) 2 6 CONSTANT PAREF = 0.03 REST OF THE BODY (UNITLESS) 2 7 CONSTANT PALIF = 0.35 LIVER (UNITLESS) 2 8 CONSTANT PAPLAF = 0.3 OPTIMIZED PARAMETER 29 3 0 !TISSUE BLOOD FLOW EXPRESSED AS A FRACTION OF CARDIAC OUTPUT, KRISHNAN 2007 31 CONSTANT QFF = 0.05 ! ADIPOSE TISSUE BLOOD FLOW FRACTION 3 2 (UNITLESS), KRISHNAN 2008 33 CONSTANT QLIF = 0.26 ! LIVER (UNITLESS), KRISHNAN 2008 34 3 5 !===FRACTION OF TISSUE BLOOD WEIGHT Wang et al . (1997) 3 6 CONSTANT WFB0 = 0.050 !ADIPOSE TISSUE, WANG ET AL. 1997 3 7 CONSTANT WREB0 = 0.030 !REST OF THE BODY, WANG ET AL. 1997 3 8 CONSTANT WLIB0 = 0.266 !LIVER, WANG ET AL. 1997 3 9 CONSTANT WPLAB0 = 0.500 !ASSUME HIGHLY VASCULARIZED 40 41 ! EXPOSURE SCENARIO FOR UNIQUE OR REPETITIVE WEEKLY OR MONTHLY EXPOSURE 4 2 ! NUMBER OF EXPOSURES PER WEEK 43 CONSTANT WEEK_LACK = 0.0 !DELAY BEFORE EXPOSURE ENDS (WEEK) 4 4 CONSTANT WEEK_PERIOD = 168.0 ! NUMBER OF HOURS IN THE WEEK (HOURS) 4 5 CONSTANT WEEK FINISH = 168.0 ! TIME EXPOSURE ENDS (HOURS) 46 4 7 ! NUMBER OF EXPOSURES PER MONTH 4 8 CONSTANT MONTH_LACK = 0.0 !DELAY BEFORE EXPOSURE BEGINS (MONTHS) 49 5 0 CONSTANT FOR BACKGROUND EXPOSURE=========== 51 CONSTANT Day_LACK_BG = 0.0 ! DELAY BEFORE EXPOSURE BEGINS (HOURS) 5 2 CONSTANT Day_PERIOD_BG = 24.0 !LENGTH OF EXPOSURE (HOURS) 53 5 4 ! NUMBER OF EXPOSURES PER WEEK 55 CONSTANT WEEK_LACK_BG = 0.0 !DELAY BEFORE BACKGROUD EXPOSURE BEGINS 5 6 (WEEK) This document is a draftfor review purposes only and does not constitute Agency policy. C-22 DRAFT--DO NOT CITE OR QUOTE 1 CONSTANT WEEK_PERIOD_BG = 168.0 ! NUMBER OF HOURS IN THE WEEK (HOURS) 2 CONSTANT WEEK FINISH BG = 168.0 !TIME EXPOSURE ENDS (HOURS) 3 4 5 ! CONSTANT USED IN CARDIAC OUTPUT EQUATION 6 CONSTANT QCC = 15.36 ![L/KG-H], EMOND ET AL. 2004 7 8 ! COMPARTMENT LIPID EXPRESSED AS THE FRACTION OF TOTAL LIPID 9 !Data from Emonds Thesis 2001 10 CONSTANT F TOTLIP = 0.8000 ! ADIPOSE TISSUE (UNITLESS) 11 CONSTANT B TOTLIP = 0.0057 ! BLOOD (UNITLESS) 12 CONSTANT RE TOTLIP = 0.0190 ! REST OF THE BODY (UNITLESS) 13 CONSTANT LI TOTLIP = 0.0670 ! LIVER (UNITLESS) 14 CONSTANT PLA TOTLIP = 0.019 ! PLACENTA (UNITLESS) 15 CONSTANT FETUS TOTLIP = 0.019 ! FETUS (UNITLESS) 16 17 CONSTANT MEANLIPID = 974 18 19 END ! END OF THE INITIAL SECTION 20 21 DYNAMIC ! DYNAMIC SIMULATION SECTION 22 2 3 ALGORITHM IALG 2 ! GEAR METHOD 2 4 CINTERVAL CINT 0.1 ! COMMUNICATION INTERVAL 2 5 MAXTERVAL MAXT 1.0e+10 ! MAXIMUM CALCULATION INTERVAL 2 6 MINTERVAL MINT 1.0E-10 ! MINIMUM CALCULATION INTERVAL 2 7 VARIABLE T 0.0 2 8 CONSTANT TIMELIMIT 100 !SIMULATION LIMIT TIME (HOUR) 2 9 CONSTANT Y0 0.0 ! AGE (YEARS) AT BEGINNING OF 3 0 SIMULATION 31 CONSTANT GROWON 1.0 ! INCLUDE BODY WEIGHT AND HEIGHT 3 2 GROWTH? (1 :YES, 0=NO) 33 3 4 CINTXY = CINT 3 5 PFUNC = CINT 36 3 7 !TIME TRANSFORMATION 3 8 DAY= T/24.0 3 9 WEEK =T/168.0 4 0 YEAR=Y0+T/8760.0 ! TIME IN YEARS 41 GYR =Y0 + growon*T/8760.0 ! TIME FOR USE IN GROWTH 4 2 EQUATION 43 4 4 DERIVATIVE ! PORTION OF CODE THAT SOLVES DIFFERENTIAL EQUATIONS 45 4 6 !====== CHRONIC OR SUBCHRONIC EXPOSURE SCENARIO 4 7 ! NUMBER OF EXPOSURES PER DAY 48 4 9 DAY_LACK EXP_TIME_ON DELAY BEFORE EXPOSURE BEGINS (HOURS) 5 0 DAY_PERIOD DAY_CYCLE EXPOSURE PERIOD (HOURS) 51 DAY_FINISH CINTXY LENGTH OF EXPOSURE (HOURS) 52 MONTH_PERIOD TIMELIMIT EXPOSURE PERIOD (MONTHS) 53 MONTH FINISH EXP TIME OFF LENGTH OF EXPOSURE (MONTHS) 54 55 5 6 ! NUMBER OF EXPOSURES PER DAY AND MONTH This document is a draftfor review purposes only and does not constitute Agency policy. C-23 DRAFT--DO NOT CITE OR QUOTE 1 DAY_FINISH_BG CINTXY 2 MONTH_LACK_BG BCK TIME ON !DELAY BEFORE BACKGROUND EXPOSURE BEGINS 3 (MONTHS) 4 MONTH_PERIOD_BG TIMELIMIT !BACKGROUND EXPOSURE PERIOD (MONTHS) 5 MONTH FINISH BG BCK TIME OFF !LENGTH OF BACKGROUND EXPOSURE (MONTHS) 6 7 ! INTRAVENOUS LATE 8 IV_FINISH = CINTXY 9 B = 1-A ! FRACTION OF DIOXIN ABSORBED IN THE PORTAL FRACTION OF THE LIVER 10 11 ! MOTHER BODY WEIGHT GROWTH EQUATION 12 ! MODIFICATION TO ADAPT THIS MODEL AT HUMAN MODEL 13 ! BECAUSE LINEAR DESCRIPTION IS NOT GOOD ENOUGH FOR MOTHER GROWTH 14 ! MOTHER BODY WEIGHT GROWTH 15 ! HUMAN BODY WEIGHT (0 TO 45 YEARS) 16 ! POLYNOMIAL REGRESSION EXPRESSION WRITTEN 17 !APRIL 10 2008, OPTIMIZED WITH DATA OF PELEKIS ET AL. 2001 18 ! POLYNOMIAL REGRESSION EXPRESSION WRITTEN WITH 19 !HUH AND BOLCH 2003 FOR BMI CALCULATION 20 21 ! BODY WEIGHT CALCULATION. UNIT IN KG FOR GESTATIONAL PORTION 22 2 3 WT0 = (0.0006*GYR**3 - 0.0912*GYR**2 + 4.32*GYR + 3.652) 24 2 5 !BODY MASS INDEX CALCULATION 26 2 7 BH = -2D-5*GYR**4+4.2D-3*GYR**3.0-0.315*GYR**2.0+9.7465*GYR+72.098 2 8 !HEIGHT EQUATION FORMULATED FOR USE FROM 0 TO 70 YEARS 2 9 BHM= (BH/100.0)!HUMAN HEIGHT IN METER (BHM) 3 0 HBMI= WT0/(BHM**2.0) ! HUMAN BODY MASS INDEX (BMI) 31 32 33 MODIFICATION IN KG 3 4 RTESTGEST= T-MATTING ! STARTING TIME FOR FETAL GROWTH 35 TESTGEST=DIM(RTESTGEST,0.0) 3 6 ! GROWTH OF FETAL TISSUE 3 7 GESTATTION_FE=((4d-15*TESTGEST**4 -3d-11*TESTGEST**3 +1d-7*TESTGEST**2 -8d- 3 8 5*TESTGEST +0.0608)) 3 9 WTFER= DIM(GESTATTION_FE,0.0) ! FETAL COMPARTMENT WEIGHT 4 0 WTFE= WTFER 41 4 2 !/////////////////////////////////////////////////////////////////////// 4 3 ! FAT GROWTH EXPRESSION LINEAR DURING PREGNANCY 4 4 ! FROM O'FLAHERTY_1992 4 5 !/////////////////////////////////////////////////////////////////////// 46 4 7 WT0GR= WT0*1.0e3 ! MOTHER BODY WEIGHT IN G 48 4 9 WF0 =(-6.36D-20*WT0GR**4.0 +1.12D-14*WT0GR**3.0 & 5 0 -5.8D-10*WT0GR**2.0+1.2D-5*WT0GR+5.91D-2) ! MOTHER FAT COMPARTMENT 51 GROWTH 52 53 !/////////////////////////////////////////////////////////////////////// 5 4 ! WPLA PLACENTA GROWTH EXPRESSION, SINGLE EXPONENTIAL WITH OFFSET 55 ! FROM O'FLAHERTY_1992 ! FOR EACH PUP 5 6 !/////////////////////////////////////////////////////////////////////// This document is a draftfor review purposes only and does not constitute Agency policy. C-24 DRAFT--DO NOT CITE OR QUOTE 1 !SAME EQUATION THEN THE FORST MODEL. BODY WEIGHT KEPT IN G 2 !A CORRECTION FOR THE BODY WEIGHT (WTO(KG)*1000 = WTOGR) 3 4 WPLA0N_HUMAN= (850*exp(-9.434*(exp(-5.23d-4*(TESTGEST))))) 5 WPLA0R = WPLA0N_HUMAN/WT0GR 6 WPLA0W = DIM(WPLA0R,0.0) ! PLACENTA WEIGHT 7 WPLA0=WPLA0W 8 9 !/////////////////////////////////////////////////////////////////////// 10 ! QPLA PLACENTA GROWTH EXPRESSION, DOUBLE EXPONENTIAL WITH OFFSET 11 ! FROM O'FLAHERTY_1992 12 !/////////////////////////////////////////////////////////////////////// 13 14 QPLAF_HUMAN= SWITCH_trans*((1d-10*TESTGEST**3.0 -5D-7*TESTGEST**2.0 15 +0.0017*TESTGEST+1.1937)/QC) 16 GEST_QPLAF=DIM(QPLAF_HUMAN,0.0) ! PLACENTA BLOOD FLOW RATE 17 QPLAF =GEST_QPLAF 18 19 ! LIVER,VOLUME (HUMAN 0 TO 70 YEARS) 2 0 ! APPROACH BASED ON LUECKE (2007) 21 WLI0= (3.59D-2 -(4.76D-7*WT0GR)+(8.50D-12*WT0GR**2.0)-(5.45D-17*WT0GR**3.0)) 2 2 ! LIVER VOLUME IN GROWING HUMAN 23 2 4 ! VARIABILITY OF REST OF THE BODY DEPENDS ON OTHER ORGAN 2 5 WRE0 = (0.91-(WLIB0*WLI0+WFB0*WF0+ WPLAB0*WPLA0 + WLI0 + WF0 + 2 6 WPLA0))/(1+WREB0) 2 7 QREF = 1-(QFF+QLIF+QPLAF) !REST BODY BLOOD FLOW (ML/HR) 2 8 QTTQF = QFF+QREF+QLIF+QPLAF ! SUM MUST EQUAL 1 29 3 0 COMPARTMENT TISSUE BLOOD VOLUME (L) 31 WF = WF0 * WT0 ! ADIPOSE TISSUE 3 2 WRE = WRE0 * WT0 ! REST OF THE BODY 33 WLI = WLI0 * WT0 ! LIVER 3 4 WPLA= WPLA0* WT0 ! PLACENTA 35 3 6 COMPARTMENT TISSUE VOLUME (L) 3 7 WFB = WFB0 * WF 3 8 WREB = WREB0 * WRE 3 9 WLIB = WLIB0 * WLI 4 0 WPLAB = WPLAB0* WPLA ! ADIPOSE TISSUE ! REST OF THE BODY ! LIVER ! PLACANTA 41 4 2 ! TOTAL VOLUME OF COMPARTMENT (L) 43 WFT = WF TOTAL ADIPOSE TISSUE 4 4 WRET = WRE TOTAL REST OF THE BODY 4 5 WLIT = WLI TOTAL LIVER TISSUE 4 6 WPLAT= WPLAB TOTAL PLACENTA TISSUE 47 4 8 ! CONSTANT USED IN CARDIAC OUTPUT EQUATION 49 5 0 ! UNIT CHANGED ON JULY 14 2009 (L/HR) 51 QC= QCC*(WT0)**0.75 52 53 QF = QFF*QC ! ADIPOSE TISSUE BLOOD FLOW RATE (L/HR) 5 4 QLI = QLIF*QC ! LIVER TISSUE BLOOD FLOW RATE (L/HR) 55 QRE = QREF*QC !REST OF THE BODY BLOOD FLOW RATE (L/HR) 5 6 QPLA = QPLAF*QC !PLACENTA TISSUE BLOOD FLOW RATE (L/HR) This document is a draftfor review purposes only and does not constitute Agency policy. C-25 DRAFT--DO NOT CITE OR QUOTE 1 QTTQ = QF+QRE+QLI+QPLA !TOTAL FLOW RATE (L/HR) 2 3 ! ========= DIFFUSIONAL PERMEABILITY FACTORS FRACTION ORGAN FLOW ========= 4 PAF = PAFF*QF ! ADIPOSE TISSUE BLOOD FLOW RATE (L/HR) 5 PARE = PAREF*QRE ! REST OF THE BODY BLOOD FLOW RATE 6 (L/HR) 7 PALI = PALIF*QLI ! LIVER TISSUE BLOOD FLOW RATE (L/HR) 8 PAPLA = PAPLAF*QPLA ! PLACENTA TISSUE BLOOD FLOW RATE (L/HR) 9 10 ************************************** 11 ABSORPTION SECTION 12 ORAL 13 INTRAPERITONEAL 14 SUBCUTANEOUS 15 INTRAVENOUS 16 ************************************** 17 18 (BACKGROUND EXPOSURE 19 (EXPOSURE FOR STEADY STATE CONSIDERATION 2 0 (REPETITIVE EXPOSURE SCENARIO 21 2 2 MSTOT_NMBCKGR = MSTOTBCKGR/322 (AMOUNT IN NMOL/G 2 3 MSTTBCKGR =MSTOT_NMBCKGR *WT0 24 2 5 DAY_EXPOSURE_BG = PULSE(DAY_LACK_BG,DAY_PERIOD_BG,DAY_FINISH_BG) 2 6 WEEK_EXPOSURE_BG = PULSE(WEEK_LACK_BG,WEEK_PERIOD_BG,WEEK_FINISH_BG) 2 7 MONTH_EXPOSURE_BG = PULSE(MONTH_LACK_BG,MONTH_PERIOD_BG,MONTH_FINISH_BG) 28 2 9 MSTTCH_BG = (DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG)*MSTTBCKGR 3 0 MSTTFR_BG = MSTTBCKGR/CINT 31 3 2 CYCLE_BG =DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG 33 3 4 ( CONDITIONAL ORAL EXPOSURE (BACKGROUND EXPOSURE) 35 3 6 IF (MSTTCH_BG.EQ.MSTTBCKGR) THEN 3 7 ABSMSTT_GB= MSTTFR_BG 3 8 ELSE 3 9 ABSMSTT_GB = 0.0 4 0 END IF 41 4 2 CYCLETOTBG=INTEG(CYCLE_BG,0.0) 43 4 4 (************************************** 4 5 (MULTIROUTE EXPOSURE 4 6 (REPETITIVE EXPOSURE SCENARIO 4 7 (************************************** 4 8 MSTT= MSTOT_NM * WT0 (AMOUNT IN NMOL 4 9 DAY_EXPOSURE = PULSE(DAY_LACK,DAY_PERIOD,DAY_FINISH) 5 0 WEEK_EXPOSURE = PULSE(WEEK_LACK,WEEK_PERIOD,WEEK_FINISH) 51 MONTH_EXPOSURE = PULSE(MONTH_LACK,MONTH_PERIOD,MONTH_FINISH) 52 53 MSTTCH = (DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE)*MSTT 54 55 MSTTFR = MSTT/CINT 56 This document is a draftfor review purposes only and does not constitute Agency policy. C-26 DRAFT--DO NOT CITE OR QUOTE 1 CYCLE = DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE 2 3 SUMEXPEVENT= INTEG (CYCLE,0.0) INUMBER OF CYCLES GENERATED DURING SIMULATION 4 5 ! CONDITIONAL ORAL EXPOSURE 6 IF (MSTTCH.EQ.MSTT) THEN 7 ABSMSTT= MSTTFR 8 ELSE 9 ABSMSTT = 0.0 10 END IF 11 12 13 CYCLETOT=INTEG(CYCLE,0.0) 14 15 ! MASS CHANGE IN THE LUMEN 16 RMSTT= -(KST+KABS)*MST +ABSMSTT +ABSMSTT_GB ! RATE OF CHANGE (NMOL/H) 17 MST = INTEG(RMSTT,0.0) IAMOUNT REMAINING IN DUODENUM 18 (NMOL) 19 2 0 ! ABSORPTION IN LYMPH CIRCULATION 21 LYRMLUM = KABS*MST*A 2 2 LYMLUM = INTEG(LYRMLUM,0.0) 23 2 4 ! ABSORPTION IN PORTAL CIRCULATION 2 5 LIRMLUM = KABS*MST*B 2 6 LIMLUM = INTEG(LIRMLUM,0.0) 27 28 2 9 IIV ABSORPTION SCENARIO--------- 3 0 IV= DOSEIV_NM * WT0 IAMOUNT IN NMOL 31 IVR= IV/PFUNC ! RATE FOR IV INFUSION IN BLOOD 3 2 EXPIV= IVR * (l-STEP(PFUNC)) 33 IVDOSE = integ(EXPIV,0.0) 34 3 5 IIV LATE IN THE CYCLE 3 6 IMODIFICATION JANUARY 13 2004 3 7 IV_RlateR = DOSEIVNMlate*WT0 3 8 IV_EXPOSURE=PULSE(IV_LACK,IV_PERIOD,IV_FINISH) 39 4 0 IV_lateT = IV_EXPOSURE *IV_RlateR 41 IV_late = IV_lateT/CINT 42 4 3 SUMEXPEVENTIV= integ(IV_EXPOSURE,0.0) INUMBER OF CYCLE GENERATE DURING 4 4 SIMULATION 45 4 6 ISYSTEMIC BLOOD COMPARTMENT 4 7 I MODIFICATION OCT 8 2009 4 8 CB=(QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM+QPLA*CPLAB+IV_late)/(QC+CLURI) I 4 9 CA = CB I CONCENTRATION (NMOL/L) 50 51 ICB=(QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM+QPLA*CPLAB+IV_late-RAURI)/QC 5 2 I(NMOL/L) 53 5 4 IURINARY EXCRETION BY KIDNEY 55 I MODIFICATION OCT 8 2009 5 6 RAURI = CLURI *CB This document is a draftfor review purposes only and does not constitute Agency policy. C-27 DRAFT--DO NOT CITE OR QUOTE 1 AURI = INTEG(RAURI,0.0) 2 3 !RAURI = CLURI * CRE 4 !AURI = INTEG(RAURI,0.0) 5 6 !UNIT CONVERSION POST SIMULATION 7 CONSTANT MW=322 !MOLECULAR WEIGHT (NG/NMOL) 8 CONSTANT SERBLO = 0.55 9 CONSTANT UNITCORR = 1.0e3 10 11 CBSNGKGLIADJ = CB*MW/(0.55*B_TOTLIP) !NG SERUM LIPID ADJUSTED/KG 12 AUCBS_NGKGLIADJ=integ(CBSNGKGLIADJ,0.) 13 CBNGKG= CB*MW !NG/KG 14 PRCT_B = 100.0*CB/(MSTT+1E-30) !PERCENT OF ORAL DOSE IN BLOOD 15 PRCT_BIV = 100.0*CB/(IV_RlateR+1E-30) ! PERCENT OF IV DOSE IN BLOOD 16 17 !ADIPOSE COMPARMTENT 18 !TISSUE BLOOD SUBCOMPARTMENT 19 RAFB= QF*(CA-CFB)-PAF*(CFB-CF/PF) !(NMOL/H) 2 0 AFB = INTEG(RAFB,0.0) !(NMOL) 21 CFB = AFB/WFB !(NMOL/L) 2 2 !TISSUE SUBCOMPARTMENT 2 3 RAF = PAF*(CFB-CF/PF) !(NMOL/H) 2 4 AF = INTEG(RAF,0.0) !(NMOL) 2 5 CF = AF/WF !(NMOL/L) 26 2 7 !UNIT CONVERSION POST SIMULATION 2 8 CFTOTAL= (AF + AFB)/(WF + WFB) ! TOTAL CONCENTRATION IN NMOL/ML 2 9 PRCT_F = 100.0*CFTOTAL/(MSTT+1E-30) !PERCENT OF ORAL DOSE IN FAT 3 0 PRCT_FIV = 100.0*CFTOTAL/(IV_RlateR+1E-30) !PERCENT OF IV DOSE IN FAT 31 CFNGKG=CFTOTAL*MW ! FAT CONCENTRATION IN NG/KG 3 2 AUCF_NGKGH=integ(CFNGKG,0.) 33 34 3 5 !REST OF THE BODY COMPARTMENT 3 6 !TISSUE BLOOD SUBCOMPARTMENT 3 7 RAREB= QRE *(CA-CREB)-PARE*(CREB-CRE/PRE) !(NMOL/H) 3 8 AREB = INTEG(RAREB,0.0) !(NMOL) 3 9 CREB = AREB/WREB !(NMOL/L) 4 0 !TISSUE SUBCOMPARTMENT 41 RARE = PARE*(CREB - CRE/PRE) !(NMOL/H) 4 2 ARE = INTEG(RARE,0.0) !(NMOL) 4 3 CRE = ARE/WRE !(NMOL/L) 4 4 ARETOT = ARE +AREB 45 4 6 !POST SIMULATION UNIT CONVERSION 4 7 CRETOTAL= (ARE + AREB)/(WRE + WREB) ! TOTAL CONCENTRATION (NMOL/L) 4 8 PRCT_RE = 100.0*CRETOTAL/(MSTT+1E-30) ! PERCENT OF ORAL DOSE IN REST OF BODY 4 9 PRCT_REIV = 100.0*CRETOTAL/(IV_RlateR+1E-30) ![ PERCENT OF IV DOSE IN REST 5 0 OF BODY 51 CRENGKG=CRETOTAL*MW ! REST OF THE BODY 5 2 CONCENTRATION (NG/KG) 53 54 55 !LIVER COMPARTMENT 5 6 !TISSUE BLOOD SUBCOMPARTMENT This document is a draftfor review purposes only and does not constitute Agency policy. C-28 DRAFT--DO NOT CITE OR QUOTE 1 RALIB = QLI*(CA-CLIB)-PALI*(CLIB-CFLLIR)+LIRMLUM ! (NMOL/HR) 2 ALIB = INTEG(RALIB,0.0) !(NMOL) 3 CLIB = ALIB/WLIB !(NMOL/L) 4 !TISSUE SUBCOMPARMTENT 5 RALI = PALI*(CLIB - CFLLIR)-REXCLI ! (NMOL/HR) 6 ALI = INTEG(RALI,0.0) !(NMOL) 7 CLI = ALI/WLI !(NMOL/L) 8 9 !FREE TCDD CONCENTRATION IN LIVER 10 ! MODIFICATION OCTOBER 8 2009 11 CFLLI= IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLIR)) & 12 +((CYP1A2_1O3*CFLLIR/(KDLI2+CFLLIR)*PAS_INDUC)))-CFLLI,CFLLI0) 13 CFLLIR=DIM(CFLLI,0.0) ! FREE TCDD CONCENTRATION IN LIVER 14 !MODIFIED FROM: 15 !PARAMETER (LIVER_1RMN = 1.0E-30) 16 ! CFLLI= IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLIR & 17 !+LIVER_1RMN))+((CYP1A2_1O3*CFLLIR/(KDLI2 + CFLLIR & 18 !+LIVER_1RMN)*PAS_INDUC)))-CFLLI,CFLLI0) 19 !CFLLIR=DIM(CFLLI,0.0) 20 21 ! MODIFICATION OCTOBER 8 2009 2 2 CBNDLI= LIBMAX*CFLLIR/(KDLI+CFLLIR) !BOUND CONCENTRATION (NMOL/L) 23 2 4 !POST SIMULATION UNIT CONVERSION 2 5 CLITOTAL= (ALI + ALIB)/(WLI + WLIB) ! TOTAL CONCENTRATION (NMOL/L) 2 6 PRCT_LI = 100.0*CLITOTAL/(MSTT+1E-30) ! PERCENT OF ORAL DOSE IN LIVER 2 7 PRCT_LIIV = 100.0*CLITOTAL/(IV_RlateR+1E-30) ! PERCENT OF IV DOSE IN LIVER 2 8 Rec_occ= CFLLIR/(KDLI+CFLLIR) 2 9 CLINGKG=CLITOTAL*MW ! LIVER CONCENTRATION IN NG/KG 3 0 AUCLI_NGKGH=integ(CLINGKG,0.0) 31 CBNDLINGKG = CBNDLI*MW ! BOUND CONCENTRATION IN NG/KG 3 2 AUCBNDLI_NGKGH =INTEG(CBNDLINGKG,0.0) 33 3 4 !FRACTION INCREASE OF INDUCTION OF CYP1A2 35 fold_ind=CYP1A2_1OUT/CYP1A2_1A2 3 6 VARIATIONOFAC =(CYP1A2_1OUT-CYP1A2_1A2)/CYP1A2_1A2 37 3 8 !VARIABLE ELIMINATION BASED ON THE CYP1A2 3 9 ! MODIFICATION OCTOBER 8 2009 4 0 KBILE_LI_T = Kelv*VARIATIONOFAC! ! DOSE-DEPENDENT EXCRETION RATE CONSTANT 41 4 2 REXCLI = KBILE_LI_T*CFLLIR*WLI ! DOSE-DEPENDENT BILLIARY EXCRETION RATE 4 3 EXCLI = INTEG(REXCLI,0.0) 44 4 5 !KBILE LI T =((CYP1A2 1OUT-CYP1A2 1A2)/CYP1A2 1A2)*Kelv ! 46 47 4 8 ICHEMICAL IN CYP450 (1A2) COMPARTMENT 49 5 0 CYP1A2_1KINP = CYP1A2_1KOUT* CYP1A2_1OUTZ I BASAL PRODCUTION RATE OF CYP1A2 51 SET EQUAL TO BASAL DEGREDATION RATEI 52 53 I MODIFICATION OCTOBER 8 2009 5 4 CYP1A2_1OUT =INTEG(CYP1A2_1KINP * (1.0 + CYP1A2_1EMAX *(CBNDLI+1.0e-30)**HILL 55 & 5 6 /(CYP1A2_1EC50**HILL + (CBNDLI+1.0e-30)**HILL)) & This document is a draftfor review purposes only and does not constitute Agency policy. C-29 DRAFT--DO NOT CITE OR QUOTE 1 - CYP1A2_1KOUT*CYP1A2_1OUT, CYP1A2_1OUTZ) 2 IMODIFIED FROM: 3 IPARAMETER (CYP1A2_1RMN = 1E-30) 4 ICYP1A2_1OUT =INTEG(CYP1A2_1KINP * (1 + CYP1A2_1EMAX *(CBND& 5 ILI +CYP1A2_1RMN)**HILL/(CYP1A2_1EC50 + (CBNDLI + CYP1A2_1& 6 IRMN)**HILL) +CYP1A2_1RMN) - CYP1A2_1KOUT*CYP1A2_1& 7 IOUT, CYP1A2_1OUTZ) 8 9 I EQUATIONS INCORPORATING DELAY OF CYP1A2 PRODUCTION (NOT USED IN 10 SIMULATIONS) 11 CYP1A2_1RO2 = (CYP1A2_1OUT - CYP1A2_1O2)/ CYP1A2_1TAU 12 CYP1A2_1O2 =INTEG(CYP1A2_1RO2, CYP1A2_1A1) 13 14 CYP1A2_1RO3 = (CYP1A2_1O2 - CYP1A2_1O3)/ CYP1A2_1TAU 15 CYP1A2_1O3 =INTEG(CYP1A2_1RO3, CYP1A2_1A2) 16 17 IPLACENTA COMPARTMENT 18 ITISSUE BLOOD SUBCOMPARTMENT 19 RAPLAB= QPLA*(CA - CPLAB)-PAPLA*(CPLAB -CFLPLAR) I NMOL/HR) 2 0 APLAB = INTEG(RAPLAB,0.0) I (NMOL) 21 CPLAB = APLAB/(WPLAB+1E-30) I (NMOL/ML) 2 2 ITISSUE SUBCOMPARTMENT 23 RAPLA = PAPLA*(CPLAB-CFLPLAR)-RAMPF + RAFPM I (NMOL/HR) 2 4 APLA = INTEG(RAPLA,0.0) I (NMOL) 2 5 CPLA = APLA/(WPLA+1e-30) I (NMOL/ML) 26 2 7 I NEW EQUATION AUGUST 28 2009 2 8 PARAMETER (PARA_ZERO = 1.0E-30) 2 9 CFLPLA= IMPLC(CPLA-(CFLPLAR*PPLA +(PLABMAX*CFLPLAR/(KDPLA& 3 0 +CFLPLAR+PARA_ZERO)))-CFLPLA,CFLPLA0) 31 CFLPLAR=DIM(CFLPLA,0.0) 32 33 IPOST SIMULATION UNIT CONVERSION 3 4 CPLATOTAL = ((APLAB+APLA)/(WPLAB+WPLA)) 35 PRCT_PLA = (CPLATOTAL/(MSTT+1E-30))*100 3 6 PRCT_PLAIV = (CPLATOTAL/(IV_RlateR+1E-30))*100 37 3 8 IFETUS COMPARTMENT 3 9 RAFETUS= RAMPF-RAFPM 4 0 AFETUS=INTEG(RAFETUS,0.0) 41 CFETUS=AFETUS/(WTFE+1.0e-30) 4 2 CFETOTAL= CFETUS 43 CFETUS_v = CFETUS/PFETUS 44 4 5 IPOST SIMULATION UNIT CONVERSION 4 6 CFETUSNGKG = CFETUS*MW I(NG/KG) 4 7 PRCT_FE = 100.0*CFETOTAL/(MSTT+1E-30) 4 8 PRCT_FEIV = 100.0*CFETOTAL/(IV_RlateR+1E-30) 49 5 0 ITRANSFER OF DIOXIN FROM PLACENTA TO FETUS 51 IFETAL EXPOSURE ONLY DURING EXPOSURE 52 53 IF (T.LT.TRANSTIME_ON) THEN 5 4 SWITCH_trans = 0.0 55 ELSE 5 6 SWITCH_trans = 1 This document is a draftfor review purposes only and does not constitute Agency policy. C-30 DRAFT--DO NOT CITE OR QUOTE 1 END IF 2 3 ITRANSFER OF DIOXIN FROM PLACENTA TO FETUS 4 ! MODIFICATION 26 SEPTEMBER 2003 5 6 RAMPF = (CLPLA_FET*CPLA)*SWITCH_trans 7 AMPF=INTEG(RAMPF,0.0) 8 9 ITRANSFER OF DIOXIN FROM FETUS TO PLACENTA 10 RAFPM = (CLPLA_FET*CFETUS_v)*SWITCH_transI 11 AFPM = INTEG(RAFPM,0.0) 12 13 ICHECK MASS BALANCE ---------- 14 BDOSE= IVDOSE +LYMLUM+LIMLUM 15 BMASSE = EXCLI+AURI+AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB+AFETUS I 16 BDIFF = BDOSE-BMASSE 17 18 IBODY BURDEN (NMOL) 19 BODY_BURDEN = AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB 20 21 IBODY BURDEN CONCENTRATION (NG/KG) 2 2 BBNGKG =(AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB)*MW/WT0 23 2 4 I END SIMULATION COMMAND 25 2 6 TERMT (T.GE. TimeLimit, 'Time limit has been reached.') 27 2 8 END END OF THE DERIVATIVE SECTION 2 9 END END OF THE DYNAMIC SECTION 3 0 END END OF THE PROGRAM 31 32 C.2.2.2. Input File 33 output @clear 3 4 prepare @clear T year CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 35 3 6 CINT = 1 %168 %100 %INTEGRATION TIME 3 7 %EXPOSURE SCENARIO 3 8 EXP_TIME_ON 0 % TIME AT WHICH EXPOSURE BEGINS (HOUR) 3 9 EXP_TIME_OFF 401190 %TIME AT WHICH EXPOSURE ENDS (HOUR) 4 0 DAY_CYCLE 24 NUMBER OF HOURS BETWEEN DOSES (HOUR) 41 BCK_TIME_ON 401190 %TIME AT WHICH BACKGROUND EXPOSURE BEGINS 4 2 (HOUR) 43 BCK_TIME_OFF 401190 %TIME AT WHICH BACKGROUND EXPOSURE ENDS (HOUR) 4 4 IV_LACK 401190 4 5 IV PERIOD 401190 4 6 %GESTATION CONTROL 4 7 MATTING = 393120 5 BEGINNING OF MATING (HOUR) AT 45 YEARS OLD 4 8 TIMELIMIT = 399840 SIMULATION TIME LIMIT (HOUR) 4 9 TRANSTIME_ON = 394632 5 TRANSFER FROM MOTHER TO FETUS AT 1512 HOURS 5 0 GESTATION 51 %EXPOSURE DOSE 5 2 MSTOT = 9.9733 9283634997E-07 % NG OF TCDD PER KG OF BW 53 MSTOTBCKGR = 0. %0.1 % ORAL BACKGROUND EXPOSURE DOSE (NG/KG) 54 DOSEIV = 0. %10 This document is a draftfor review purposes only and does not constitute Agency policy. C-31 DRAFT--DO NOT CITE OR QUOTE 1 DOSEIVLATE = 0. %10 2 3 % TRANFER MOTHER TO FETUS CLEARANCE 4 CLPLA FET = 0.001 % MOTHER TO FETUS TRANFER CLEARANCE (L/HR) 5 6 C.2.3. Rat Standard M odel 7 C.2.3.1. Model Code 8 P R O G R A M : 'T h re e C o m p a rtm e n t P B P K M o d e l in R at: S ta n d a rd M o d e l (N o n -G e sta tio n )' 9 10 !Rat Dioxin 3C June09 2clean icf afterKKfix v3 ratnongest.csl 11 !RAT_NON_GEST_ICF_F083109.CSL 12 !RAT_NON_GEST_ICF_F100609.CSL 13 !***************************************************** 14 15 INITIAL ! INITIALIZATION OF PARAMETERS 16 17 SIMULATION PARAMETERS 18 CONSTANT PARA_ZERO 19 CONSTANT EXP_TIME_ON = = 1d-30 0.0 ! TIME AT WHICH EXPOSURE BEGINS 2 0 (HOURS) 21 CONSTANT EXP_TIME_OFF = 900.0 ! TIME AT WHICH EXPOSURE ENDS 2 2 (HOURS) 23 CONSTANT DAY_CYCLE = 900.0 ! NUMBER OF HOURS BETWEEN 2 4 DOSES (HOURS) 2 5 CONSTANT BCK_TIME_ON = 0.0 ! TIME AT WHICH BACKGROUND 2 6 EXPOSURE BEGINS (HOURS) 2 7 CONSTANT BCK_TIME_OFF 2 8 EXPOSURE ENDS (HOURS) = 0.0 ! TIME AT WHICH BACKGROUND 29 3 0 CONSTANT MW=322 !MOLECULAR WEIGHT (NG/NMOL) 31 CONSTANT SERBLO = 0.55 3 2 CONSTANT UNITCORR = 1000 33 34 3 5 !EXPOSURE DOSES 3 6 CONSTANT MSTOTBCKGR = 0.0 !ORAL BACKGROUND EXPOSURE DOSE 3 7 (UG/KG) 3 8 CONSTANT MSTOT 3 9 CONSTANT MSTOTsc 4 0 (UG/KG) = 10 = 0.0 !ORAL EXPOSURE DOSE (UG/KG) SUBCUTANEOUS EXPOSURE DOSE 41 CONSTANT DOSEIV = 0.0 ! INJECTED DOSE (UG/KG) 42 4 3 !ORAL DOSE 4 4 MSTOT_NM 45 MSTOT_NMBCKGR MSTOT/MW !AMOUNT IN NMOL/G MSTOTBCKGR/MW !AMOUNT IN NMOL/G 46 4 7 !INTRAVENOUS DOSE 4 8 DOSEIV NM 49 DOSEIV/MW !AMOUNT IN NMOL/G 5 0 !INITIAL GUESS OF THE FREE CONCENTRATION IN THE LIGAND (COMPARTMENT 51 INDICATED BELOW)==== 5 2 CONSTANT CFLLI0 = 0.0 !LIVER (NMOL/ML) 53 This document is a draftfor review purposes only and does not constitute Agency policy. C-32 DRAFT--DO NOT CITE OR QUOTE 1 IBINDING CAPACITY (AhR) FOR NON LINEAR BINDING (COMPARTMENT INDICATED 2 BELOW) (NMOL/ML) === 3 CONSTANT LIBMAX = 3.5e-4 I LIVER (NMOL/ML), WANG ET AL. 4 1997 5 6 ! PROTEIN AFFINITY CONSTANTS (1A2 OR AhR, COMPARTMENT INDICATED BELOW) 7 (NMOL/ML)=== 8 CONSTANT KDLI = 1.0e-4 I LIVER (AhR) (NMOL/ML), WANG 9 ET AL. 1997 10 CONSTANT KDLI2 = 4.0e-2 ILIVER (1A2) (NMOL/ML), EMOND 11 ET AL. 2004 12 13 !EXCRETION AND ABSORPTION CONSTANT [RAT] 14 CONSTANT KST = 0.36 I GASTRIC RATE CONSTANT (HR-1), 15 WANG ET AL. (1997) 16 CONSTANT KABS = 0.48 IINTESTINAL ABSORPTION CONSTANT 17 (HR-1), WANG ET AL. 1997 18 19 !URINARY ELIMINATION CLEARANCE (ML/HR) 2 0 CONSTANT CLURI = 0.01 !URINARY CLEARANCE (ML/HR), 21 EMOND ET AL. 2004 22 2 3 IINTERSPECIES VARIABLE ELIMINATION 2 4 CONSTANT KELV = 0.15 I INTERSPECIES VARIABLE 2 5 ELIMINATION CONSTANT (1/HOUR) (OPTIMIZED), EMOND ET AL. 2004 26 2 7 I CONSTANT TO DIVIDE THE ABSORPTION INTO LYMPHATIC AND PORTAL FRACTIONS 2 8 CONSTANT A = 0.7 I LYMPHATIC FRACTION, WANG ET 2 9 AL. 1997 30 31 !PARTITION COEFFICIENTS 3 2 CONSTANT PF = 100 ! ADIPOSE TISSUE/BLOOD, WANG ET 33 AL. 1997 3 4 CONSTANT PRE = 1.5 ! REST OF THE BODY/BLOOD, WANG 3 5 ET AL. 1997 3 6 CONSTANT PLI = 6.0 ! LIVER/BLOOD, WANG ET AL. 3 7 1997 38 3 9 IPARAMETER FOR INDUCTION OF CYP 1A2 [MOUSE] === 4 0 CONSTANT PAS_INDUC = 1.0 I INCLUDE INDUCTION? (1 = YES, 41 0 = NO) 4 2 CONSTANT CYP1A2_1OUTZ = 1.6 I DEGRADATION CONCENTRATION 4 3 CONSTANT OF 1A2 (NMOL/ML), WANG ET AL. 997 4 4 CONSTANT CYP1A2_1A1 = 1.6 I BASAL CONCENTRATION OF 1A1 4 5 (NMOL/ML), WANG ET AL. 1997 4 6 CONSTANT CYP1A2_1EC50 = 0.13 I DISSOCIATION CONSTANT TCDD- 4 7 CYP1A2 (NMOL/ML) , WANG ET AL. 1997 4 8 CONSTANT CYP1A2_1A2 = 1.6 I BASAL CONCENTRATION OF 1A2 4 9 (NMOL/ML) Wang et al (1997) 5 0 CONSTANT CYP1A2_1KOUT = 0.1 I FIRST ORDER RATE OF 51 DEGRADATION (H-1), WANG ET AL. 1997 5 2 CONSTANT CYP1A2_1TAU = 0.25 I HOLDING TIME (H), WANG ET AL. 53 1997 5 4 CONSTANT CYP1A2_1EMAX = 600 I MAXIMUM INDUCTION OVER BASAL 55 EFFECT (UNITLESS), WANG ET AL. 1997 This document is a draftfor review purposes only and does not constitute Agency policy. C-33 DRAFT--DO NOT CITE OR QUOTE 1 CONSTANT HILL = 0.6 !HILL CONSTANT; COOPERATIVELY LIGAND 2 BINDING EFFECT CONSTANT (UNITLESS) 3 4 (TISSUE BLOOD FLOW EXPRESSED AS A FRACTION OF CARDIAC OUTPUT 5 CONSTANT QFF = 0.069 ! ADIPOSE TISSUE BLOOD FLOW 6 FRACTION (UNITLESS), WANG ET AL. 1997 7 CONSTANT QLIF = 0.183 ! LIVER (UNITLESS), WANG ET AL. 8 1997 9 10 !DIFFUSIONAL PERMEABILITY FRACTION 11 CONSTANT PAFF = 0.0910 ! ADIPOSE (UNITLESS), WANG ET 12 AL. 1997 13 CONSTANT PAREF = 0.0298 ! REST OF THE BODY (UNITLESS), 14 WANG ET AL. 1997 15 CONSTANT PALIF = 0.35 ! LIVER (UNITLESS), WANG ET AL. 16 1997 17 18 !FRACTION OF TISSUE VOLUME (UNITLESS) 19 CONSTANT WLI0 = 0.0360 LIVER, WANG ET AL. 1997 2 0 CONSTANT WF0 = 0.069 BLOOD, WANG ET AL. 1997 21 2 2 !COMPARTMENT TISSUE BLOOD EXPRESSED AS A FRACTION OF THE TOTAL 23 COMPARTMENT VOLUME ========= 2 4 CONSTANT WFB0 = 0.050 !ADIPOSE TISSUE, WANG ET AL. 2 5 1997 2 6 CONSTANT WREB0 = 0.030 !REST OF THE BODY, WANG ET AL. 2 7 1997 2 8 CONSTANT WLIB0 = 0.266 !LIVER , WANG ET AL. 1997 29 3 0 !EXPOSURE SCENARIO FOR UNIQUE OR REPETITIVE WEEKLY OR MONTHLY EXPOSURE 31 ! NUMBER OF EXPOSURES PER WEEK 3 2 CONSTANT WEEK_LACK = 0.0 ! DELAY BEFORE EXPOSURE ENDS 33 (WEEK) 3 4 CONSTANT WEEK_PERIOD = 168.0 ! NUMBER OF HOURS IN THE WEEK 3 5 (HOURS) 3 6 CONSTANT WEEK FINISH = 168.0 ! TIME EXPOSURE ENDS (HOURS) 37 3 8 !NUMBER OF EXPOSURES PER MONTH 3 9 CONSTANT MONTH_LACK = 0.0 ! DELAY BEFORE EXPOSURE BEGINS 4 0 (MONTH) 41 4 2 !SET FOR BACKGROUND EXPOSURE===== 4 3 (CONSTANT FOR BACKGROUND EXPOSURE^ 4 4 CONSTANT Day_LACK_BG = 0.0 ! DELAY BEFORE EXPOSURE BEGINS 4 5 (HOURS) 4 6 CONSTANT Day_PERIOD_BG = 24.0 ! LENGTH OF EXPOSURE (HOURS) 47 4 8 !NUMBER OF EXPOSURES PER WEEK 4 9 CONSTANT WEEK_LACK_BG = 0.0 ! DELAY BEFORE BACKGROUND 5 0 EXPOSURE (WEEK) 51 CONSTANT WEEK_PERIOD_BG = 168.0 !NUMBER OF HOURS IN THE WEEK 5 2 (HOURS) 53 CONSTANT WEEK FINISH BG = 168.0 ! TIME EXPOSURE ENDS (HOURS) 54 55 !GROWTH CONSTANT FOR RAT 5 6 !CONSTANT FOR MOTHER BODY WEIGHT GROWTH This document is a draftfor review purposes only and does not constitute Agency policy. C-34 DRAFT--DO NOT CITE OR QUOTE 1 CONSTANT BW_T0 = 250.0 ICHANGED FOR SIMULATION 2 3 ! CONSTANT USED IN CARDIAC OUTPUT EQUATION 4 CONSTANT QCCAR =311.4 ICONSTANT (ML/MIN/KG), WANG ET 5 AL. 6 7 I COMPARTMENT LIPID EXPRESSED AS THE FRACTION OF TOTAL LIPID 8 CONSTANT F_TOTLIP = 0.855 IADIPOSE TISSUE (UNITLESS) 9 CONSTANT B_TOTLIP = 0.0033 IBLOOD (UNITLESS) 10 CONSTANT RE_TOTLIP = 0.019 IREST OF THE BODY (UNITLESS) 11 CONSTANT LI TOTLIP = 0.06 ILIVER (UNITLESS) 12 13 END IEND OF THE INITIAL SECTION 14 15 DYNAMIC IDYNAMIC SIMULATION SECTION 16 17 ALGORITHM IALG 2 !GEAR METHOD 18 CINTERVAL CINT 0.1 !COMMUNICATION INTERVAL 19 MAXTERVAL MAXT 1.0e+10 !MAXIMUM CALCULATION INTERVAL 2 0 MINTERVAL MINT 1.0E-10 !MINIMUM CALCULATION INTERVAL 21 VARIABLE T 0.0 2 2 CONSTANT TIMELIMIT 900.0 !SIMULATION TIME LIMIT 2 3 (HOURS) 2 4 CINTXY = CINT 2 5 PFUNC = CINT 26 2 7 !TIME CONVERSION 2 8 DAY=T/24.0 ! TIME IN DAYS 2 9 WEEK =T/168.0 ! TIME IN WEEKS 3 0 MONTH =T/730.0 ! TIME IN MONTHS 31 YEAR=T/8760.0 ! TIME IN YEARS 32 33 3 4 DERIVATIVE ! PORTION OF CODE THAT SOLVES DIFFERENTIAL EQUATIONS 35 3 6 ICHRONIC OR SUBCHRONIC EXPOSURE SCENARIO ======= 3 7 INUMBER OF EXPOSURES PER DAY 3 8 DAY_LACK = EXP_TIME_ON I DELAY BEFORE EXPOSURE BEGINS 3 9 (HOURS) 4 0 DAY_PERIOD = DAY_CYCLE I EXPOSURE PERIOD (HOURS) 41 DAY_FINISH = CINTXY I LENGTH OF EXPOSURE (HOURS) 4 2 MONTH_PERIOD = TIMELIMIT I EXPOSURE PERIOD (MONTHS) 4 3 MONTH FINISH = EXP TIME OFF I LENGTH OF EXPOSURE (MONTHS) 44 4 5 !NUMBER OF EXPOSURES PER DAY AND MONTH 4 6 DAY_FINISH_BG = CINTXY ! LENGTH OF EXPOSURE (HOURS) 4 7 MONTH_LACK_BG = BCK_TIME_ON ! DELAY BEFORE BACKGROUND 4 8 EXPOSURE BEGINS (MONTHS) 4 9 MONTH_PERIOD_BG = TIMELIMIT ! BACKGROUND EXPOSURE PERIOD 5 0 (MONTHS) 51 MONTH_FINISH_BG = BCK_TIME_OFF ! LENGTH OF BACKGROUND EXPOSURE 5 2 (MONTHS) 53 54 55 B = 1-A FRACTION OF DIOXIN ABSORBED IN 5 6 THE PORTAL FRACTION OF THE LIVER This document is a draftfor review purposes only and does not constitute Agency policy. C-35 DRAFT--DO NOT CITE OR QUOTE 1 2 ! BODY WEIGHT GROWTH EQUATION======== 3 PARAMETER (BW_RMN = 1.0E-30) 4 WT0= (BW T0 *(1.0+(0.41*T)/(1402.5+T+BW RMN))) 5 6 IVARIABILITY OF REST OF THE BODY DEPEND OTHERS ORGAN 7 WRE0 = (0.91 - (WLIB0*WLI0 + WFB0*WF0 + WLI0 + WF0))/(1.0+WREB0) IREST OF 8 THE BODY FRACTION; UPDATED FOR EPA ASSESSMENT 9 QREF = 1.0-(QFF+QLIF) IREST OF BODY BLOOD FLOW 10 QTTQF = QFF+QREF+QLIF I SUM MUST EQUAL 1 11 12 COMPARTMENT VOLUME (G) ========= 13 WF = WF0 * WT0 ! ADIPOSE 14 WRE = WRE0 * WT0 ! REST OF THE BODY 15 WLI = WLI0 * WT0 ! LIVER 16 17 COMPARTMENT TISSUE BLOOD VOLUME (G) 18 WFB = WFB0 * WF ! ADIPOSE 19 WREB = WREB0 * WRE ! REST OF THE BODY 2 0 WLIB = WLIB0 * WLI ! LIVER 21 2 2 ICARDIAC OUTPUT FOR THE GIVEN BODY WEIGHT 2 3 QC= QCCAR*60.0*(WT0/UNITCORR)**0.75 24 2 5 I COMPARTMENT BLOOD FLOW (ML/HR) 2 6 QF = QFF*QC IADIPOSE TISSUE BLOOD FLOW RATE 2 7 QLI = QLIF*QC ILIVER TISSUE BLOOD FLOW RATE 2 8 QRE = QREF*QC IREST OF THE BODY BLOOD FLOW 2 9 RATE 3 0 QTTQ = QF+QRE+QLI I TOTAL FLOW RATE 31 3 2 !PERMEABILITY ORGAN FLOW (ML/HR) 33 PAF = PAFF*QF ADIPOSE 3 4 PARE = PAREF*QRE REST OF THE BODY 3 5 PALI = PALIF*QLI LIVER TISSUE 36 3 7 CONDITIONAL ORAL EXPOSURE (BACKGROUND EXPOSURE) 3 8 !EXPOSURE + !REPETITIVE EXPOSURE SCENARIO 3 9 IV= DOSEIV_NM * WT0 !AMOUNT IN NMOL 4 0 MSTT= MSTOT_NM * WT0 !AMOUNT IN NMOL 41 MSTTBCKGR =MSTOT NMBCKGR *WT0 42 4 3 IREPETITIVE ORAL BACKGROUND EXPOSURE SCENARIOS 4 4 DAY_EXPOSURE_BG = PULSE(DAY_LACK_BG,DAY_PERIOD_BG,DAY_FINISH_BG) 4 5 WEEK_EXPOSURE_BG = PULSE(WEEK_LACK_BG,WEEK_PERIOD_BG,WEEK_FINISH_BG) 4 6 MONTH_EXPOSURE_BG = PULSE(MONTH_LACK_BG,MONTH_PERIOD_BG,MONTH_FINISH_BG) 47 4 8 MSTTCH_BG = (DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG)*MSTTBCKGR 4 9 MSTTFR_BG = MSTTBCKGR/CINT 50 51 CYCLE BG =DAY EXPOSURE BG*WEEK EXPOSURE BG*MONTH EXPOSURE BG 52 53 IF (MSTTCH_BG.EQ.MSTTBCKGR) THEN 54 ABSMSTT_GB= MSTTFR_BG 55 ELSE 5 6 ABSMSTT GB = 0.0 This document is a draftfor review purposes only and does not constitute Agency policy. C-36 DRAFT--DO NOT CITE OR QUOTE 1 END IF 2 3 4 IREPETITIVE ORAL MAIN EXPOSURE SCENARIO 5 DAY_EXPOSURE = PULSE(DAY_LACK,DAY_PERIOD,DAY_FINISH) 6 WEEK_EXPOSURE = PULSE(WEEK_LACK,WEEK_PERIOD,WEEK_FINISH) 7 MONTH_EXPOSURE = PULSE(MONTH_LACK,MONTH_PERIOD,MONTH_FINISH) 8 9 MSTTCH = (DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE)*MSTT 10 CYCLE = DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE 11 MSTTFR = MSTT/CINT 12 13 SUMEXPEVENT= integ (CYCLE,0.0) INUMBER OF CYCLE GENERATE DURING SIMULATION 14 15 16 ICONDITIONAL ORAL EXPOSURE 17 IF (MSTTCH.EQ.MSTT) THEN 18 ABSMSTT= MSTTFR 19 ELSE 2 0 ABSMSTT = 0.0 21 END IF 22 23 CYCLETOT=INTEG(CYCLE,0.0) 24 2 5 IMASS CHANGE IN THE LUMEN 2 6 RMSTT = -(KST+KABS)*MST+ABSMSTT +ABSMSTT_GB I RATE OF CHANGE (NMOL/H) 2 7 MST = INTEG(RMSTT,0.0) IAMOUNT OF STAY IN DUODENUM (NMOL) 28 2 9 IABSORPTION IN LYMPH CIRCULATION 3 0 LYRMLUM = KABS*MST*A 31 LYMLUM = INTEG(LYRMLUM,0.0) 32 33 IABSORPTION IN PORTAL CIRCULATION 3 4 LIRMLUM = KABS*MST*B 3 5 LIMLUM = INTEG(LIRMLUM,0.0) 36 3 7 IPERCENT OF DOSE REMAINING IN THE GI TRACT 3 8 PRCT remain GIT = (MST/(MSTT+PARA ZERO))*100.0 39 4 0 IABSORPTION of Dioxin by IV route--------41 IVR= IV/PFUNC I RATE FOR IV INFUSION IN BLOOD 4 2 EXPIV= IVR * (1.0-STEP(PFUNC)) 4 3 IVDOSE = integ(EXPIV,0.0) 44 4 5 ISYSTEMIC BLOOD COMPARTMENT 4 6 I MODIFICATION ON OCTOBER 6, 2009 4 7 CB=(QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM)/(QC+CLURI) I 4 8 CA = CB 49 5 0 IURINARY EXCRETION BY KIDNEY 51 I MODIFICATION ON OCTOBER 6, 2009 5 2 RAURI = CLURI *CB 53 AURI = INTEG(RAURI,0.0) 54 55 ICONVERSION EQUATION POST SIMULATION 5 6 PRCT_B = (CB/(MSTT+PARA_ZERO))*100.0 This document is a draftfor review purposes only and does not constitute Agency policy. C-37 DRAFT--DO NOT CITE OR QUOTE 1 CBNGKG = CB*MW*UNITCORR ![NG/KG] 2 3 4 CBSNGKGLIADJ= (CB*MW*UNITCORR*(1.0/B_TOTLIP)*(1.0/SERBLO))![NG of TCDD 5 Serum/Kg OF LIPIP] 6 7 !ADIPOSE TISSUE COMPARTMENT 8 !TISSUE BLOOD SUBCOMPARTMENT 9 RAFB = QF*(CA-CFB)-PAF*(CFB-CF/PF) !(NMOL/HR) 10 AFB = INTEG(RAFB,0.0) !(NMOL) 11 CFB = AFB/WFB !(NMOL/ML) 12 !TISSUE SUBCOMPARTMENT 13 RAF = PAF*(CFB-CF/PF) !(NMOL/HR) 14 AF = INTEG(RAF,0.0) !(NMOL) 15 CF = AF/WF !(NMOL/ML) 16 17 !CONVERSION EQUATION POST SIMULATION 18 CFTOTAL = (AF + AFB)/(WF + WFB) !TOTAL CONCENTRATION IN NMOL/ML 19 PRCT_F = (CFTOTAL/(MSTT+PARA_ZERO))*100.0 ! PRCENT OF DOSE IN FAT 2 0 CFNGKG = CFTOTAL*MW*UNITCORR ! CONCENTRATION [NG/KG] 21 2 2 !REST OF THE BODY COMPARTMENT 2 3 ! TISSUE BLOOD SUBCOMPARTMENT 2 4 RAREB= QRE*(CA-CREB)-PARE*(CREB-CRE/PRE) !(NMOL/HR) 2 5 AREB = INTEG(RAREB,0.0) !(NMOL) 2 6 CREB = AREB/WREB !(NMOL/ML) 2 7 ! TISSUE COMPARTMENT 2 8 RARE = PARE*(CREB - CRE/PRE) !(NMOL/HR) 2 9 ARE = INTEG(RARE,0.0) !(NMOL) 3 0 CRE = ARE/WRE !(NMOL/ML) 31 3 2 !CONVERSION EQUATION POST SIMULATION 33 CRETOTAL= (ARE + AREB)/(WRE + WREB) ! TOTAL CONCENTRATION IN 3 4 NMOL/ML 3 5 PRCT_RE = (CRETOTAL/(MSTT+PARA_ZERO))*100.0 3 6 CTREPGG= CRETOTAL*MW*UNITCORR !(PG/ML) 3 7 AUC_REPPG = integ(CTREPGG,0.0) 38 3 9 !LIVER COMPARTMENT 4 0 !TISSUE BLOOD COMPARTMENT 41 RALIB = QLI*(CA-CLIB)-PALI*(CLIB-CFLLIR)+LIRMLUM !(NMOL/HR) 4 2 ALIB = INTeg(RALIB,0.0) !(NMOL) 4 3 CLIB = ALIB/WLIB 4 4 !TISSUE COMPARTMENT 4 5 RALI = PALI*(CLIB-CFLLIR)-REXCLI !(NMOL/HR) 4 6 ALI = integ(RALI,0.0) !(NMOL) 4 7 CLI = ALI/WLI !(NMOL/ML) 48 49 5 0 PARAMETER (LIVER_1RMN = 1.0E-30) 51 CFLLI= IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLIR & 52 +LIVER_1RMN))+((CYP1A2_1O3*CFLLIR/(KDLI2+CFLLIR & 53 +LIVER_1RMN)*PAS_INDUC)))-CFLLIR,CFLLI0) ! FREE TCDD CONCENTRATION IN LIVER 5 4 CFLLIR=DIM(CFLLI,0.0) 55 5 6 CBNDLI= LIBMAX*CFLLIR/(KDLI+CFLLIR+LIVER_1RMN) !BOUND CONCENTRATION This document is a draftfor review purposes only and does not constitute Agency policy. C -3 8 D R A FT -- D O N O T C IT E O R Q U O T E 1 2 CONVERSION EQUATION POST SIMULATION 3 CLITOTAL= (ALI + ALIB)/(WLI + WLIB) ! TOTAL CONCENTRATION IN 4 NMOL/ML 5 PRCT_LI = (CLITOTAL/(MSTT+PARA_ZERO))*100.0 6 rec_occ_AHR= (CFLLIR/(KDLI+CFLLIR+1))*100.0 ! PERCENT OF AhR 7 OCCUPANCY 8 PROT_occ_1A2= (CFLLIR/(KDLI2+CFLLIR))*100.0 ! PERCENT OF 1A2 9 OCCUPANCY 10 CLINGKG =(CLITOTAL*MW*UNITCORR) 11 CBNDLINGKG = CBNDLI*MW*UNITCORR 12 AUCLI_NGKGH=INTEG(CLINGKG,0.0) 13 CLINGG=CLITOTAL*MW 14 15 VARIABLE ELIMINATION HALF-LIFE BASED ON THE CONCENTRATION OF CYP1A2 16 KBILE_LI_T =((CYP1A2_1OUT-CYP1A2_1A2)/CYP1A2_1A2)*Kelv ! INDUCED BILIARY 17 EXCRETION RATE CONSTANT 18 19 REXCLI= (KBILE_LI_T*CFLLIR*WLI) ! DOSE-DEPENDENT BILIARY EXCRETION RATE 2 0 EXCLI = INTEG(REXCLI,0.0) 21 2 2 CHEMICAL IN CYP450 (1A2) COMPARTMENT 2 3 !===PARAMETER FOR INDUCTION OF CYP1A2 24 2 5 CYP1A2_1KINP = CYP1A2_1KOUT* CYP1A2_1OUTZ ! BASAL RATE OF CYP1A2 PRODUCTION 2 6 SET EQUAL TO BASAL RATE OF DEGREDATION 27 28 2 9 ! MODIFICATION ON OCTOBER 6, 2009 3 0 CYP1A2_1OUT =INTEG(CYP1A2_1KINP * (1.0 + CYP1A2_1EMAX *(CBNDLI+1.0e- 31 30)**HILL & 3 2 /(CYP1A2_1EC50**HILL + (CBNDLI+1.0e-30)**HILL)) &- 33 - CYP1A2 1KOUT*CYP1A2 1OUT, CYP1A2 1OUTZ) 34 3 5 ! EQUATIONS INCORPORATING DELAY OF CYP1A2 PRODUCTION (NOT USED IN 3 6 SIMULATIONS) 37 3 8 CYP1A2_1RO2 = (CYP1A2_1OUT - CYP1A2_1O2)/ CYP1A2_1TAU 3 9 CYP1A2_1O2 =INTEG(CYP1A2_1RO2, CYP1A2_1A1) 4 0 CYP1A2_1RO3 = (CYP1A2_1O2 - CYP1A2_1O3)/ CYP1A2_1TAU 41 CYP1A2_1O3 =INTEG(CYP1A2_1RO3, CYP1A2_1A2) 42 4 3 ! ------------ CHECK MASS BALANCE ---------- 4 4 BDOSE= LYMLUM+LIMLUM+IVDOSE 4 5 BMASSE = EXCLI+AURI+AFB+AF+AREB+ARE+ALIB+ALI 4 6 BDIFF = BDOSE-BMASSE 47 4 8 !--------------- BODY BURDEN------------------------- 4 9 BBNGKG =(((AFB+AF+AREB+ARE+ALIB+ALI)*MW)/(WT0/UNITCORR)) ! 5 0 ! ------------ e n d OF THE SIMULATION C O M M A N D ----------- 51 5 2 TERMT (T.GE. TimeLimit, 'Time limit has been reached.') 53 5 4 END END OF THE DERIVATIVE SECTION 55 END END OF THE DYNAMIC SIMULATION SECTION 5 6 END END OF THE PROGRAM. This document is a draftfor review purposes only and does not constitute Agency policy. C-39 DRAFT--DO NOT CITE OR QUOTE 1 C.2.3.2. Input Files 2 C .2.3.2.1. Cantoni et al. (1981). 3 output @clear 4 prepare @clear 5 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 6 7 %Cantoni et al. 1981 8 %protocol: oral exposure 1 dose/week for 45 weeks; female CD-COBS rats 9 %Rat Dioxin 3C June09 2clean.csl 10 %RAT_NON_GEST_ICF_F083109.CSL (now 09-11-09) 11 %dose levels: 0.01, 0.1, 1 ug/kg 1 dose/week for 45 weeks 12 %dose levels: 10, 100, 1000 ng/kg 1 dose/week for 45 weeks 13 %dose levels equivalent to: 1.43, 14.3 143 ng/kg 7 days/weeks for 45 weeks 14 15 MAXT 0.01 16 CINT 0.1 17 EXP_TIME_ON 0. %delay before begin exposure (HOUR) 18 EXP_TIME_OFF 7560 %TIME EXPOSURE STOP (HOUR) 19 DAY_CYCLE 168 2 0 BCK_TIME_ON 0. %DELAY BEFORE BACGROUND EXPOSURE (HOUR) 21 BCK_TIME_OFF 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 2 2 TIMELIMIT 7560 %SIMULATION LIMIT TIME (HOUR) 2 3 BW_T0 125 % Body weight at the beginning of the simulation 2 4 (g) 25 2 6 %EXPOSURE DOSE SCENARIOS (UG/KG) 2 7 %MSTOT = 0.01 % exposure dose ug/kg 2 8 %MSTOT = 0.1 % exposure dose ug/kg 2 9 MSTOT =1 % exposure dose ug/kg 30 31 C .2.3.2.2. Chu et al. (2007). 3 2 output @clear 33 prepare @clear 3 4 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG 35 3 6 % Chu et al. 2007 3 7 %protocol: oral exposure daily for 28 days 3 8 %dose levels: 0.0025, 0.025, 0.250, 1.0 ug/kg every day for 28 days 3 9 % dose levels = 2.5, 25, 250, 1000 ng/kg every day for 28 days 4 0 MAXT = 0.01 41 CINT = 0.1 4 2 EXP TIME ON = 0. %delay before begin exposure (HOUR) 5 weeks 4 3 after start of experiment (age = 12 weeks) 4 4 EXP TIME OFF = 672. %TIME EXPOSURE STOP (HOUR); 30 doses, 1 4 5 every two weeks 4 6 DAY_CYCLE = 24. % once every two weeks 4 7 BCK_TIME_ON = 0. %DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 4 8 BCK_TIME_OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 4 9 TIMELIMIT = 672. %SIMULATION LIMIT TIME (HOUR) 5 0 BW_T0 = 200. % Body weight at the beginning of the 51 simulation (g); corresponds to 12 week old female 52 53 %EXPOSURE DOSE SCENARIOS (UG/KG) This document is a draftfor review purposes only and does not constitute Agency policy. C-40 DRAFT--DO NOT CITE OR QUOTE 1 %MSTOT 2 %MSTOT 3 %MSTOT 4 MSTOT 0.0025 0.025 0.250 1.0 % ORAL EXPOSURE DOSE (UG/KG) % ORAL EXPOSURE DOSE (UG/KG) % ORAL EXPOSURE DOSE (UG/KG) % ORAL EXPOSURE DOSE (UG/KG) 5 C .2.3.2.3. Crofton et al. (2005). 6 output @clear 7 prepare @clear 8 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG 9 10 % Crofton et al. 2005 11 %protocol: oral expo: ure daily for 4 days o I 1-- CO o o CO o I 1-- o CO 12 %dose levels: 0.0001, 0.003, 0.01, 1, 3, and 10 ug/kg every 13 day for four days 14 %dose levels: 10, 30, 100, 300, 1000, 3000 and 10000 ng/kg every day 15 for four days 16 17 MAXT = 0.01 18 CINT = 0.1 19 EXP_TIME_ON = 0. %delay before begin exposure (HOUR) 5 weeks 2 0 after start of experiment (age 12 weeks) 21 EXP_TIME_OFF = 96. %TIME EXPOSURE STOP (HOUR); 30 doses, 1 2 2 every two weeks 23 DAY_CYCLE = 24. % once every two weeks 2 4 BCK_TIME_ON = 0. %DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 2 5 BCK_TIME_OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 2 6 TIMELIMIT = 96. %SIMULATION LIMIT TIME (HOUR) 2 7 BW T0 = 250 % Body weight at the beginning of the 2 8 simulation (g); corresponds to 12 week old female 29 3 0 %EXPOSURE DOSE SCENARIOS (UG/KG) 31 MSTOT = 0.0001 % ORAL EXPOSURE DOSE (UG/KG) 3 2 %MSTOT = 0.003 % ORAL EXPOSURE DOSE (UG/KG) 33 %MSTOT = 0.01 % ORAL EXPOSURE DOSE (UG/KG) 3 4 %MSTOT = 0.03 % ORAL EXPOSURE DOSE (UG/KG) 3 5 %MSTOT = 0.1 % ORAL EXPOSURE DOSE (UG/KG) 3 6 %MSTOT = 0.3 % ORAL EXPOSURE DOSE (UG/KG) 3 7 %MSTOT = 1. % ORAL EXPOSURE DOSE (UG/KG) 3 8 %MSTOT = 3. % ORAL EXPOSURE DOSE (UG/KG) 3 9 MSTOT = 10. % ORAL EXPOSURE DOSE (UG/KG) 40 41 4 2 C .2.3.2.4. Fattore et al. (2000). 4 3 output @clear 4 4 prepare @clear 4 5 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG 46 4 7 % Fattore et al. 2000 4 8 %built and check in August 7 2009 4 9 %protocol: oral exposure in diet for 13 weeks; SD rats 5 0 %dose levels: 0.02, 0.1, 0.2, 2 ug/kg 7 days/week for 13 weeks 51 %dose levels equivalent to: 20, 100, 200, 2000 ng/kg 7 days/week for 13 weeks 52 53 MAXT = 0.01 This document is a draftfor review purposes only and does not constitute Agency policy. C-41 DRAFT--DO NOT CITE OR QUOTE 1 CINT = 0 .1 2 EXP TIME ON = 0. %TIME AT WHICH EXPOSURE BEGINS (HOUR) 3 EXP TIME OFF = 2184 %TIME AT WHICH EXPOSURE ENDS (HOUR) 4 DAY CYCLE = 24 5 BCK TIME ON = 0. %TIME AT WHICH BACKGROUND EXPOSURE BEGINS (HOUR) 6 BCK TIME OFF = 0. %TIME AT WHICH BACKGROUND EXPOSURE ENDS (HOUR) 7 TIMELIMIT = 2184 %SIMULATION TIME LIMIT (HOUR) 8 BW T0 = 150 % BODY WEIGHT AT THE BEGINNING OF THE SIMULATION 9 (G) 10 11 %EXPOSURE DOSE SCENARIOS (UG/KG) 12 %MSTOT = 0.02 % EXPOSURE DOSE IN UG/KG 13 %MSTOT = 0.1 % EXPOSURE DOSE IN UG/KG 14 %MSTOT = 0.2 % EXPOSURE DOSE IN UG/KG 15 MSTOT =2 % EXPOSURE DOSE IN UG/KG 16 17 C .2.3.2.5. Franc et al. (2001). Sprague Dawley rats 18 output @clear 19 prepare @clear 2 0 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 21 2 2 % Franc et al. 2001 2 3 % Non-gestational rat model 2 4 % dose levels: 0.140, 0.420, and 1.400 ug/kg every 2 weeks for 22 weeks 2 5 % dose levels: 140, 420, and 1400 ng/kg every 2 weeks for 22 weeks 2 6 % dose levels equivalent to 10, 30, and 100 ng/kg/day 27 2 8 MAXT = 0.01 2 9 CINT = 0.1 3 0 EXP_TIME_ON = 0. %delay before begin exposure (HOUR) 31 EXP_TIME_OFF = 3696. %TIME EXPOSURE STOP (HOUR) 3 2 DAY_CYCLE = 336. 33 BCK_TIME_ON = 0. %DELAY BEFORE BACGROUND EXPOSURE (HOUR) 3 4 BCK_TIME_OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 35 TIMELIMIT = 3696. %SIMULATION LIMIT TIME (HOUR) 3 6 BW_T0 = 200. % Body weight at the beginning of the 3 7 simulation (g) corresponds to approximate weight of females 10 weeks old 38 3 9 %EXPOSURE DOSE SCENARIOS (UG/KG) 4 0 %MSTOT = 0.14 % ORAL EXPOSURE DOSE (UG/KG) 41 %MSTOT = 0.42 % ORAL EXPOSURE DOSE (UG/KG) 4 2 MSTOT = 1.4 % ORAL EXPOSURE DOSE (UG/KG) 43 4 4 C .2.3.2.6. Franc et al. (2001). Long-Evans rats 4 5 output @clear 4 6 prepare @clear 4 7 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 48 4 9 % Franc et al. 2001 5 0 % Non-gestational rat model 51 % dose levels: 0.140, 0.420, and 1.400 ug/kg every 2 weeks for 22 weeks 5 2 % dose levels: 140, 420, and 1400 ng/kg every 2 weeks for 22 weeks 53 % dose levels equivalent to 10, 30, and 100 ng/kg/day 54 This document is a draftfor review purposes only and does not constitute Agency policy. C-42 DRAFT--DO NOT CITE OR QUOTE 1 MAXT = 0.01 2 CINT = 0.1 3 EXP TIME ON = 0. %delay before begin exposure (HOUR) 4 EXP TIME OFF = 3696. %TIME EXPOSURE STOP (HOUR) 5 DAY CYCLE = 336. 6 BCK TIME ON = 0. %DELAY BEFORE BACGROUND EXPOSURE (HOUR) 7 BCK TIME OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 8 TIMELIMIT = 3696. %SIMULATION LIMIT TIME (HOUR) 9 BW T0 = 190. % Body weight at the beginning of the 10 simulation (g); corresponds to approximate weight of females 10 weeks old 11 12 %EXPOSURE DOSE SCENARIOS (UG/KG) 13 %MSTOT = 0.14 % ORAL EXPOSURE DOSE (UG/KG) 14 %MSTOT = 0.42 % ORAL EXPOSURE DOSE (UG/KG) 15 MSTOT = 1.4 % ORAL EXP 16 17 C .2.3.2.7. Franc et al. (2001). Hans Wistar rats 18 output @clear 19 prepare @clear 2 0 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 21 2 2 % Franc et al. 2001 2 3 % Non-gestational rat model 2 4 % dose levels: 0.140, 0.420, and 1.400 ug/kg every 2 weeks for 22 weeks 2 5 % dose levels: 140, 420, and 1400 ng/kg every 2 weeks for 22 weeks 2 6 % dose levels equivalent to 10, 30, and 100 ng/kg/day 27 2 8 MAXT = 0.01 2 9 CINT = 0.1 3 0 EXP_TIME_ON = 0. %delay before begin exposure (HOUR) 31 EXP_TIME_OFF = 3696. %TIME EXPOSURE STOP (HOUR) 3 2 DAY_CYCLE = 336. 33 BCK_TIME_ON = 0. %DELAY BEFORE BACGROUND EXPOSURE (HOUR) 3 4 BCK_TIME_OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 35 TIMELIMIT = 3696. %SIMULATION LIMIT TIME (HOUR) 3 6 BW_T0 = 205. % Body weight at the beginning of the 3 7 simulation (g) corresponds to approximate weight of females 10 weeks old 38 3 9 %EXPOSURE DOSE SCENARIOS (UG/KG) 4 0 %MSTOT = 0.14 % ORAL EXPOSURE DOSE (UG/KG) 41 %MSTOT = 0.42 % ORAL EXPOSURE DOSE (UG/KG) 4 2 MSTOT = 1.4 % ORAL EXP 43 4 4 C .2.3.2.8. Hassoun et al. (2000). 4 5 output @clear 4 6 prepare @clear 4 7 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 48 4 9 % Hassoun et al. 2000 5 0 %protocol: oral exposure for 13 weeks; SD rats 51 %dose levels: 0.003, 0.010, 0.022, 0.046 0.1 ug/kg 5 days/weeks for 13 weeks 5 2 %dose levels equivalent to: 3, 10, 22, 46 100 ng/kg 5 days/weeks for 13 weeks 53 %dose levels equivalent to: 2.14, 7.14, 15.7, 32.9 71.4 ng/kg 7 days/weeks 5 4 for 13 weeks This document is a draftfor review purposes only and does not constitute Agency policy. C-43 DRAFT--DO NOT CITE OR QUOTE 1 2 MAXT = 0.01 3 CINT = 0.1 4 EXP_TIME_ON = 0. %delay before begin exposure (HOUR) 5 EXP_TIME_OFF = 2184 %TIME EXPOSURE STOP (HOUR) 6 DAY_CYCLE = 24. 7 WEEK_PERIOD = 168. 8 WEEK_FINISH = 119. 9 BCK_TIME_ON = 0. %DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 10 BCK_TIME_OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 11 TIMELIMIT = 2184 %SIMULATION LIMIT TIME (HOUR) 12 BW T0 = 215. % Body weight at the beginning of the 13 simulation (g) 14 15 %EXPOSURE DOSE SCENARIOS (UG/KG) 16 %MSTOT = 0.003 exposure dose ug/kg 17 %MSTOT = 0.010 % exposure dose ug/kg 18 %MSTOT = 0.022 % exposure dose ug/kg 19 %MSTOT = 0.046 % exposure dose ug/kg 20 MSTOT = 0.1 % exposure dose ug/kg 21 2 2 C .2.3.2.9. H utt et al. (2008). 2 3 output @clear 2 4 prepare @clear 2 5 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 26 2 7 % Hutt et al. 2008 2 8 % Non-gestational rat model 2 9 % dose levels: 0.050 ug/kg every week for 13 weeks 3 0 % dose levels: 50 ng/kg every week for 13 weeks 31 % dose levels equivalent to 7.14 ng/kg/day 32 33 MAXT = 0.01 3 4 CINT = 0.1 35 EXP_TIME_ON = 0. %delay before begin exposure (HOUR) 3 6 EXP_TIME_OFF = 2184. %TIME EXPOSURE STOP (HOUR) 3 7 DAY_CYCLE = 168. 3 8 BCK_TIME_ON = 0. %DELAY BEFORE BACGROUND EXPOSURE (HOUR) 3 9 BCK_TIME_OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 4 0 TIMELIMIT = 2184. %SIMULATION LIMIT TIME (HOUR) 41 BW_T0 = 4.5 % Body weight at the beginning of the 4 2 simulation (g) corresponds to approximate weight of females 10 weeks old 43 4 4 %EXPOSURE DOSE SCENARIOS (UG/KG) 4 5 MSTOT = 0.05 % ORAL EXPOSURE DOSE (UG/KG) 46 4 7 C .2.3.2.10. Kitchin and Woods (1979) 4 8 output @clear 4 9 prepare @clear 5 0 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 51 5 2 % Kitchen and Woods 1979 53 %protocol: single oral gavage This document is a draftfor review purposes only and does not constitute Agency policy. C-44 DRAFT--DO NOT CITE OR QUOTE 1 %dose levels: 0.0006, 0.002, 0. 004, 0.020, 0.060, 0.200, 0.600, 2.000, 2 5.000, 20.000 ug/kg single oral gavage 3 % dose levels = 0.6, 2, 4, 20, 60, 200, 600, 2000, 5000, 20000 ng/kg single 4 oral gavage 5 MAXT = 0.001 6 CINT = 0.1 7 EXP TIME ON = 0. %delay before begin exposure (HOUR) 8 EXP TIME OFF = 24. %TIME EXPOSURE STOP (HOUR) 9 DAY CYCLE = 24. % daily 10 BCK TIME ON = 0. %DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 11 BCK TIME OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 12 TIMELIMIT = 24. %SIMULATION LIMIT TIME (HOUR) 13 BW T0 = 225. % Body weight at the beginning of the 14 simulation (g) 15 16 %EXPOSURE DOSE SCENARIOS (UG/KG) 17 %MSTOT 18 %MSTOT 19 %MSTOT 2 0 %MSTOT 21 %MSTOT 2 2 %MSTOT 2 3 %MSTOT 2 4 %MSTOT 2 5 %MSTOT = 0.0006 = 0.002 = 0.004 = 0.020 = 0.060 = 0.200 = 0.600 = 2.000 = 5.000 % ORAL EXPOSURE DOSE (UG/KG) % ORAL EXPOSURE DOSE (UG/KG) % ORAL EXPOSURE DOSE (UG/KG) % ORAL EXPOSURE DOSE (UG/KG) % ORAL EXPOSURE DOSE (UG/KG) % ORAL EXPOSURE DOSE (UG/KG) % ORAL EXPOSURE DOSE (UG/KG) % ORAL EXPOSURE DOSE (UG/KG) % ORAL EXPOSURE DOSE (UG/KG) 2 6 MSTOT = 20.000 % ORAL EXPOSURE DOSE (UG/KG) 27 2 8 C .2.3.2.11. Kociba et al. (1976) (13 weeks). 2 9 output @clear 3 0 prepare @clear 31 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 32 33 % Kociba et al. 1976. 3 4 %built and check in August 7 2009 3 5 %protocol: 5 days/week exposure for 13 weeks; SD rats 3 6 %Rat Dioxin 3C June09 2clean.csl 3 7 %RAT_NON_GEST_ICF_F083109.CSL (now 09-11-09) 3 8 %dose levels: 0.001, 0.01, 0.1, 1 ug/kg 5 days/weeks for 13 weeks 3 9 %dose levels: 1, 10, 100, 1000 ng/kg 5 days/weeks for 13 weeks 4 0 %dose levels equivalent to: 0.714, 7.14, 71.4, 714 ng/kg/d (adj) 7 days/weeks 41 for 13 weeks 42 4 3 MAXT = 0.001 4 4 CINT = 0.1 4 5 EXP_TIME_ON = 0. %delay before begin exposure (HOUR) 4 6 EXP_TIME_OFF = 2184 %TIME EXPOSURE STOP (HOUR) 4 7 WEEK_PERIOD = 168 4 8 WEEK_FINISH = 119 4 9 DAY_CYCLE = 24 5 0 BCK_TIME_ON = 0. %DELAY BEFORE BACGROUND EXPOSURE (HOUR) 51 BCK_TIME_OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 5 2 TIMELIMIT = 2184 %SIMULATION LIMIT TIME (HOUR) 53 BW_T0 = 180 % Body weight at the begeniong of the 5 4 simulation (g) 55 This document is a draftfor review purposes only and does not constitute Agency policy. C-45 DRAFT--DO NOT CITE OR QUOTE 1 %EXPOSURE DOSE SCENARIOS (UG/KG) 2 %MSTOT = 0.001 3 %MSTOT = 0.01 4 %MSTOT = 0.1 5 MSTOT =1 6 7 C .2.3.2.12. Kociba et al. (1978) (female) (104 weeks). 8 output @clear 9 prepare @clear 10 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG 11 12 % Kociba et al, 1978. 13 %built and check in August 7 2009 14 %protocol: daily dietary exposure for 104 weeks; SD rats 15 %dose levels: 0.001, 0.01, 0.1 ug/kg 7 days/week for 104 weeks 16 %dose levels: 1, 10, 100 ng/kg 7 days/week for 104 weeks 17 18 MAXT 0.01 19 CINT 0.1 2 0 EXP_TIME_ON 0. %TIME AT WHICH EXPOSURE BEGINS (HOUR) 21 EXP_TIME_OFF 17472 %TIME AT WHICH EXPOSURE ENDS (HOUR) 2 2 DAY_CYCLE 24 23 BCK_TIME_ON 0. %TIME AT WHICH BACKGROUND EXPOSURE BEGINS 2 4 (HOUR) 2 5 BCK_TIME_OFF 0. %TIME AT WHICH BACKGROUND EXPOSURE ENDS 2 6 (HOUR) 2 7 TIMELIMIT 17472 %SIMULATION TIME LIMIT (HOUR) 2 8 BW_T0 180 % BODY WEIGHT AT THE BEGINNING OF THE 2 9 SIMULATION (G) 30 31 %EXPOSURE DOSE SCENARIOS (UG/KG) 3 2 %MSTOT = 0.001 % EXPOSURE DOSE IN UG/KG 33 %MSTOT = 0.01 % EXPOSURE DOSE IN UG/KG 3 4 MSTOT = 0.1 % EXPOSURE DOSE IN UG/KG 35 3 6 C .2.3.2.13. Kociba et al. (1978) (male) (104 weeks). 3 7 output @clear 3 8 prepare @clear 3 9 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG 40 41 % Kociba et al, 1978. 4 2 %built and check in August 7 2009 4 3 %protocol: daily dietary exposure for 104 weeks; SD rats 4 4 %dose levels: 0.001, 0.01, 0.1 ug/kg 7 days/week for 104 weeks 4 5 %dose levels: 1, 10, 100 ng/kg 7 days/week for 104 weeks 46 4 7 MAXT = 0.01 4 8 CINT = 0.1 4 9 EXP TIME ON = 0. %TIME AT WHICH EXPOSURE BEGINS (HOUR) 5 0 EXP TIME OFF = 17472 %TIME AT WHICH EXPOSURE ENDS (HOUR) 51 DAY CYCLE = 24 5 2 BCK TIME ON = 0. %TIME AT WHICH BACKGROUND EXPOSURE BEGINS 53 (HOUR) This document is a draftfor review purposes only and does not constitute Agency policy. C-46 DRAFT--DO NOT CITE OR QUOTE 1 BCK_TIME_OFF 0. %TIME AT WHICH BACKGROUND EXPOSURE ENDS 2 (HOUR) 3 TIMELIMIT 17472 %SIMULATION TIME LIMIT (HOUR) 4 BW_T0 250 % BODY WEIGHT AT THE BEGINNING OF THE 5 SIMULATION (G) 6 7 %EXPOSURE DOSE SCENARIOS (UG/KG) 8 %MSTOT = 0.001 % EXPOSURE DOSE IN UG/KG 9 %MSTOT = 0.01 % EXPOSURE DOSE IN UG/KG 10 MSTOT = 0.1 % EXPOSURE DOSE IN UG/KG 11 12 C .2.3.2.14. Latchoumycandane and M athur (2002). 13 output @clear 14 prepare @clear 15 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 16 17 % Latchoumycandane and Mathur 2002. 18 %built and check in August 7 2009 19 %protocol: 1 time per day for 45 days oral gavage 2 0 %Rat Dioxin 3C June09 2clean.csl 21 %RAT_NON_GEST_ICF_F083109.CSL (now 09-11-09) 2 2 %dose levels: 0.001, 0.01, 0.1 ug/kg daily for 45 days 2 3 %dose levels: 1, 10, 100 ng/kg daily for 45 days 24 2 5 MAXT 0.01 2 6 CINT 0.1 2 7 EXP_TIME_ON 0. % delay before begin exposure (HOUR) 2 8 EXP_TIME_OFF 1080 % TIME EXPOSURE STOP (HOUR) 2 9 DAY_CYCLE 24 3 0 BCK_TIME_ON 0. % DELAY BEFORE BACGROUND EXPOSURE (HOUR) 31 BCK_TIME_OFF 0. % TIME OF BACKGROUND EXPOSURE STOP (HOUR) 3 2 TIMELIMIT 1080 % SIMULATION LIMIT TIME (HOUR) 33 BW_T0 200 % Body weight at the beginning of the 3 4 simulation (g) 35 3 6 %EXPOSURE DOSE SCENARIOS (UG/KG) 3 7 %MSTOT = 0.001 % exposure dose ug/kg 3 8 %MSTOT = 0.01 % exposure dose ug/kg 3 9 MSTOT = 0.1 % exposure dose ug/kg 40 41 4 2 C .2.3.2.15. L i et al. (1997). 4 3 output @clear 4 4 prepare @clear 4 5 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG 46 4 7 % Li et al 1997 4 8 % created 1/10/10 4 9 % Non-gestational rat model 5 0 % dose levels: 3, 10, 30, 100, 300, 1000, 3000, 10000, 30000 nkd one dose via 51 gavage, sacrificed 24 hrs later 52 53 MAXT = 0.1 5 4 CINT = 0.1 This document is a draftfor review purposes only and does not constitute Agency policy. C-47 DRAFT--DO NOT CITE OR QUOTE 1 EXP_TIME_ON 0. delay before begin exposure (HOUR) 2 EXP_TIME_OFF 24 , %TIME EXPOSURE STOP (HOUR) 3 DAY_CYCLE 24 , 4 BCK_TIME_ON 0. %DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 5 BCK_TIME_OFF 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 6 TIMELIMIT 24 %SIMULATION LIMIT TIME (HOUR) 7 BW T0 = 56.5 % Body weight at the 8 simulation (g) 9 10 %EXPOSURE DOSE SCENARIOS (UG/KG) 11 MSTOT = 0.003 % ORAL EXPOSURE DOSE (UG/KG) 12 %MSTOT = 0.01 OO ORAL EXPOSURE DOSE (UG/KG) 13 %MSTOT = 0.03 O% ORAL EXPOSURE DOSE (UG/KG) 14 %MSTOT = 0.1 O% ORAL EXPOSURE DOSE (UG/KG) 15 %MSTOT = 0.3 O% ORAL EXPOSURE DOSE (UG/KG) 16 %MSTOT = 1. O% ORAL EXPOSURE DOSE (UG/KG) 17 %MSTOT = 3. O% ORAL EXPOSURE DOSE (UG/KG) 18 %MSTOT = 10. O% ORAL EXPOSURE DOSE (UG/KG) 19 %MSTOT = 30. O% ORAL EXPOSURE DOSE (UG/KG) 20 21 2 2 C .2.3.2.16. Murray et a l (1979). 2 3 output @clear 2 4 prepare @clear 2 5 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG 26 2 7 % Murray et al 1979 2 8 %built and check in August 7 2009 2 9 %protocol: dietary exposure for 3 generations (assume 120 day exposure for 3 0 each) 31 %dose levels: 0.001 0.01, 0.1 ug/kg/d 3 2 %dose levels: 1, 10, 100 ng/kg/d 33 3 4 MAXT = 0.01 3 5 CINT = 0.1 3 6 EXP_TIME_ON = 0. %TIME AT WHICH EXPOSURE BEGINS (HOUR) 3 7 EXP_TIME_OFF = 2880 %TIME AT WHICH EXPOSURE ENDS (HOUR); 3 8 CORRESPONDS TO 120 DAYS OF EXPOSURE 3 9 DAY_CYCLE = 24. 4 0 BCK_TIME_ON = 0. %TIME AT WHICH BACKGROUND EXPOSURE BEGINS 41 (HOUR) 4 2 BCK_TIME_OFF = 0. %TIME AT WHICH BACKGROUND EXPOSURE ENDS 4 3 (HOUR) 4 4 TIMELIMIT = 2880 %SIMULATION TIME LIMIT (HOUR) 4 5 BW_T0 = 4.5 % BODY WEIGHT AT THE BEGINNING OF THE 4 6 SIMULATION (G) 47 4 8 %EXPOSURE DOSE SCENARIOS (UG/KG) 4 9 %MSTOT = 0.001 % ORAL EXPOSURE DOSE IN UG/KG 5 0 %MSTOT = 0.01 > ORAL EXPOSURE DOSE IN UG/KG 51 MSTOT = 0.1 % ORAL EXPOSURE DOSE IN UG/KG 52 53 This document is a draftfor review purposes only and does not constitute Agency policy. C-48 DRAFT--DO NOT CITE OR QUOTE 1 C .2.3.2.17. N T P (1982) (female) (chronic). 2 output @clear 3 prepare @clear 4 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 5 6 %NTP 1982 7 %built and check in August 7 2009 8 %protocol: twice weekly gavage for 104 weeks + 3 week observation period 9 %Rat Dioxin 3C June09 2clean.csl 10 %RAT_NON_GEST_ICF_F083109.CSL (now 09-11-09) 11 %dose levels: 0.005, 0.025, 0.25 ug/kg biweekly for 104 weeks + 3 week 12 observation period 13 %dose levels: 5, 25, 250 ng/kg biweekly for 104 weeks + 3 week observation 14 period 15 %dose levels equivalent to: 1.43 , 7.14, 71.4 ng/kg/day (adj) 16 17 MAXT = 0.01 18 CINT = 0.1 19 EXP TIME ON = 0. %delay before begin exposure (HOUR) 2 0 EXP TIME OFF = 17472 %TIME EXPOSURE STOP (HOUR) 21 DAY CYCLE = 84 2 2 BCK TIME ON = 0. %DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 2 3 BCK TIME OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 2 4 TIMELIMIT = 17472 %SIMULATION LIMIT TIME (HOUR) 2 5 BW T0 = 250 % Body weight at the beginning of the 2 6 simulation (g) 27 2 8 %EXPOSURE DOSE SCENARIOS (UG/KG) 29 3 0 %MSTOT = 0.005 % exposure dose ug/kg 31 %MSTOT = 0.025 3 2 MSTOT = 0.25 33 3 4 C .2.3.2.18. N TP (1982) (male) (chronic). 3 5 output @clear 3 6 prepare @clear 3 7 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 38 3 9 %NTP 1982 4 0 %built and check in august 7 2009 41 %protocol: twice weekly gavage for 104 weeks + 3 week observation period 4 2 %Rat Dioxin 3C June09 2clean.csl 4 3 %RAT_NON_GEST_ICF_F083109.CSL (now 09-11-09) 4 4 %dose levels: 0.005, 0.025, 0.25 ug/kg biweekly for 104 weeks + 3 week 4 5 observation period 4 6 %dose levels: 5, 25, 250 ng/kg biweekly for 104 weeks + 3 week observation 4 7 period 4 8 %dose levels equivalent to: 1.43, 7.14, 71.4 ng/kg/day (adj) 49 5 0 MAXT 0.01 51 CINT 0.1 5 2 EXP_TIME_ON 0. %delay before begin exposure (HOUR) 53 EXP_TIME_OFF 17472 %TIME EXPOSURE STOP (HOUR) 5 4 DAY CYCLE 84 This document is a draftfor review purposes only and does not constitute Agency policy. C-49 DRAFT--DO NOT CITE OR QUOTE 1 BCK_TIME_ON 0. %DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 2 BCK_TIME_OFF 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 3 TIMELIMIT 17472 %SIMULATION LIMIT TIME (HOUR) 4 BW_T0 350 % Body weight at the beginning of the 5 simulation (g) 6 7 %EXPOSURE DOSE SCENARIOS (UG/KG) 8 9 %MSTOT 0.005 % exposure dose ug/kg 10 %MSTOT 0.025 11 MSTOT 0.25 12 13 C .2.3.2.19. N T P (2006) 14 weeks. 14 output @clear 15 prepare @clear 16 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 17 18 % NTP 2006 19 %built and check in August 7 2009 2 0 %protocol: oral exposure for 14 weeks; SD rats 21 %Rat_Dioxin_3C June09_2clean.csl 2 2 %RAT_NON_GEST_ICF_F083109.CSL (now 09-11-09) 2 3 %dose levels: 0.003, 0.010, 0.022, 0.046 0.1 ug/kg 5 days/weeks for 14 weeks 2 4 %dose levels equivalent to: 3, 10, 22, 46 100 ng/kg 5 days/weeks for 14 weeks 2 5 %dose levels equivalent to: 2.14, 7.14, 15.7, 32.9 71.4 ng/kg 7 days/weeks for 2 6 14 weeks 27 2 8 MAXT = 0.01 2 9 CINT = 0.1 3 0 EXP TIME ON = 0. %delay before begin exposure (HOUR) 31 EXP TIME OFF = 2352 %TIME EXPOSURE STOP (HOUR) 3 2 DAY CYCLE = 24 33 WEEK PERIOD = 168 3 4 WEEK FINISH = 119 3 5 BCK TIME ON = 0. %DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 3 6 BCK TIME OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 3 7 TIMELIMIT = 2352 %SIMULATION LIMIT TIME (HOUR) 3 8 BW T0 = 215 % Body weight at the beginning of the simulation 3 9 (g) 40 41 %EXPOSURE DOSE SCENARIOS (UG/KG) 42 %MSTOT = 0.003 % exposure dose ug/kg 43 %MSTOT = 0.010 % exposure dose ug/kg 44 %MSTOT = 0.022 % exposure dose ug/kg 45 %MSTOT = 0.046 % exposure dose ug/kg 46 MSTOT = 0.1 % exposure dose ug/kg 47 4 8 C .2.3.2.20. N T P (2006) 31 weeks. 4 9 output @clear 5 0 prepare @clear 51 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 52 53 % NTP 2006 5 4 %built and check in August 7 2009 55 %protocol: oral exposure for 31 weeks; SD rats 5 6 %Rat Dioxin 3C June09 2clean.csl 5 7 %RAT_NON_GEST_ICF_F083109.CSL (now 09-11-09) 5 8 %dose levels: 0.003, 0.010, 0.022, 0.046 0.1 ug/kg 5 days/weeks for 31 weeks This document is a draftfor review purposes only and does not constitute Agency policy. C -50 D R A FT -- D O N O T C IT E O R Q U O T E 1 %dose levels equivalent to: 3, 10, 22, 46 100 ng/kg 5 days/weeks for 31 weeks 2 %dose levels equivalent to: 2.14, 7.14, 15.7, 32.9 71.4 ng/kg 7 days/weeks 3 for 31 weeks 4 5 MAXT = 0.01 6 CINT = 0.1 7 EXP_TIME_ON = 0. %delay before begin exposure (HOUR) 8 EXP_TIME_OFF = 5208 %TIME EXPOSURE STOP (HOUR) 9 DAY_CYCLE = 24 10 WEEK_PERIOD = 168 11 WEEK_FINISH = 119 12 BCK_TIME_ON = 0. %DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 13 BCK_TIME_OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 14 TIMELIMIT = 5208 %SIMULATION LIMIT TIME (HOUR) 15 BW T0 = 215 % Body weight at the beginning of the 16 simulation (g) 17 18 %EXPOSURE DOSE SCENARIOS (UG/KG) 19 %MSTOT = 0.003 exposure dose ug/kg 20 %MSTOT = 0.010 % exposure dose ug/kg 21 %MSTOT = 0.022 % exposure dose ug/kg 22 %MSTOT = 0.046 % exposure dose ug/kg 23 MSTOT = 0.1 % exposure dose ug/kg 24 2 5 C .2.3.2.21. N T P (2006) 53 weeks. 2 6 output @clear 2 7 prepare @clear 2 8 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 29 3 0 % NTP 2006 31 %built and check in August 7 2009 3 2 %protocol: oral exposure for 53 weeks; SD rats 33 %Rat Dioxin 3C June09 2clean.csl 3 4 %RAT_NON_GEST_ICF_F083109.CSL (now 09-11-09) 3 5 %dose levels: 0.003, 0.010, 0.022, 0.046 0.1 ug/kg 5 days/weeks for 53 weeks 3 6 %dose levels equivalent to: 3, 10, 22, 46 100 ng/kg 5 days/weeks for 53 weeks 3 7 %dose levels equivalent to: 2.14, 7.14, 15.7, 32.9 71.4 ng/kg 7 days/weeks 3 8 for 53 weeks 39 4 0 MAXT = 0.01 41 CINT = 0.1 4 2 EXP_TIME_ON = 0. %delay before begin exposure (HOUR) 43 EXP_TIME_OFF = 8904 %TIME EXPOSURE STOP (HOUR) 4 4 DAY_CYCLE = 24 45 WEEK_PERIOD = 168 4 6 WEEK_FINISH = 119 4 7 BCK_TIME_ON = 0. %DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 4 8 BCK_TIME_OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 4 9 TIMELIMIT = 8904 %SIMULATION LIMIT TIME (HOUR) 5 0 BW_T0 = 215 % Body weight at the beginning of the 51 simulation (g) 52 53 %EXPOSURE DOSE SCENARIOS (UG/KG) 54 %MSTOT = 0.003 % exposure dose ug/kg 55 %MSTOT = 0.010 % exposure dose ug/kg This document is a draftfor review purposes only and does not constitute Agency policy. C -51 D R A FT -- D O N O T C IT E O R Q U O T E 1 %MSTOT = 0.022 2 %MSTOT = 0.046 3 MSTOT = 0.1 4 5 C .2.3.2.22. N T P (2006) 2 year. % exposure dose ug/kg % exposure dose ug/kg % exposure dose ug/kg 6 output @clear 7 prepare @clear 8 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG 9 10 % NTP 2006 11 %built and check in August 7 2009 12 %protocol: oral exposure for 105 weeks; SD rats 13 %dose levels: 0.003, 0.010, 0.022, 0.046, 0.1 ug/kg 5 days/week for 105 14 weeks 15 %dose levels equivalent to: 3, 10, 22, 46, 100 ng/kg 5 days/week for 105 16 weeks 17 %dose levels equivalent to: 2.14, 7.14, 15.7, 32.9, 71.4 ng/kg 7 days/week 18 for 105 weeks 19 2 0 MAXT 0.01 21 CINT 0.1 2 2 EXP_TIME_ON 0. %TIME AT WHICH EXPOSURE BEGINS (HOUR) 23 EXP_TIME_OFF 17640 %TIME AT WHICH EXPOSURE ENDS (HOUR) 2 4 DAY_CYCLE 24 25 WEEK_PERIOD 168 2 6 WEEK_FINISH 119 2 7 BCK_TIME_ON 0. %TIME AT WHICH BACKGROUND EXPOSURE BEGINS 2 8 (HOUR) 2 9 BCK_TIME_OFF 0. %TIME AT WHICH BACKGROUND EXPOSURE ENDS (HOUR) 3 0 TIMELIMIT 17640 %SIMULATION TIME LIMIT (HOUR) 31 BW_T0 215 % BODY WEIGHT AT THE BEGINNING OF THE 3 2 SIMULATION (G) 33 3 4 %EXPOSURE DOSE SCENARIOS (UG/KG) 35 %MSTOT = 0.003 % EXPOSURE DOSE IN UG/KG 36 %MSTOT = 0.010 % EXPOSURE DOSE IN UG/KG 37 %MSTOT = 0.022 % EXPOSURE DOSE IN UG/KG 38 %MSTOT = 0.046 % EXPOSURE DOSE IN UG/KG 39 MSTOT = 0.1 % EXPOSURE DOSE IN UG/KG 40 41 C .2.3.2.23. Sewall et al. (1995). 4 2 output @clear 4 3 prepare @clear 4 4 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 4 5 % Sewall et al. 1995 4 6 %Rat Dioxin 3C June09 2clean.csl 4 7 %RAT_NON_GEST_ICF_F083109.CSL (now 09-11-09) 4 8 %protocol: gavage every 2 weeks for 30 weeks 4 9 %dose levels: 0.049, 0.1498, 0.49, and 1.75 ug/kg every 2 weeks 5 0 %dose levels: 3.5, 10.7, 35, and 125 ng/kg/d or 49, 149.8, 490, and 1750 51 ng/kg every 2 weeks 52 53 MAXT = 0.01 5 4 CINT = 0.1 This document is a draftfor review purposes only and does not constitute Agency policy. C-52 DRAFT--DO NOT CITE OR QUOTE 1 EXP TIME ON = 0. %delay before begin exposure (HOUR) 5 weeks 2 after start of experiment (age = 12 weeks) 3 EXP TIME OFF = 5040 %TIME EXPOSURE STOP (HOUR); 30 doses, 1 4 every two weeks 5 DAY_CYCLE = 336. % once every two weeks 6 BCK_TIME_ON = 0. %DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 7 BCK_TIME_OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 8 TIMELIMIT = 5040 %SIMULATION LIMIT TIME (HOUR) 9 BW_T0 = 250 % Body weight at the beginning of the 10 simulation (g); corresponds to 12 week old female 11 12 %EXPOSURE DOSE SCENARIOS (UG/KG) 13 %MSTOT = 0.049 % ORAL EXPOSURE DOSE (UG/KG) 14 %MSTOT = 0.1498 % ORAL EXPOSURE DOSE (UG/KG) 15 %MSTOT = 0.49 % ORAL EXPOSURE DOSE (UG/KG) 16 MSTOT = 1.75 % ORAL EXPOSURE DOSE (UG/KG) 17 18 C .2.3.2.24. Shi et al. (2007), adult portion. 19 output @clear 2 0 prepare @clear 21 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG 22 2 3 % Shi et al 2007 2 4 %built and check in August 7 2009 2 5 %protocol: gavage once per week for 322 days 2 6 %dose levels: 0.001, 0.005, 0.05 and 0.2 ug TCDD:kg body weight by gavage 2 7 once per week 2 8 %dose levels: 1, 5, 50 and 200 ng/kg ng TCDD:kg body weight by gavage once 2 9 per week 3 0 % dose equivalent adjusted 0.143, 0.714, 7.14 and 28.6 ng/kg/d 31 3 2 MAXT = 0.0001 33 CINT = 0.1 3 4 EXP TIME ON = 504. % TIME AT WHICH EXPOSURE BEGINS (HOUR) 3 5 EXP TIME OFF = 7728 %TIME AT WHICH EXPOSURE ENDS (HOUR); 3 6 CORRESPONDS TO 322 DAYS OF EXPOSURE 3 7 DAY CYCLE = 168. 3 8 BCK TIME ON = 0. % TIME AT WHICH BACKGROUND EXPOSURE 3 9 BEGINS (HOUR) 4 0 BCK TIME OFF = 0. % TIME AT WHICH BACKGROUND EXPOSURE ENDS 41 (HOUR) 4 2 TIMELIMIT = 7728 %SIMULATION TIME LIMIT (HOUR) LO 4 3 BW T0 = % BODY WEIGHT AT THE BEGINNING OF THE 4 4 SIMULATION (G) 45 4 6 %EXPOSURE DOSE SCENARIOS (UG/KG) 4 7 %MSTOT = 0.001 % ORAL EXPOSURE DOSE IN UG/KG 4 8 %MSTOT = 0.005 % ORAL EXPOSURE DOSE IN UG/KG 4 9 %MSTOT = 0.05 % ORAL EXPOSURE DOSE IN UG/KG 5 0 MSTOT = 0.2 % ORAL EXPOSURE DOSE IN UG/KG 51 5 2 C .2.3.2.25. Van Birgelen et al. (1995). 53 output @clear 5 4 prepare @clear This document is a draftfor review purposes only and does not constitute Agency policy. C-53 DRAFT--DO NOT CITE OR QUOTE 1 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG 2 3 % Van Birgelen et al. (1995) 4 %protocol: daily dietary exposure for 13 weeks 5 %dose levels: 0.0135, 0.0264, 0.0469, 0.320, 1.024 ug/kg every day for 13 6 weeks 7 % dose levels 13.5, 26.4, 46.9, 320, 1024 ng/kg every day for 13 weeks 8 MAXT = 0.01 9 CINT = 0.1 10 EXP_TIME_ON = 0. %delay before begin exposure (HOUR) 11 EXP_TIME_OFF = 2184. %TIME EXPOSURE STOP (HOUR) 12 DAY_CYCLE = 24. % once every two weeks 13 BCK_TIME_ON = 0. %DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 14 BCK_TIME_OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 15 TIMELIMIT = 2184. %SIMULATION LIMIT TIME (HOUR) 16 BW_T0 = 150. % Body weight at the beginning of the 17 simulation (g) 18 19 %EXPOSURE DOSE SCENARIOS (UG/KG) 2 0 %MSTOT = 0.0135 % ORAL EXPOSURE DOSE (UG/KG) 21 %MSTOT = 0.0264 % ORAL EXPOSURE DOSE (UG/KG) 2 2 %MSTOT = 0.0469 % ORAL EXPOSURE DOSE (UG/KG) 2 3 %MSTOT = 0.320 % ORAL EXPOSURE DOSE (UG/KG) 2 4 MSTOT = 1.024 % ORAL EXPOSURE DOSE (UG/KG) 25 2 6 C .2.3.2.26. Vanden Heuvel et al. (1994). 2 7 output @clear 2 8 prepare @clear 2 9 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 30 31 % Vanden Heuvel et al. 1994. 3 2 %built and check in August 7 2009 33 %protocol: single gavage 3 4 %Rat Dioxin 3C June09 2clean.csl 3 5 %RAT_NON GEST ICF F083109.CSL (now 09-11-09) 3 6 %dose levels:0. 00005, 0.0001, 0. 001, 0.010, 0.1, 1, 10 ug/kg/d 3 7 %dose levels equivalent to: 0.05 , 0.1, 1, 10, 100, 1000, 10000 ng/kg/d 38 3 9 MAXT = 0.001 4 0 CINT = 0.1 41 EXP TIME ON = 0. %delay before begin exposure (HOUR) 4 2 EXP TIME OFF = 24 %TIME EXPOSURE STOP (HOUR) 4 3 DAY CYCLE = 24 4 4 BCK TIME ON = 0. %DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 4 5 BCK TIME OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 4 6 TIMELIMIT = 24 %SIMULATION LIMIT TIME (HOUR) 4 7 BW T0 = 250 % Body weight at the beginning of the 4 8 simulation (g) 49 5 0 %EXPOSURE DOSE SCENARIOS (UG/KG) 51 5 2 %MSTOT = 0.00005 % exposure dose ug/kg 53 %MSTOT = 0.0001 % exposure dose ug/kg 5 4 %MSTOT = 0.001 % exposure dose ug/kg 55 %MSTOT = 0.01 % exposure dose ug/kg This document is a draftfor review purposes only and does not constitute Agency policy. C-54 DRAFT--DO NOT CITE OR QUOTE 1 %MSTOT 2 %MSTOT 3 MSTOT = 0.1 =1 = 10 4 5 C.2.4. Rat G estational M odel 6 C.2.4.1. Model Code % exposure dose ug/kg % exposure dose ug/kg % exposure dose ug/kg 7 P R O G R A M : 'T h re e C o m p a rtm e n t P B P K M o d e l fo r T C D D in R a t (G e statio n )' 8 9 ! Parameters were change May 16, 2002 10 ! Come from {8MAI_CHR_PRE-EXP_GD} 11 ! Come from {12 Mouse GD}file 12 !******************************************** 13 !{{IMPORTANT-IMPORTANT-IMPORTANT-IMPORTANT}} 14 ! REDUCTION OF MOTHER AND FETUS COMPARTMENT 15 ! 2M_R_TCDD_JULY2002 ////(JULY 18,2002)//// 16 !TCDD_RED_4Species_2003_4 ////(APR 8 ,2003)//// 17 !TCDD_RED_4Species_2003_9 ////(APR 17 ,2003)//// 18 !TCDD_RED_4Species_2003_12 ////(APR 17 ,2003)//// 19 !***************************************************** 2 0 !APRIL 18 2003 21 !TCDD_4C_4SP_2003 ////(APR 18 ,2003)//// 2 2 ! was ''Gest 4 species 1.csl'' but update July 2009 23 2 4 !DevTCDD4Species ICF afterKKfix v3 ratgest.csl 2 5 !RAT_GESTATIONAL_ICF_F083109.csl _ 2 6 !RAT_GESTATIONAL_ICF_F100609.csl 27 28 2 9 !Legend/Legend/Legend/Legend/Legend/Legend/Legend/Legend/ 3 0 !Legend for this PBPK model 31 !Mating: control the tenure of exchange between fetus and 3 2 !Mother and also control imitated tissue growth 33 !Control: WTFE, WFO, WPLA0, QPLAF,WT0 3 4 !(for rat, mouse, human, and monkey) 3 5 !Control transfer from mother to fetus or fetus to mother by TRANSTIME ON 3 6 !SWITCH_trans = 0 NO TRANSFER _ 3 7 !SWITCH_trans = 1 TRANSFER OCCURS 3 8 !Gest off = 1 3 9 !Gest on= 0.0 4 0 ! These switches are also controlled by mating parameters 41 4 2 INITIAL ! 43 4 4 ISIMULATION PARAMETERS ==== 4 5 CONSTANT PARA ZERO = 1E 4 6 CONSTANT EXP TIME ON = 0.0 ! TIME AT WHICH EXPOSURE BEGINS (HOURS) 4 7 CONSTANT EXP TIME OFF = 530 ! TIME AT WHICH EXPOSURE ENDS (HOURS) 4 8 CONSTANT DAY CYCLE = 24.0 ! NUMBER OF HOURS BETWEEN DOSES (HOURS) 4 9 CONSTANT BCK TIME ON = 0.0 ! TIME AT WHICH BACKGROUND EXPOSURE 5 0 BEGINS (HOURS) 51 CONSTANT BCK TIME OFF = 0.0 ! TIME AT WHICH BACKGROUND EXPOSURE ENDS 5 2 (HOURS) 53 CONSTANT TRANSTIME ON = 144.0 !CONTROL TRANSFER FROM MOTHER TO FETUS 5 4 AT GESTATIONAL DAY 6 This document is a draftfor review purposes only and does not constitute Agency policy. C-55 DRAFT--DO NOT CITE OR QUOTE 1 2 !UNIT CONVERSION 3 CONSTANT MW=322 ! MOLECULAR WEIGHT (NG/NMOL) 4 CONSTANT SERBLO = 0.55 5 CONSTANT UNITCORR = 1000 6 7 8 !INTRAVENOUS SEQUENCE 9 constant IV_LACK = 0.0 10 constant IV PERIOD = 0.0 11 12 IPREGNANCY PARAMETER ==== 13 CONSTANT MATTING = 0.0 IBEGINNING OF MATING (HOUR) 14 CONSTANT N FETUS = 10.0 INUMBER OF FETUS PRESENT 15 16 !CONSTANT EXPOSURE CONTROL =========== 17 !ACUTE, SUBCHRONIC, CHRONIC EXPOSURE ===== 18 !OR BACKGROUND EXPOSURE (IN THIS CASE 3 TIMES A DAY)=== 19 CONSTANT MSTOTBCKGR = 0.0 ! ORAL BACKGROUND EXPOSURE DOSE (UG/KG) 2 0 CONSTANT MSTOT = 0.0 ! ORAL EXPOSURE DOSE (UG/KG) 21 2 2 !ORAL ABSORPTION 23 MSTOT NM = MSTOT/MW CONVERTS THE DOSE TO NMOL/G 24 2 5 !INTRAVENOUS ABSORPTION 2 6 CONSTANT DOSEIV = 0.0 I INJECTED DOSE (UG/KG) 2 7 DOSEIV_NM = DOSEIV/MW I CONVERTS THE INJECTED DOSE TO NMOL/G 2 8 CONSTANT DOSEIVLATE = 0.0 I INJECTED DOSE LATE (UG/KG) 2 9 DOSEIVNMlate = DOSEIVLATE/MW IAMOUNT IN NMOL/G 30 31 IINITIAL GUESS OF THE FREE CONCENTRATION IN THE LIGAND (COMPARTMENT 3 2 INDICATED BELOW)==== 33 CONSTANT CFLLI0 = 0.0 ILIVER (NMOL/ML) 3 4 CONSTANT CFLPLA0 = 0.0 IPLACENTA (NMOL/ML) 35 3 6 !BINDING CAPACITY (AhR) FOR NON LINEAR BINDING (COMPARTMENT INDICATED 3 7 BELOW) (NMOL/ML) === 3 8 CONSTANT LIBMAX = 3.5E-4 I LIVER (NMOL/ML), WANG ET AL. 1997 3 9 CONSTANT PLABMAX = 2.0E-4 ITEMPORARY PARAMETER 40 41 I PROTEIN AFFINITY CONSTANTS (1A2 OR AhR, COMPARTMENT INDICATED BELOW) 4 2 (NMOL/ML)=== 4 3 CONSTANT KDLI = 1.0E-4 ILIVER (AhR) (NMOL/ML), WANG ET AL. 1997 4 4 CONSTANT KDLI2 = 4.0E-2 ILIVER (1A2) (NMOL/ML), EMOND ET AL. 2004 4 5 CONSTANT KDPLA = 1.0E-4 ITEMPORARY PARAMETER; ASSUME IDENTICAL TO 4 6 KDLI (AhR) 47 4 8 IEXCRETION AND ABSORPTION CONSTANT 4 9 CONSTANT KST = 0.36 I GASTRIC RATE CONSTANT (HR-1), WANG ET 5 0 AL. 1997 51 CONSTANT KABS = 0.48 IINTESTINAL ABSORPTION CONSTANT (HR-1) ), 5 2 WANG ET AL. 1997I 53 5 4 I ELIMINATION CONSTANTS 55 CONSTANT CLURI = 0.01 I URINARY CLEARANCE (ML/HR), EMOND ET 5 6 AL. 2004 This document is a draftfor review purposes only and does not constitute Agency policy. C-56 DRAFT--DO NOT CITE OR QUOTE 1 2 IINTERSPECIES ELIMINATION VARIABLE 3 CONSTANT kelv = 0.15 I INTERSPECIES VARIABLE ELIMINATION 4 CONSTANT (1/HOUR) 5 6 I CONSTANT TO DIVIDE THE ABSORPTION INTO LYMPHATIC AND PORTAL FRACTIONS 7 CONSTANT A = 0.7 ! LYMPHATIC FRACTION, WANG ET AL. 1997 8 9 IPARTITION COEFFICIENTS 10 CONSTANT PF = 100 ! ADIPOSE TISSUE/BLOOD, WANG ET AL. 1997 11 CONSTANT PRE = 1.5 ! REST OF THE BODY/BLOOD, WANG ET AL. 12 1997 13 CONSTANT PLI = 6.0 ! LIVER/BLOOD, WANG ET AL. 1997 14 CONSTANT PPLA = 1.5 ! TEMPORARY PARAMETER NOT CONFIGURED, 15 WANG ET AL. 1997 16 17 IPARAMETER FOR INDUCTION OF CYP 1A2, WANG ET AL. 1997 18 CONSTANT PAS INDUC = 1.0 ! INCLUDE INDUCTION? (1 = YES, 0 = NO) 19 CONSTANT CYP1A2 1OUTZ = 1.6 ! DEGRADATION CONCENTRATION CONSTANT OF 2 0 1A2 (NMOL/ML) 21 CONSTANT CYP1A2 1A1 = 1.6 ! BASAL CONCENTRATION OF 1A1 (NMOL/ML) 2 2 CONSTANT CYP1A2' 1EC50 = 0.13 ! DISSOCIATION CONSTANT TCDD-CYP1A2 2 3 (NMOL/ML) 2 4 CONSTANT CYP1A2 1A2 = 1.6 !BASAL CONCENTRATION OF 1A2 (NMOL/ML) 2 5 CONSTANT CYP1A2' 1KOUT = 0.1 ! FIRST ORDER RATE OF DEGRADATION (H-1) 2 6 CONSTANT CYP1A2 1TAU = 0.25 !HOLDING TIME (H) 2 7 CONSTANT CYP1A2' 1EMAX = 600 ! MAXIMUM INDUCTION OVER BASAL EFFECT 2 8 (UNITLESS) 2 9 CONSTANT HILL = 0.6 !HILL CONSTANT; COOPERATIVELY LIGAND 3 0 BINDING EFFECT CONSTANT (UNITLESS) 31 3 2 !DIFFUSIONAL PERMEABILITY FRACTION 33 CONSTANT PAFF = 0.0910 !ADIPOSE (UNITLESS), WANG ET AL. 1997 3 4 CONSTANT PAREF = 0.0298 !REST OF THE BODY (UNITLESS), WANG ET 3 5 AL. 1997 3 6 CONSTANT PALIF = 0.3500 !LIVER (UNITLESS), WANG ET AL. 1997 3 7 CONSTANT PAPLAF = 0.3 !TEMPORARY PARAMETER NOT CONFIGURED 38 3 9 IFRACTION OF TISSUE WEIGHT ==== 4 0 CONSTANT WLI0 = 0.0360 ILIVER, WANG ET AL. 1997 41 4 2 !TISSUE BLOOD FLOW EXPRESSED AS A FRACTION OF CARDIAC OUTPUT 43 CONSTANT QFF = 0.069 ! ADIPOSE TISSUE BLOOD FLOW FRACTION 4 4 (UNITLESS), WANG ET AL. 1997 4 5 CONSTANT QLIF = 0.183 !LIVER (UNITLESS), WANG ET AL. 1997 46 4 7 !COMPARTMENT TISSUE BLOOD EXPRESSED AS A FRACTION OF THE TOTAL COMPARTMENT 4 8 VOLUME 4 9 CONSTANT WFB0 0.050 !ADIPOSE TISSUE, WANG ET AL. 1997 5 0 CONSTANT WREB0 0.030 !REST OF THE BODY, WANG ET AL. 1997 51 CONSTANT WLIB0 0.266 !LIVER, WANG ET AL. 1997 5 2 CONSTANT WPLAB0 0.500 !TEMPORARY PARAMETER NOT CONFIGURED 53 5 4 !EXPOSURE SCENARIO FOR UNIQUE OR REPETITIVE WEEKLY OR MONTHLY EXPOSURE 55 !NUMBER OF EXPOSURES PER WEEK 5 6 CONSTANT WEEK_LACK = 0.0 !DELAY BEFORE EXPOSURE ENDS (WEEK) This document is a draftfor review purposes only and does not constitute Agency policy. C-57 DRAFT--DO NOT CITE OR QUOTE 1 CONSTANT WEEK_PERIOD = 168 ! NUMBER OF HOURS IN THE WEEK (HOURS) 2 CONSTANT WEEK_FINISH = 168 ! TIME EXPOSURE ENDS (HOURS) 3 4 INUMBER OF EXPOSURES PER MONTH 5 CONSTANT MONTH_LACK = 0.0 !DELAY BEFORE EXPOSURE BEGINS (MONTHS) 6 7 ICONSTANT FOR BACKGROUND EXPOSURE 8 CONSTANT Day_LACK_BG = 0.0 !DELAY BEFORE EXPOSURE BEGINS (HOURS) 9 CONSTANT Day_PERIOD_BG = 24 !LENGTH OF EXPOSURE (HOURS) 10 11 INUMBER OF EXPOSURES PER WEEK 12 CONSTANT WEEK_LACK_BG = 0.0 !DELAY BEFORE BACKGROUD EXPOSURE BEGINS 13 (WEEKS) 14 CONSTANT WEEK_PERIOD_BG = 168 !NUMBER OF HOURS IN THE WEEK (HOURS) 15 CONSTANT WEEK_FINISH_BG = 168 !TIME EXPOSURE ENDS (HOURS) 16 17 IINITIAL BODY WEIGHT 18 CONSTANT BW_T0 = 250 ! WANG ET AL. 1997 19 CONSTANT RATIO_RATF_MOUSEF = 1.0 !RATIO OF FETUS MOUSE/RAT AT 2 0 GESTATIONAL DAY 22 21 2 2 ! COMPARTMENT LIPID EXPRESSED AS THE FRACTION OF TOTAL LIPID, POULIN ET AL 2 3 2000 2 4 CONSTANT F_TOTLIP = 0.855 ! ADIPOSE TISSUE (UNITLESS) 2 5 CONSTANT B_TOTLIP = 0.0023 ! BLOOD (UNITLESS) 2 6 CONSTANT RE_TOTLIP = 0.019 ! REST OF THE BODY 2 7 (UNITLESS) 2 8 CONSTANT LI_TOTLIP 060 ! LIVER (UNITLESS) 2 9 CONSTANT PLA_TOTLIP 019 3 0 CONSTANT FETUS TOTLIP 019 31 3 2 END ! END OF THE INITIAL SECTION 33 3 4 DYNAMIC ! DYNAMIC SIMULATION SECTION 3 5 ALGORITHM IALG = 2 ! GEAR METHOD 3 6 CINTERVAL CINT = 0.1 ! COMMUNICATION INTERVAL 3 7 MAXTERVAL MAXT = 1.0e+10 ! MAXIMUM CALCULATION INTERVAL 3 8 MINTERVAL MINT = 1.0E-10 ! MINIMUM CALCULATION INTERVAL 3 9 VARIABLE T = 0.0 4 0 CONSTANT TIMELIMIT = 100 !SIMULATION LIMIT TIME (HOURS) 41 CINTXY = CINT 4 2 PFUNC = CINT 43 4 4 !TIME CONVERSION 4 5 DAY = T/24 TIME IN DAYS 4 6 WEEK = T/168 TIME IN WEEKS 4 7 MONTH = T/730 TIME IN MONTHS 4 8 YEAR = T/8760 TIME IN YEARS 49 5 0 DERIVATIVE PORTION OF CODE THAT SOLVES DIFFERENTIAL EQUATIONS 51 5 2 !CHRONIC OR SUBCHRONIC EXPOSURE SCENARIO ======= 53 !NUMBER OF EXPOSURES PER DAY 5 4 DAY_LACK = EXP_TIME_ON ! DELAY BEFORE EXPOSURE BEGINS (HOURS) 55 DAY_PERIOD = DAY_CYCLE ! EXPOSURE PERIOD (HOURS) 5 6 DAY FINISH = CINTXY ! LENGTH OF EXPOSURE (HOURS) This document is a draftfor review purposes only and does not constitute Agency policy. C-58 DRAFT--DO NOT CITE OR QUOTE 1 MONTH_PERIOD = TIMELIMIT ! EXPOSURE PERIOD (MONTHS) 2 MONTH FINISH = EXP TIME OFF ! LENGTH OF EXPOSURE (MONTHS) 3 4 INUMBER OF EXPOSURES PER DAY AND MONTH 5 DAY_FINISH_BG = CINTXY 6 MONTH_LACK_BG = BCK_TIME_ON IDELAY BEFORE BACKGROUD EXPOSURE BEGINS 7 (MONTHS) 8 MONTH_PERIOD_BG = TIMELIMIT IBACKGROUND EXPOSURE (MONTHS) 9 MONTH FINISH BG = BCK TIME OFF ILENGTH OF BACKGROUND EXPOSURE (MONTHS) 10 11 IINTRAVENOUS LATE 12 IV_FINISH = CINTXY 13 B = 1-A I FRACTION OF DIOXIN ABSORBED IN THE PORTAL FRACTION OF THE LIVER 14 15 16 !FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUM 17 E 18 ! FROM OFLAHERTY_1992 19 2 0 RTESTGEST= T-MATTING 21 TESTGEST=DIM(RTESTGEST,0.0) 22 23 WTFER_RODENT= (2.3d-3*EXP(1.49d-2*(TESTGEST))+1.3d-2)*Gest_on 2 4 WTFER = (WTFER_RODENT*RATIO_RATF_MOUSEF*N_FETUS) 2 5 WTFE = DIM(WTFER,0.0) 26 27 ! 2 8 FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME 2 9 ! FAT GROWTH EXPRESSION LINEAR DURING PREGNANCY 3 0 ! FROM O'FLAHERTY_1992 31 3 2 WF0= (((9.66d-5*(TESTGEST))*gest_on)+0.069) 33 3 4 ! PLACENTA,VOLUME, PLACENTA,VOLUME, PLACENTA,VOLUME, PLACENTA,VOLUME 3 5 ! WPLA PLACENTA GROWTH EXPRESSION, SINGLE EXPONENTIAL WITH OFFSET 3 6 ! FROM O'FLAHERTY_1992 ! FOR EACH PUP 37 3 8 WPLA0N_RODENT = (0.6/(1+(5d+3*EXP(-0.0225*(TESTGEST)))))*N_FETUS 3 9 WPLA0R = (WPLA0N_RODENT/WT0)*Gest_on 4 0 WPLA0 = DIM(WPLA0R,0.0) 41 4 2 ! PLACENTA,FLOW RATE, PLACENTA,FLOW RATE, PLACENTA,FLOW RATE, PLACENTA,FLOW 4 3 RATE 4 4 ! QPLA PLACENTA GROWTH EXPRESSION, DOUBLE EXPONENTIAL WITH OFFSET 4 5 ! FROM O'FLAHERTY_1992 46 4 7 QPLARF = (1.67d-7 *exp(9.6d-3*(TESTGEST)) & 4 8 +1.6d-3*exp(7.9d-3*(TESTGEST))+0.0)*Gest_on*SWITCH_trans 4 9 QPLAF=DIM(QPLARF,0.0) !FRACTION OF FLOW RATE IN PLACENTA 50 51 ! GESTATION CONTROL 5 2 IF (T.LT.MATTING) THEN 53 Gest off = 1.0 5 4 Gest on= 0.0 55 ELSE 5 6 Gest off = 0.0 This document is a draftfor review purposes only and does not constitute Agency policy. C-59 DRAFT--DO NOT CITE OR QUOTE 1 Gest on = 1.0 2 END IF _ 3 4 MOTHER BODY WEIGHT GROWTH EQUATION======== 5 MODIFICATION TO ADAPT THIS MODEL AT HUMAN MODEL 6 BECAUSE LINEAR DESCRIPTION IS NOT GOOD ENOUGH FOR MOTHER GROWTH 7 MOTHER BODY WEIGHT GROWTH 8 9 PARAMETER (BW_RMN = 1.0E-30) 10 WT0= BW_T0 *(1+(0.41*T)/(1402.5+T+BW_RMN)) 11 12 ! VARIABILITY OF REST OF THE BODY DEPENDS ON OTHER ORGANS 13 WRE0 = (0.91 - (WLIB0*WLI0 + WFB0*WF0 +WPLAB0*WPLA0 + WLI0 + WF0 + 14 WPLA0))/(1+WREB0) ! REST OF THE BODY FRACTION; UPDATED FOR EPA ASSESSMENT 15 QREF = 1-(QFF+QLIF+QPLAF) !REST OF BODY BLOOD FLOW RATE (ML/HR) 16 QTTQF = QFF+QREF+QLIF+QPLAF ! SUM MUST EQUAL 1 17 18 ! COMPARTMENT VOLUME (ML OR G) 19 WF = WF0 * WT0 ADIPOSE TISSUE 2 0 WRE = WRE0 * WT0 REST OF THE BODY 21 WLI = WLI0 * WT0 LIVER 2 2 WPLA= WPLA0* WT0 PLACENTA 23 2 4 ! COMPARTMENT TISSUE BLOOD (ML OR G) 2 5 WFB = WFB0 * WF ! ADIPOSE TISSUE 2 6 WREB = WREB0 * WRE ! REST OF THE BODY 2 7 WLIB = WLIB0 * WLI ! LIVER 2 8 WPLAB = WPLAB0* WPLA ! PLACANTA 29 3 0 ! CARDIAC OUTPUT FOR THE GIVEN BODY WEIGHT (ML/H) ========= 31 !QC= QCCAR*60*(WT0/1000.0)**0.75 3 2 CONSTANT QCC=18684.0 ! EQUIVALENT TO 311.4 * 60 33 QC= QCC*(WT0/UNITCORR)**0.75 34 3 5 !COMPARTMENT BLOOD FLOW RATE (ML/HR) 3 6 QF = QFF*QC !ADIPOSE TISSUE BLOOD FLOW RATE 3 7 QLI = QLIF*QC !LIVER TISSUE BLOOD FLOW RATE 3 8 QRE = QREF*QC !REST OF THE BODY BLOOD FLOW RATE 3 9 QPLA = QPLAF*QC !PLACENTA TISSUE BLOOD FLOW RATE 4 0 QTTQ = QF+QRE+QLI+QPLA !TOTAL FLOW RATE 41 4 2 !PERMEABILITY ORGAN FLOW (ML/HR) : 4 3 PAF = PAFF*QF ADIPOSE TISSUE 4 4 PARE = PAREF*QRE REST OF THE BODY 4 5 PALI = PALIF*QLI LIVER TISSUE 4 6 PAPLA = PAPLAF*QPLA PLACENTA 47 4 8 ************************************** 4 9 ABSORPTION SECTION 5 0 ORAL 51 INTRAPERITONEAL 5 2 INTRAVENOUS 53 54 55 !REPETITIVE ORAL BACKGROUND EXPOSURE SCENARIO 56 This document is a draftfor review purposes only and does not constitute Agency policy. C-60 DRAFT--DO NOT CITE OR QUOTE 1 MSTOT_NMBCKGR = MSTOTBCKGR/MW ! CONVERTS THE BACKGROUND DOSE TO NMOL/G 2 MSTTBCKGR =MSTOT_NMBCKGR *WT0 3 4 DAY_EXPOSURE_BG = PULSE(DAY_LACK_BG,DAY_PERIOD_BG,DAY_FINISH_BG) 5 WEEK_EXPOSURE_BG = PULSE(WEEK_LACK_BG,WEEK_PERIOD_BG,WEEK_FINISH_BG) 6 MONTH_EXPOSURE_BG = PULSE(MONTH_LACK_BG,MONTH_PERIOD_BG,MONTH_FINISH_BG) 7 8 MSTTCH_BG = (DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG)*MSTTBCKGR 9 MSTTFR_BG = MSTTBCKGR/CINT 10 11 CYCLE_BG =DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG 12 13 ! CONDITIONAL ORAL EXPOSURE (BACKGROUND EXPOSURE) 14 15 IF (MSTTCH_BG.EQ.MSTTBCKGR) THEN 16 ABSMSTT_GB= MSTTFR_BG 17 ELSE 18 ABSMSTT_GB = 0.0 19 END IF 20 21 CYCLETOTBG=INTEG(CYCLE_BG,0.0) 22 2 3 !REPETITIVE ORAL EXPOSURE SCENARIO 24 2 5 MSTT= MSTOT_NM * WT0 !AMOUNT IN NMOL 26 2 7 DAY_EXPOSURE = PULSE(DAY_LACK,DAY_PERIOD,DAY_FINISH) 2 8 WEEK_EXPOSURE = PULSE(WEEK_LACK,WEEK_PERIOD,WEEK_FINISH) 2 9 MONTH_EXPOSURE = PULSE(MONTH_LACK,MONTH_PERIOD,MONTH_FINISH) 30 31 MSTTCH = (DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE)*MSTT 3 2 MSTTFR = MSTT/CINT 33 3 4 CYCLE = DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE 3 5 SUMEXPEVENT= INTEG (CYCLE,0.0) !NUMBER OF CYCLE GENERATE DURING SIMULATION 36 3 7 ! CONDITIONAL ORAL EXPOSURE 3 8 IF (MSTTCH.EQ.MSTT) THEN 3 9 ABSMSTT= MSTTFR 4 0 ELSE 41 ABSMSTT = 0.0 4 2 END IF 43 44 4 5 CYCLETOT=INTEG(CYCLE,0.0) 46 4 7 ! MASS CHANGE IN THE LUMEN 4 8 RMSTT= -(KST+KABS)*MST +ABSMSTT +ABSMSTT_GB ! RATE OF CHANGE (NMOL/H) 4 9 MST = INTEG(RMSTT,0.0) IAMOUNT REMAINING IN DUODENUM 5 0 (NMOL) 51 5 2 ! ABSORPTION IN LYMPH CIRCULATION 53 LYRMLUM = KABS*MST*A 5 4 LYMLUM = INTEG(LYRMLUM,0.0) 55 5 6 ! ABSORPTION IN PORTAL CIRCULATION This document is a draftfor review purposes only and does not constitute Agency policy. C-61 DRAFT--DO NOT CITE OR QUOTE 1 LIRMLUM = KABS*MST*B 2 LIMLUM = INTEG(LIRMLUM,0.0) 3 4 5 ! ---- IV EXPOSURE --------- 6 7 IV= DOSEIV_NM * WT0 !AMOUNT IN NMOL 8 IVR= IV/PFUNC ! RATE FOR IV INFUSION IN BLOOD 9 EXPIV= IVR * (1.0-STEP(PFUNC)) 10 IVDOSE = integ(EXPIV,0.0) 11 12 !------IV LATE IN THE CYCLE 13 ! MODIFICATION ON January 13 2004 14 IV_RlateR = DOSEIVNMlate*WT0 15 IV_EXPOSURE=PULSE(IV_LACK,IV_PERIOD,IV_FINISH) 16 17 IV_lateT = IV_EXPOSURE *IV_RlateR 18 IV_late = IV_lateT/CINT 19 2 0 SUMEXPEVENTIV= integ (IV_EXPOSURE,0.0) !NUMBER OF CYCLE GENERATE DURING 21 SIMULATION 22 2 3 !SYSTEMIC CONCENTRATION OF TCDD 24 2 5 ! MODIFICATION ON OCTOBER 6, 2009 2 6 CB= (QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM+QPLA*CPLAB+IV_late)/(QC+CLURI) ! 2 7 CA = CB ! CONCENTRATION (NMOL/ML) 28 29 3 0 !URINARY EXCRETION BY KIDNEY 31 ! MODIFICATION ON OCTOBER 6, 2009 3 2 RAURI = CLURI *CB 33 AURI = INTEG(RAURI,0.0) 34 35 36 3 7 !UNIT CONVERSION POST SIMULATION 3 8 CBSNGKGLIADJ=(CB*MW*UNITCORR*(1.0/B_TOTLIP)*(1.0/SERBLO))![NG of TCDD 3 9 Serum/Kg OF LIPIP] 4 0 AUCBS_NGKGLIADJ=integ(CBSNGKGLIADJ,0.0) 41 4 2 PRCT_B = (CB/(MSTT+1E-30))*100.0 !PERCENT OF ORAL DOSE IN BLOOD 4 3 PRCT_BIV = (CB/(IV_RlateR+1E-30))*100.0 ! PERCENT OF IV DOSE IN BLOOD 4 4 CBNGKG= CB*MW*UNITCORR 45 46 4 7 !ADIPOSE COMPARTMENT 4 8 !TISSUE BLOOD COMPARTMENT 4 9 RAFB= QF*(CA-CFB)-PAF*(CFB-CF/PF) !(NMOL/H) 5 0 AFB = INTEG(RAFB,0.0) !(NMOL) 51 CFB = AFB/WFB !(NMOL/ML) 5 2 !TISSUE COMPARTMENT 53 RAF = PAF*(CFB-CF/PF) !(NMOL/H) 5 4 AF = INTEG(RAF,0.0) !(NMOL) 55 CF = AF/WF !(NM/ML) 56 This document is a draftfor review purposes only and does not constitute Agency policy. C-62 DRAFT--DO NOT CITE OR QUOTE 1 !UNIT CONVERSION POST SIMULATION 2 CFTOTAL= (AF + AFB)/(WF + WFB) ! TOTAL CONCENTRATION IN NMOL/ML 3 CFTFREE = CFB + CF !TOTAL FREE CONCENTRATION IN FAT (NM/ML) 4 PRCT_F = (CFTOTAL/(MSTT+1E-30))*100.0 ! PERCENT OF ORAL DOSE IN FAT 5 PRCT_FIV = (CFTOTAL/(IV_RlateR+1E-30))*100.0 ! PERCENT OF IV DOSE IN FAT 6 CFNGKG=CFTOTAL*MW*UNITCORR ! FAT CONCENTRATION NG/KG 7 AUCF_NGKGH=integ(CFNGKG,0.0) 8 9 !REST OF THE BODY COMPARTMENT 10 RAREB= QRE *(CA-CREB)-PARE*(CREB-CRE/PRE) !(NMOL/H) 11 AREB = INTEG(RAREB,0.0) !(NMOL) 12 CREB = AREB/WREB !(NMOL/H) 13 !TISSUE COMPARTMENT 14 RARE = PARE*(CREB - CRE/PRE) !(NMOL/H) 15 ARE = INTEG(RARE,0.0) !(NMOL) 16 CRE = ARE/WRE !(NMOL/ML) 17 18 !UNIT CONVERSION POST SIMULATION 19 CRETOTAL= (ARE + AREB)/(WRE + WREB) ! TOTAL CONCENTRATION IN 2 0 NMOL/ML 21 PRCT_RE = (CRETOTAL/(MSTT+1E-30))*100.0 ! PERCENT OF ORAL DOSE IN REST OF 2 2 THE BODY 2 3 PRCT_REIV = (CRETOTAL/(IV_RlateR+1E-30))*100.0 !PERCENT OF IV DOSE IN 2 4 REST OF THE BODY 2 5 CRENGKG=CRETOTAL*MW*UNITCORR ! REST OF THE BODY CONCENTRATION IN NG/KG 26 27 2 8 !LIVER COMPARTMENT 2 9 !TISSUE BLOOD COMPARTMENT 3 0 RALIB = QLI*(CA-CLIB)-PALI*(CLIB-CFLLIR)+LIRMLUM ! 31 ALIB = INTEG(RALIB,0.0) !(NMOL) 3 2 CLIB = ALIB/WLIB !(NMOL/ML) 33 !TISSUE COMPARTMENT 3 4 RALI = PALI*(CLIB - CFLLIR)-REXCLI ! (NMOL/HR) 3 5 ALI = INTEG(RALI,0.0) !(NMOL) 3 6 CLI = ALI/WLI !(NMOL/ML) 37 3 8 !FREE TCDD CONCENTRATION IN LIVER COMPARTMENT 3 9 PARAMETER (LIVER_1RMN = 1.0E-30) 4 0 CFLLI= IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLIR & 41 +LIVER_1RMN))+((CYP1A2_1O3*CFLLIR/(KDLI2 + CFLLIR & 4 2 +LIVER_1RMN)*PAS_INDUC)))-CFLLI,CFLLI0) 4 3 CFLLIR=DIM(CFLLI,0.0) ! FREE CONCENTRATION IN LIVER 44 4 5 CBNDLI= LIBMAX*CFLLIR/(KDLI+CFLLIR+LIVER_1RMN) !BOUND CONCENTRATION 46 4 7 !VARIABLE ELIMINATION BASED ON THE CYP1A2 4 8 KBILE_LI_T =((CYP1A2_1OUT-CYP1A2_1A2)/CYP1A2_1A2)*Kelv ! INDUCED BILIARY 4 9 EXCRETION RATE CONSTANT IN LIVER 5 0 REXCLI = KBILE_LI_T*CFLLIR*WLI ! DOSE-DEPENDENT BILIARY EXCRETION RATE 51 EXCLI = INTEG(REXCLI,0.0) 52 53 !UNIT CONVERSION POST SIMULATION 5 4 CLITOTAL= (ALI + ALIB)/(WLI + WLIB) ! TOTAL CONCENTRATION IN NMOL/ML 55 PRCT_LI = (CLITOTAL/(MSTT+1E-30))*100 5 6 PRCT_LIIV = (CLITOTAL/(IV_RlateR+1E-30))*100.0 This document is a draftfor review purposes only and does not constitute Agency policy. C-63 DRAFT--DO NOT CITE OR QUOTE 1 Rec_occ= CFLLIR/(KDLI+CFLLIR) 2 CLINGKG=CLITOTAL*MW*UNITCORR ! LIVER CONCENTRATION NG/KG 3 AUCLI_NGKGH=INTEG(CLINGKG,0.0) 4 CBNDLINGKG = CBNDLI*MW*UNITCORR 5 AUCBNDLI NGKGH =INTEG(CBNDLINGKG,0.0) 6 7 8 !CHEMICAL IN CYP450 (1A2) COMPARTMENT 9 CYP1A2 1KINP = CYP1A2 1KOUT* CYP1A2 1OUTZ 10 11 12 ! MODIFICATION ON OCTOBER 6, 2009 13 CYP1A2_1OUT =INTEG(CYP1A2_1KINP * (1.0 + CYP1A2_1EMAX *(CBNDLI+1.0e-30)**HILL 14 & 15 /(CYP1A2_1EC50**HILL + (CBNDLI+1.0e-30)**HILL)) & 16 - CYP1A2_1KOUT*CYP1A2_1OUT, CYP1A2_1OUTZ) 17 18 ! EQUATIONS INCORPORATING DELAY OF CYP1A2 PRODUCTION (NOT USED IN 19 SIMULATIONS) 20 21 CYP1A2_1RO2 = (CYP1A2_1OUT - CYP1A2_1O2)/ CYP1A2_1TAU 2 2 CYP1A2_1O2 =INTEG(CYP1A2_1RO2, CYP1A2_1A1) 23 2 4 CYP1A2_1RO3 = (CYP1A2_1O2 - CYP1A2_1O3)/ CYP1A2_1TAU 2 5 CYP1A2_1O3 =INTEG(CYP1A2_1RO3, CYP1A2_1A2) 26 2 7 ! TRANSFER OF DIOXIN FROM PLACENTA TO FETUS 2 8 ! FETAL EXPOSURE ONLY DURING EXPOSURE 29 3 0 IF (T.LT.TRANSTIME_ON) THEN 31 SWITCH_trans = 0.0 3 2 ELSE 33 SWITCH_trans = 1.0 3 4 END IF 35 3 6 !TRANSFER OF DIOXIN FROM PLACENTA TO FETUS 3 7 ! MODIFICATION 26 SEPTEMBER 2003 38 3 9 CONSTANT PFETUS= 4.0 ! 4 0 CONSTANT CLPLA_FET = 0.17 ! 41 4 2 RAMPF = (CLPLA_FET*CPLA) *SWITCH_trans 43 AMPF=INTEG(RAMPF,0.0) 44 4 5 !TRANSFER OF DIOXIN FROM FETUS TO PLACENTA 4 6 RAFPM = (CLPLA_FET*CFETUS_v)*SWITCH_trans ! 4 7 AFPM = INTEG(RAFPM,0.0) 48 4 9 ! TCDD IN PLACENTA (MOTHER) COMPARTMENT 5 0 RAPLAB= QPLA*(CA - CPLAB)-PAPLA*(CPLAB -CFLPLAR) ! NMOL/H) 51 APLAB = INTEG(RAPLAB,0.0) ! (NMOL) 5 2 CPLAB = APLAB/(WPLAB+1E-30) ! (NMOL/ML) 53 RAPLA = PAPLA*(CPLAB-CFLPLAR)-RAMPF + RAFPM ! (NMOL/H) 5 4 APLA = INTEG(RAPLA,0.0) ! (NMOL) 55 CPLA = APLA/(WPLA+1e-30) ! (NMOL/ML) 56 This document is a draftfor review purposes only and does not constitute Agency policy. C-64 DRAFT--DO NOT CITE OR QUOTE 1 2 PARAMETER (PARA_ZERO = 1.0E-30) 3 CFLPLA= IMPLC(CPLA-(CFLPLAR*PPLA +(PLABMAX*CFLPLAR/(KDPLA& 4 +CFLPLAR+PARA_ZERO)))-CFLPLA,CFLPLA0) 5 CFLPLAR=DIM(CFLPLA,0.0) 6 7 !UNIT CONVERSION POST SIMULATION 8 CPLATOTAL= (APLA + APLAB)/((WPLA + WPLAB)+1e-30)! TOTAL CONCENTRATION IN 9 NMOL/ML 10 PRCT_PLA = (CPLATOTAL/(MSTT+1E-30))*100 11 PRCT PLAIV = (CPLATOTAL/(IV RlateR+1E-30))*100 12 13 14 !FETUS COMPARTMENT 15 RAFETUS= RAMPF-RAFPM 16 AFETUS=INTEG(RAFETUS,0.0) 17 CFETUS=AFETUS/(WTFE+1E-30) 18 CFETOTAL= CFETUS 19 CFETUS_v = CFETUS/PFETUS 20 21 ! UNIT CONVERSION POST SIMULATION 2 2 CFETUSNGKG = CFETUS*MW*UNITCORR !(NG/KG) 2 3 AUC_FENGKGH = INTEG(CFETUSNGKG,0.0) 2 4 PRCT_FE = (CFETOTAL/(MSTT+1E-30))*100 2 5 PRCT FEIV = (CFETOTAL/(IV RlateR+1E-30))*100 26 27 2 8 ! ------------ CONTROL MASS BALANCE ---------- 2 9 BDOSE= IVDOSE +LYMLUM+LIMLUM 3 0 BMASSE = EXCLI+AURI+AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB+AFETUS 31 BDIFF = BDOSE-BMASSE 32 33 !BODY BURDEN (NG) 3 4 BODY_BURDEN = AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB ! 35 BBFETUSNG = AFETUS*MW*UNITCORR ! UNIT (NG) 3 6 ! BODY BURDEN IN TERMS OF CONCENTRATION (NG/KG) 3 7 BBNGKG =(((AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB)/WT0)*MW*UNITCORR) ! 3 8 AUC BBNGKGH=INTEG(BBNGKG,0.0) 39 40 41 --------COMMAND OF THE END OF SIMULATION -------- 4 2 TERMT (T.GE. TimeLimit, 'Time limit has been reached.') 4 3 END ! END OF THE DERIVATIVE SECTION 4 4 END ! END OF THE DYNAMIC SECTION 4 5 END ! END OF THE PROGRAM 46 47 4 8 C .2.4.2. Input Files 4 9 C .2.4.2.1. Bell et al. (2007). 5 0 %clear variable 51 output @clear 5 2 prepare @clear T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH 53 AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH 5 4 CBNGKG AUC_CBNGKGH This document is a draftfor review purposes only and does not constitute Agency policy. C-65 DRAFT--DO NOT CITE OR QUOTE 1 2 %output @nciout=1 T BBFETUSNG %AJS turned off 9/21/09 3 4 %Bell et al. 2007 (rat species) 5 %protocol: daily dietary dose for 12 weeks followed by a two-week mating 6 time and 21-day gestation period 7 %DevTCDD4Species.csl 8 %RAT GESTATIONAL ICF F083109.csl (now 09-11-09) 9 %dose levels: 0.0024, 0.008, 0.046 ug/kg/d with 0.00003 ug/kg/d background 10 %dose levels: 2.4, 8, 46 ng/kg/d with 0.03 ng/kg/day background 11 12 %EXPOSURES SCENARIOS 13 MAXT = 0.01 14 CINT = 0.1 % 15 EXP TIME ON =0 OO delay before begin exposure (HOUR) 16 EXP TIME OFF = 2856 O% TIME EXPOSURE STOP (HOUR) 12 weeks 17 exposure + 2 weeks for mating + 21 days gestation with exposure 18 DAY CYCLE = 24 19 BCK TIME ON = 0. O% DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 2 0 BCK TIME OFF = 2856. % TIME OF BACKGROUND EXPOSURE STOP (HOUR) 21 IV LACK = 505. 2 2 IV PERIOD = 505. 23 TIMELIMIT = 2856 O% SIMULATION LIMIT TIME (HOUR) 2 4 BW T0 = 85 25 MATTING = 2352 O% BEGINNING MATING (HOUR) 2 6 TRANSTIME ON = 2496 O% SHOULD BE MATING TIME + 6 DAYS(144 HOURS) 2 7 N FETUS = 10 28 2 9 %EXPOSURE DOSE SCENARIOS (UG/KG) 3 0 MSTOT = 0.00243 % ORAL EXPOSURE DOSE (UG/KG) 31 3 2 %MSTOT = 0.008 % ORAL EXPOSURE DOSE (UG/KG) 33 3 4 %MSTOT = 0. 0461 % ORAL EXPOSURE DOSE (UG/KG) 35 3 6 C .2.4.2.2. Haavisto et al. (2006). 3 7 %clear variable 3 8 output @clear 3 9 prepare @clear T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH 4 0 AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH 41 CBNGKG AUC_CBNGKGH 42 4 3 %Haavisto et al. 2006 4 4 %protocol: single dose on GD 13 4 5 %dose levels: 0.04, 0.2, and 1.0 ug/kg on GD 13 4 6 %dose levels: 40, 200, and 1,000 ng/kg on GD 13 47 4 8 MAXT = 0.001 4 9 CINT = 0.1 50 51 %EXPOSURES SCENARIOS 5 2 EXP TIME ON = 312 O% TIME AT WHICH EXPOSURE BEGINS (HOUR) 53 EXP "t i m e OFF = 335 O% TIME AT WHICH EXPOSURE ENDS (HOUR) 5 4 DAY_CYCLE = 24 This document is a draftfor review purposes only and does not constitute Agency policy. C-66 DRAFT--DO NOT CITE OR QUOTE 1 BCK TIME ON = 0. 2 BEGINS (HOUR) 3 BCK TIME OFF = 0. 4 ENDS (HOUR) 5 IV LACK = 505 6 IV PERIOD = 505 7 TIMELIMIT = 336 8 BW T0 = 190 9 MATTING = 0. 10 TRANSTIME ON = 144. 11 HOURS) 12 N_FETUS = 10 13 14 %EXPOSURE DOSE SCENARIOS (UG/KG) 15 %MSTOT = 0.04 16 %MSTOT = 0.2 17 MSTOT = 1.0 18 19 2 0 C .2.4.2.3. Hojo et al. (2002). OO TIME AT WHICH BACKGROUND EXPOSURE O% TIME AT WHICH BACKGROUND EXPOSURE O% SIMULATION LIMIT TIME (HOUR) O% BEGINNING MATTING (HOUR) O% SHOULD BE MATTING TIME + 6 DAYS(144 % ORAL EXPOSURE DOSE (UG/KG) % ORAL EXPOSURE DOSE (UG/KG) % ORAL EXPOSURE DOSE (UG/KG) 21 %clear variable 2 2 output @clear 2 3 prepare @clear T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH 2 4 AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH 2 5 CBNGKG AUC_CBNGKGH 2 6 %Hojo et al. 2002 2 7 %protocol: single oral dose at GD8 2 8 %DevTCDD4Species.csl 2 9 %RAT_GESTATIONAL_ICF_F083109.csl (now 09-11-09) 3 0 %RAT_GESTATIONAL_ICF_F092009.csl (now 09-21-09 31 %dose levels: 0.02 0.06, 0.18 ug/kg at GD8 3 2 %dose levels: 20, 60, 180 ng/kg at GD8 33 % author provided the body weight for each group at the beginning og 3 4 gestation (g) 3 5 %20 ng/kg BW = 271g 3 6 %60 ng/kg BW = 275g 3 7 %180 ng/kg BW = 262g 38 3 9 %EXPOSURES SCENARIOS 4 0 MAXT= 0.001 41 CINT =0.1 OO 4 2 EXP_TIME_ON = 192 % delay before begin exposure (HOUR) 4 3 EXP_TIME_OFF = 216 % TIME EXPOSURE STOP (HOUR) 4 4 DAY_CYCLE = 24 4 5 BCK_TIME_ON = 0. % DELAY BEFORE BACGROUND EXPOSURE (HOUR) 4 6 BCK_TIME_OFF = 0. % TIME OF BACKGROUND EXPOSURE STOP (HOUR) 4 7 IV_LACK = 505 4 8 IV_PERIOD = 505 4 9 TIMELIMIT = 216 % SIMULATION LIMIT TIME (HOUR) 5 0 % BW_T0 = 190 51 MATTING = 0. % BEGINNING MATTING (HOUR) 5 2 TRANSTIME_ON = 144. % SHOULD BE MATTING TIME + 6 DAYS(144 53 HOURS) 5 4 N_FETUS = 10 55 This document is a draftfor review purposes only and does not constitute Agency policy. C-67 DRAFT--DO NOT CITE OR QUOTE 1 %EXPOSURE DOSE SCENARIOS (UG/KG) 2 3 %MSTOT = 0.02 % ORAL EXPOSURE DOSE (UG/KG) 4 %BW T0 = 275 % 20 ng/kg BW = 271g 5 6 %MSTOT = 0.06 % ORAL EXPOSURE DOSE (UG/KG) 7 %BW T0 = 262 %60 ng/kg BW = 275g 8 9 MSTOT = 0.18 % ORAL EXPOSURE DOSE (UG/KG) 10 BW T0 = 278 %180 ng/kg BW = 262g 11 12 C .2.4.2.4. Ikeda et al. (2005). 13 %clear variable 14 output @clear 15 prepare @clear T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH 16 AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH 17 18 %Ikeda et al. 2005 (rat species) 19 %protocol: loading dose of 400 ng/kg followed by weekly maintenance doses of 2 0 80 ng/kg for 6 weeks, 21 %dose levels: 0.4 ug/kg/day followed by weekly 0.08 ug/kg/day 2 2 %dose levels: 400 ng/kg/day followed by weekly 80 ng/kg/day 23 2 4 %EXPOSURES SCENARIOS 2 5 MAXT = .1 2 6 CINT = 0.1 % 2 7 EXP TIME ON =0 OO TIME AT WHICH EXPOSURE BEGINS (HOUR) 2 8 EXP TIME OFF = 1008 O% TIME AT WHICH EXPOSURE ENDS (HOUR); PRE- 2 9 MATING (2 WEEKS ) + MATING (1 WEEK) + GESTATION (3 WEEKS) 3 0 DAY CYCLE = 168 O% WEEKLY CYCLE 31 BCK TIME ON = 0. O% TIME AT WHICH BACKGROUND EXPOSURE BEGINS 3 2 (HOUR) 33 BCK TIME OFF = 167. O% TIME AT WHICH BACKGROUND EXPOSURE ENDS 3 4 (HOUR) 3 5 IV LACK = 505. 3 6 IV PERIOD = 505. 3 7 TIMELIMIT = 1008 O% SIMULATION TIME LIMIT (HOUR) 3 8 BW T0 = 250 3 9 MATTING = 504 O% BEGINNING OF MATING (HOUR) 4 0 TRANSTIME ON = 648 O% SHOULD BE MATING TIME + 6 DAYS (144 HOURS) 41 N_FETUS = 10 42 4 3 %EXPOSURE DOSE SCENARIOS (UG/KG) 4 4 MSTOT = 0.08 O% ORAL EXPOSURE DOSE IN UG/KG 4 5 MSTOTBCKGR = 0.32 O% BACKGROUND EXPOSURE IN UG/KG 46 47 4 8 C .2.4.2.5. Kattainen et al. (2001). 4 9 %clear variable 5 0 output @clear 51 prepare @clear T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH 5 2 AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH 53 CBNGKG AUC_CBNGKGH 54 This document is a draftfor review purposes only and does not constitute Agency policy. C-68 DRAFT--DO NOT CITE OR QUOTE 1 %Kattainen et al. 2001 2 %protocol: single gavage at GD15 3 %DevTCDD4Species.csl 4 %RAT_GESTATIONAL_ICF_F083109.csl (now 09-11-09) 5 %dose levels: 0.03 0.1, 0.3, 1 ug/kg at GD15 6 %dose levels: 30, 100 300, 1000 ng/kg at GD15 7 8 MAXT=0.001 9 CINT =0.1 10 11 %EXPOSURES SCENARIOS 12 EXP TIME ON = 336 OO delay before begin exposure (HOUR) 13 EXP TIME OFF = 360 OO TIME EXPOSURE STOP (HOUR) 14 DAY CYCLE = 24 15 BCK TIME ON = 0. OO DELAY BEFORE BACKGROUND EXPOSURE 16 (HOUR) 17 BCK TIME OFF = 0. OO TIME OF BACKGROUND EXPOSURE STOP 18 (HOUR) 19 IV LACK = 505 2 0 IV PERIOD = 505 21 TIMELIMIT = 360 OO SIMULATION LIMIT TIME (HOUR) 2 2 BW T0 = 190 23 MATTING = 0. OO BEGINNING MATTING (HOUR) 2 4 TRANSTIME ON = 144. OO SHOULD BE MATTING TIME + 6 DAYS(144 2 5 HOURS) 2 6 N FETUS = 10 27 2 8 %EXPOSURE DOSE SCENARIOS (UG/KG) 2 9 OMSTOT = 0.03 3 0 OMSTOT = 0.1 31 OMSTOT = 0.3 3 2 MSTOT =1 OO ORAL EXPOSURE DOSE OO ORAL EXPOSURE DOSE OO ORAL EXPOSURE DOSE OO ORAL EXPOSURE DOSE (UG/KG) (UG/KG) (UG/KG) (UG/KG 33 3 4 C .2.4.2.6. M arkowski et al. (2001). 3 5 %clear variable 3 6 output @clear 3 7 prepare @clear T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH 3 8 AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH 3 9 CBNGKG AUC_CBNGKGH 40 41 %Markowski et al. 2001 4 2 %protocol: single gavage at GD18 43 %DevTCDD4Species.csl 4 4 %RAT_GESTATIONAL_ICF_F083109.csl (now 09-11-09) 4 5 %dose levels: 0.02 0.06, 0.18 ug/kg at GD18 4 6 %dose levels: 20, 60, 180 ng/kg at GD18 47 4 8 %EXPOSURES SCENARIOS 4 9 MAXT=0.0001 5 0 CINT =0.1 % 51 EXP TIME ON = 408 OO delay before begin exposure (HOUR) 5 2 EXP TIME OFF = 432 OO TIME EXPOSURE STOP (HOUR) 53 DAY CYCLE = 24 5 4 BCK TIME ON = 0. OO DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 55 BCK TIME OFF = 0. OO TIME OF BACKGROUND EXPOSURE STOP (HOUR) This document is a draftfor review purposes only and does not constitute Agency policy. C-69 DRAFT--DO NOT CITE OR QUOTE 1 IV LACK = 505 2 IV PERIOD = 505 3 TIMELIMIT = 432 OO SIMULATION LIMIT TIME (HOUR) 4 BW T0 = 190 5 MATTING = 0. OO BEGINNING MATING (HOUR) 6 TRANSTIME ON = 144. OO SHOULD BE MATING TIME + 6 DAYS(144 HOURS) 7 N FETUS = 10 8 9 %EXPOSURE DOSE SCENARIOS (UG/KG) 10 %MSTOT = 0.02 % ORAL EXPOSURE DOSE (UG/KG) 11 %MSTOT = 0.06 % ORAL EXPOSURE DOSE (UG/KG) 12 MSTOT = 0.18 % ORAL EXPOSURE DOSE (UG/KG) 13 14 C .2.4.2.7. Miettinen et al. (2006). 15 %clear variable 16 output @clear 17 prepare @clear T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH 18 AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH 19 CBNGKG AUC_CBNGKGH 20 21 %Miettinen et al. 2006 2 2 %protocol: single oral dose at GD15 23 %DevTCDD4Species.csl 2 4 %RAT_GESTATIONAL_ICF_F083109.csl (now 09-11-09) 2 5 %dose levels: 0.03 0.1, 0.3, 1 ug/kg at GD15 2 6 %dose levels: 30, 100, 300, 1000 ng/kg at GD15 27 2 8 MAXT=0.01 2 9 CINT =0.1 % 30 31 %EXPOSURES SCENARIOS 3 2 EXP TIME ON = 336 oO delay before begin exposure (HOUR) 33 EXP TIME OFF = 360 oO TIME EXPOSURE STOP (HOUR) 3 4 DAY CYCLE = 24 3 5 BCK TIME ON = 0. oO DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 3 6 BCK TIME OFF = 0. oO TIME OF BACKGROUND EXPOSURE STOP (HOUR) 3 7 IV LACK = 505 3 8 IV PERIOD = 505 3 9 TIMELIMIT = 360 oO SIMULATION LIMIT TIME (HOUR) 4 0 BW T0 = 180 41 MATTING = 0. oO BEGINNING MATING (HOUR) 4 2 TRANSTIME ON = 144. oO SHOULD BE MATING TIME + 6 DAYS(144 HOURS) 4 3 N FETUS = 10 44 4 5 %EXPOSURE DOSE SCENARIOS (UG/KG) 4 6 %MSTOT = 0.03 oO ORAL EXPOSURE DOSE (UG/KG) 4 7 %MSTOT = 0.1 O ORAL EXPOSURE DOSE (UG/KG) o CO 4 8 %MSTOT = O ORAL EXPOSURE DOSE (UG/KG) 4 9 MSTOT = 1 O ORAL EXPOSURE DOSE (UG/KG) 50 51 C .2.4.2.8. Nohara et al. (2000). 5 2 %clear variable 53 output @clear This document is a draftfor review purposes only and does not constitute Agency policy. C-70 DRAFT--DO NOT CITE OR QUOTE 1 prepare @clear T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH 2 AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH 3 CBNGKG AUC_CBNGKGH 4 5 %Nohara et al. 2000 6 %protocol: single gavage at GD15 7 %DevTCDD4Species.csl 8 %RAT_GESTATIONAL_ICF_F083109.csl (now 09-11-09) 9 %dose levels: 0.0125, 0.050, 0.2, or 0.8 ug TCDD:kg body weight by gavage on 10 GD15. 11 %dose levels: 12.5, 50, 200, or 800 ng TCDD:kg body weight by gavage on GD15. 12 13 MAXT=0.01 14 CINT =0.1 % 15 16 %EXPOSURES SCENARIOS 17 EXP TIME ON = 336 % delay before begin exposure (HOUR) 18 EXP TIME OFF = 360 % TIME EXPOSURE STOP (HOUR) 19 DAY CYCLE = 24 2 0 BCK TIME ON = 0. % DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 21 BCK TIME OFF = 0. % TIME OF BACKGROUND EXPOSURE STOP (HOUR) 2 2 IV LACK = 505 2 3 IV PERIOD = 505 2 4 TIMELIMIT = 360 % SIMULATION LIMIT TIME (HOUR) 2 5 BW T0 = 180 2 6 MATTING = 0. % BEGINNING MATTING (HOUR) 2 7 TRANSTIME ON = 144. % SHOULD BE MATTING TIME + 6 DAYS(144 HOURS) 2 8 N_FETUS = 10 29 3 0 %EXPOSURE DOSE SCENARIOS (UG/KG) 31 %MSTOT = 0.0125 % ORAL EXPOSURE DOSE (UG/KG) 3 2 %MSTOT = 0.050 % ORAL EXPOSURE DOSE (UG/KG) 33 %MSTOT = 0.2 % ORAL EXPOSURE DOSE (UG/KG) 3 4 MSTOT o CO % ORAL EXPOSURE DOSE (UG/KG) 35 3 6 C .2.4.2.9. Ohsako et al. (2001). 3 7 %clear variable 3 8 output @clear 3 9 prepare @clear T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH 4 0 AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH 41 CBNGKG AUC_CBNGKGH 42 4 3 %Ohsako et al. 2001 4 4 %protocol: single oral dose at GD15 4 5 %DevTCDD4Species.csl 4 6 %RAT_GESTATIONAL_ICF_F083109.csl (now 09-11-09) 4 7 %RAT_GESTATIONAL_ICF_F092009.csl (now 09-21-09) 4 8 %dose levels: 0.0125, 0.05, 0.2, 0.8 ug/kg at GD15 4 9 %dose levels: 12.5, 50, 200, 800 ng/kg at GD15 50 51 %EXPOSURES SCENARIOS 5 2 MAXT=0.01 53 CINT =0.1 OO 5 4 EXP_TIME_ON = 360 % delay before begin exposure (HOUR) 55 EXP TIME OFF = 384 % TIME EXPOSURE STOP (HOUR) This document is a draftfor review purposes only and does not constitute Agency policy. C-71 DRAFT--DO NOT CITE OR QUOTE 1 DAY CYCLE = 24 2 BCK TIME ON = 0. % DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 3 BCK TIME OFF = 0. % TIME OF BACKGROUND EXPOSURE STOP (HOUR) 4 IV LACK = 505 5 IV PERIOD = 505 6 TIMELIMIT = 384 % SIMULATION LIMIT TIME (HOUR) 7 BW T0 = 200 8 MATTING = 0. % BEGINNING MATTING (HOUR) 9 TRANSTIME ON = 144. % SHOULD BE MATTING TIME + 6 DAYS((144 10 HOURS) 11 N_FETUS = 10 12 13 %EXPOSURE DOSE SCENARIOS (UG/KG) 14 15 %MSTOT = 0.0125 OO ORAL EXPOSURE DOSE (UG/KG) 16 %MSTOT = 0.05 OO ORAL EXPOSURE DOSE (UG/KG) 17 %MSTOT = 0.20 OO ORAL EXPOSURE DOSE (UG/KG) 18 MSTOT = 0.80 % ORAL EXPOSURE DOSE (UG/KG) 19 2 0 C .2.4.2.10. Schantz et al. (1996) and Amin et al. (2000). 21 %clear variable 2 2 output @clear 2 3 prepare @clear T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH 2 4 AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH 2 5 CBNGKG AUC_CBNGKGH 26 2 7 %Amin et al. 2000 (rat species) and Schantz et al. 1996 2 8 %protocol: daily doses on GDs 10 to 16 2 9 %DevTCDD4Species.csl 3 0 %RAT_GESTATIONAL_ICF_F083109.csl (now 09-11-09) 31 %dose levels: 25 and 100 ng/kg/day 3 2 %dose levels: 0.025 and 0.100 ug/kg/day 33 3 4 %EXPOSURES SCENARIOS 3 5 MAXT = 0.001 3 6 CINT = 0.1 O 3 7 EXP TIME ON = 240. % TIME AT WHICH EXPOSURE BEGINS (HOUR) 3 8 EXP TIME OFF = 384. OO TIME AT WHICH EXPOSURE ENDS (HOUR) GD 10 3 9 to 16 4 0 DAY CYCLE = 24 OO weekly cycle 41 BCK TIME ON = 1000. % DELAY BEFORE BACKGROUND EXPOSURE (HOUR) 4 2 BCK TIME OFF = 1000. % TIME OF BACKGROUND EXPOSURE STOP (HOUR) 4 3 IV LACK = 505. 4 4 IV PERIOD = 505. 4 5 TIMELIMIT = 384. OO SIMULATION LIMIT TIME (HOUR) 4 6 BW T0 = 250. 4 7 MATTING = 0 % BEGINNING MATTING (HOUR) 4 8 TRANSTIME ON = 144. OO SHOULD BE MATTING TIME + 6 DAYS(144 HOURS) 4 9 N_FETUS = 10 50 51 %EXPOSURE DOSE SCENARIOS (UG/KG) 5 2 %MSTOT = .025 % ORAL EXPOSURE DOSE (UG/KG) 53 MSTOT = .100 54 MSTOTBCKGR = 0 % Background Exposure (UG/KG) 55 This document is a draftfor review purposes only and does not constitute Agency policy. C-72 DRAFT--DO NOT CITE OR QUOTE 1 C .2.4.2.11. Seo et al. (1995). 2 %clear variable 3 output @clear 4 prepare @clear T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH 5 AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH 6 CBNGKG AUC_CBNGKGH 7 8 %Seo et al. 1995 9 %protocol: daily doses on GDs 10-16 10 %DevTCDD4Species.csl 11 %RAT_GESTATIONAL_ICF_F083109.csl (now 09-11-09) 12 %dose levels: 0.025 and 0.1 ug/kg on GDs 10-16 13 %dose levels: 25 and 100 ng/kg on GDs 10-16 14 15 MAXT = 0.01 16 CINT = 0.1 17 18 %EXPOSURES SCENARIOS 19 EXP TIME ON = 240 OO TIME AT WHICH EXPOSURE BEGINS (HOUR) 2 0 EXP TIME OFF = 384 OO TIME AT WHICH EXPOSURE ENDS (HOUR) 21 DAY CYCLE = 24 2 2 BCK TIME ON = 0. OO TIME AT WHICH BACKGROUND EXPOSURE 2 3 BEGINS (HOUR) 2 4 BCK TIME OFF = 0. OO TIME AT WHICH BACKGROUND EXPOSURE 2 5 ENDS (HOUR) 2 6 IV LACK = 505 2 7 IV PERIOD = 505 2 8 TIMELIMIT = 384 OO SIMULATION LIMIT TIME (HOUR) 2 9 BW T0 = 190 3 0 MATTING = 0. OO BEGINNING MATTING (HOUR) 31 TRANSTIME ON = 144. OO SHOULD BE MATTING TIME + 6 DAYS(144 3 2 HOURS) 33 N_FETUS = 10 34 3 5 %EXPOSURE DOSE SCENARIOS (UG/KG) 3 6 %MSTOT = 0 .025 OO ORAL EXPOSURE DOSE (UG/KG) 3 7 MSTOT = 0. 1 % ORAL EXPOSURE DOSE (UG/KG) 38 39 C.2.5. M ouse Standard M odel 4 0 C.2.5.1. Model Code 41 P R O G R A M : 'T h re e C o m p a rtm e n t P B P K M o d e l fo r T C D D in M ice: S ta n d a rd M o d e l (N o n 42 G estation)' 43 4 4 !Mice Dioxin 3C June09 1 icf afterKKfix v3 mousenongest.cs l 4 5 !MICE_NON_GESTAT_ICF_F0831097csl __ 4 6 !MICE_NON_GESTAT_ICF_F093009.csl 4 7 !MICE NON GESTAT ICF F100609.csl 4 8 !***************************************************** 49 5 0 INITIAL ! INITIALIZATION OF PARAMETERS 51 5 2 (SIMULATION PARAMETERS ==== This document is a draftfor review purposes only and does not constitute Agency policy. C-73 DRAFT--DO NOT CITE OR QUOTE 1 CONSTANT PARA_ZERO = 1D-30 2 CONSTANT EXP_TIME_ON = 0.0 ! TIME AT WHICH EXPOSURE BEGINS 3 (HOURS) 4 CONSTANT EXP_TIME_OFF = 2832 ! TIME AT WHICH EXPOSURE ENDS 5 (HOURS) 6 CONSTANT DAY_CYCLE = 24 ! NUMBER OF HOURS BETWEEN DOSES 7 (HOURS) 8 CONSTANT BCK_TIME_ON = 0.0 ! TIME AT WHICH BACKGROUND EXPOSURE 9 BEGINS (HOURS) 10 CONSTANT BCK_TIME_OFF = 0.0 ! TIME AT WHICH BACKGROUND EXPOSURE 11 ENDS (HOURS) 12 13 CONSTANT MW=322 ! MOLECULAR WEIGHT (NG/NMOL) 14 CONSTANT SERBLO = 0.55 15 CONSTANT UNITCORR = 1000 16 17 !CONSTANT EXPOSURE CONTROL =========== 18 !ACUTE, SUBCHRONIC, CHRONIC EXPOSURE ===== 19 !OR BACKGROUND EXPOSURE (IN THIS CASE 3 TIMES A DAY)=== oo oo 2 0 CONSTANT MSTOTBCKGR = !ORAL BACKGROUND EXPOSURE DOSE 21 (UG/KG) 2 2 CONSTANT MSTOT = 0.15 !ORAL EXPOSURE DOSE (UG/KG) 2 3 CONSTANT MSTOTsc = ! SUBCUTANEOUS EXPOSURE DOSE 2 4 (UG/KG) 25 2 6 !ORAL ABSORPTION 2 7 MSTOT NM MSTOT/MW !AMOUNT IN NMOL/G 28 2 9 ! INTRAVENOUS ABSORPTION 3 0 CONSTANT DOSEIV = 0.0 !INJECTED DOSE (UG/KG) 31 DOSEIV NM = DOSEIV/MW ! CONVERTS THE INJECTED DOSE TO NMOL/G 32 33 !INITIAL GUESS OF THE FREE CONCENTRATION IN THE LIGAND (COMPARTMENT 3 4 INDICATED BELOW)==== 3 5 CONSTANT CFLLI0 = 0.0 !LIVER (NMOL/ML) 36 3 7 !BINDING CAPACITY (AhR) FOR NON LINEAR BINDING (COMPARTMENT INDICATED 3 8 BELOW) (NMOL/ML) 3 9 CONSTANT LIBMAX = 3.5e-4 ! LIVER (NMOL/ML), WANG ET AL. 4 0 1997 41 4 2 ! PROTEIN AFFINITY CONSTANTS (1A2 OR AhR, COMPARTMENT INDICATED BELOW) 4 3 (NMOL/ML)=== 4 4 CONSTANT KDLI = 1.0e-4 !LIVER (AhR)(NMOL/ML), WANG ET AL. 4 5 1997 4 6 CONSTANT KDLI2 = 2.0e-2 !LIVER (1A2)(NMOL/ML), EMOND ET AL. 4 7 2004 48 4 9 !===EXCRETION AND ABSORPTION CONSTANT (OPTIMIZED) 5 0 CONSTANT KST = 0.3 ! GASTRIC RATE CONSTANT (HR-1), 51 CONSTANT KABS = 0.48 !INTESTINAL ABSORPTION CONSTANT (HR-1) ), 5 2 WANG ET AL. 1997 53 5 4 ! ELIMINATION CONSTANTS 55 CONSTANT CLURI = 0.09 ! URINARYCLEARANCE (ML/HR) 56 This document is a draftfor review purposes only and does not constitute Agency policy. C-74 DRAFT--DO NOT CITE OR QUOTE 1 ! ==test elimination variable 2 constant kelv = 0.4 ! INTERSPECIES VARIABLE ELIMINATION 3 CONSTANT (1/HOUR) 4 5 ! CONSTANT TO DIVIDE THE ABSORPTION INTO LYMPHATIC AND PORTAL FRACTIONS 6 CONSTANT A = 0.7 ! LYMPHATIC FRACTION, WANG ET AL. 7 1997 8 9 IPARTITION COEFFICIENTS OPTIMIZED 10 CONSTANT PF = 400 ! ADIPOSE TISSUE/BLOOD 11 CONSTANT PRE =3 ! REST OF THE BODY/BLOOD, WANG ET 12 AL. 2000 13 CONSTANT PLI =6 ! LIVER/BLOOD, WANG ET AL. 1997 14 15 !===PARAMETER FOR INDUCTION OF CYP 1A2 16 CONSTANT PAS_INDUC= 1.0 ! INCLUDE INDUCTION? (1 = YES, 0 = NO) 17 CONSTANT CYP1A2 1OUTZ = 1.6 ! DEGRADATION CONCENTRATION CONSTANT OF 1A2 18 (NMOL/ML) 19 CONSTANT CYP1A2 1A1 = 1.5 ! BASAL CONCENTRATION OF 1A1 (NMOL/ML) 2 0 CONSTANT CYP1A2 1EC50 = 0.13 ! DISSOCIATION CONSTANT TCDD-CYP1A2 (NMOL/ML) 21 CONSTANT CYP1A2 1A2 = 1.5 ! BASAL CONCENTRATION OF 1A2 (NMOL/ML) 2 2 CONSTANT CYP1A2 1KOUT = 0.1 ! FIRST ORDER RATE OF DEGRADATION (H-1) 2 3 CONSTANT CYP1A2 1TAU = 1.5 HOLDING TIME (H) 2 4 CONSTANT CYP1A2 1EMAX = 600 ! MAXIMUM INDUCTION OVER BASAL EFFECT 2 5 (UNITLESS) 2 6 CONSTANT HILL = 0.6 !HILL CONSTANT; COOPERATIVELY LIGAND BINDING 2 7 EFFECT CONSTANT (UNITLESS) 2 8 IDIFFUSIONAL PERMEABILITY FRACTION 2 9 CONSTANT PAFF = 0. 12 ! ADIPOSE (UNITLESS), WANG ET AL. 2000 3 0 CONSTANT PAREF = 0. 03 ! REST OF THE BODY (UNITLESS) 31 CONSTANT PALIF = 0. 35 ! LIVER (UNITLESS) 32 33 !COMPARTMENT TISSUE BLOOD VOLUME ========= 3 4 CONSTANT WLI0 = 0 .0549 ! LIVER, ILSI 1994 3 5 CONSTANT WF0 = 0 .069 ! ADIPOSE 36 3 7 !TISSUE BLOOD FLOW EXPRESSED AS A FRACTION OF CARDIAC OUTPUT 3 8 CONSTANT QFF = 0.070 ! ADIPOSE TISSUE BLOOD FLOW FRACTION 3 9 (UNITLESS), LEUNG ET AL. 1990 4 0 CONSTANT QLIF = 0.161 ! LIVER (UNITLESS) ILSI ET AL. 1994 41 4 2 !COMPARTMENT TISSUE BLOOD EXPRESSED AS A FRACTION OF THE TOTAL 43 COMPARTMENT VOLUME 4 4 CONSTANT WFB0 = 0.050 ! ADIPOSE TISSUE, WANG ET AL. 1997 4 5 CONSTANT WREB0 = 0.030 ! REST OF THE BODY, WANG ET AL. 1997 4 6 CONSTANT WLIB0 = 0.266 ! LIVER, WANG ET AL. 1997 47 4 8 ! EXPOSURE SCENARIO FOR UNIQUE OR REPETITIVE WEEKLY OR MONTHLY EXPOSURE 4 9 ! NUMBER OF EXPOSURES PER WEEK 5 0 CONSTANT WEEK_LACK = 0.0 ! DELAY BEFORE EXPOSURE ENDS (WEEK) 51 CONSTANT WEEK_PERIOD = 168 ! NUMBER OF HOURS IN THE WEEK (HOURS) 5 2 CONSTANT WEEK_FINISH = 120 ! TIME EXPOSURE ENDS (HOURS) 53 5 4 ! NUMBER OF EXPOSURES PER MONTH 55 CONSTANT MONTH_LACK = 0.0 ! DELAY BEFORE EXPOSURE (MONTH) 56 This document is a draftfor review purposes only and does not constitute Agency policy. C-75 DRAFT--DO NOT CITE OR QUOTE 1 !SET FOR BACKGROUND EXPOSURE=========== 2 ICONSTANT FOR BACKGROUND EXPOSURE=========== 3 CONSTANT Day_LACK_BG = 0.0 I DELAY BEFORE EXPOSURE BEGINS (HOURS) 4 CONSTANT Day_PERIOD_BG = 24 I LENGTH OF EXPOSURE (HOURS) 5 6 I NUMBER OF EXPOSURES PER WEEK 7 CONSTANT WEEK_LACK_BG = 0.0 I DELAY BEFORE BACKGROUD EXPOSURE (WEEK) 8 CONSTANT WEEK_PERIOD_BG = 168 INUMBER OF HOURS IN THE WEEK (HOURS) 9 CONSTANT WEEK_FINISH_BG = 168 I TIME EXPOSURE ENDS (HOURS) 10 11 IGROWTH CONSTANT FOR RAT AND MOUSE 12 ICONSTANT FOR MOTHER BODY WEIGHT GROWTH ====== 13 CONSTANT B W T 0 = 20 ICHANGED FOR SIMULATION 14 15 ICONSTANT USED IN CARDIAC OUTPUT EQUATION, HADDAD 2001 16 CONSTANT QCCAR =275 ICONSTANT (ML/MIN/KG) 17 18 I COMPARTMENT LIPID EXPRESSED AS THE FRACTION OF TOTAL LIPID 19 CONSTANT F_TOTLIP = 0.855 IADIPOSE TISSUE (UNITLESS) 2 0 CONSTANT B_TOTLIP = 0.0033 IBLOOD (UNITLESS) 21 CONSTANT RE_TOTLIP = 0.019 IREST OF THE BODY (UNITLESS) 2 2 CONSTANT LI TOTLIP = 0.06 ILIVER (UNITLESS) 23 2 4 END I END OF THE INITIAL SECTION 25 2 6 DYNAMIC I DYNAMIC SIMULATION SECTION 27 2 8 ALGORITHM IALG = 2 IGEAR METHOD 2 9 CINTERVAL CINT = 1.0 ICOMMUNICATION INTERVAL 3 0 MAXTERVAL MAXT = 1.0e+10 IMAXIMUM CALCULATION INTERVAL 31 MINTERVAL MINT = 1.0E-10 IMINIMUM CALCULATION INTERVAL 3 2 VARIABLE T = 0.0 IHOUR 33 CONSTANT TIMELIMIT = 2904 .0 ISIMULATION TIME LIMIT 3 4 (HOURS) 3 5 CINTXY = CINT 3 6 PFUNC = CINT 37 3 8 ITIME CONVERSION 3 9 DAY = T/24.0 ! TIME IN DAYS 4 0 WEEK = T/168.0 ! TIME IN WEEKS 41 MONTH = T/730.0 ! TIME IN MONTHS 4 2 YEAR = T/8760.0 ! TIME IN YEARS 43 4 4 !NMAX =MAX(T,CTFNGKG) 4 5 nmax =max(T,CFNGKG) 46 4 7 DERIVATIVE ! PORTION OF CODE THAT SOLVES DIFFERENTIAL EQUATIONS 48 4 9 ICHRONIC OR SUBCHRONIC EXPOSURE SCENARIO 5 0 INUMBER OF EXPOSURES PER DAY 51 DAY_LACK = EXP_TIME_ON DELAY BEFORE EXPOSURE BEGINS (HOURS) 5 2 DAY_PERIOD = DAY_CYCLE EXPOSURE PERIOD (HOURS) 53 DAY_FINISH = CINTXY LENGTH OF EXPOSURE (HOURS) 5 4 MONTH_PERIOD = TIMELIMIT EXPOSURE PERIOD (MONTHS) 55 MONTH FINISH = EXP TIME OFF LENGTH OF EXPOSURE (MONTHS) 56 This document is a draftfor review purposes only and does not constitute Agency policy. C-76 DRAFT--DO NOT CITE OR QUOTE 1 INUMBER OF EXPOSURES PER DAY AND MONTH 2 DAY_FINISH_BG = CINTXY 3 MONTH_LACK_BG = BCK_TIME_ON I DELAY BEFORE BACKGROUD EXPOSURE BEGINS 4 (MONTHS) 5 MONTH_PERIOD_BG = TIMELIMIT I BACKGROUND EXPOSURE PERIOD (MONTHS) 6 MONTH_FINISH_BG = BCK_TIME_OFF I LENGTH OF BACKGROUND EXPOSURE (MONTHS) 7 8 I FRACTION OF DIOXIN ABSORBED IN THE PORTAL FRACTION OF THE LIVER 9 B = 1.0-A 10 11 12 (GROWTH UP EQUATION (G) 13 14 PARAMETER (BW_RMN = 1.0E-30) 15 WT0= (BW_T0 *(1.0+(0.41*T)/(1402.5+T+BW_RMN))) 16 17 ! VARIABILITY OF REST OF THE BODY DEPENDS ON OTHER ORGANS 18 (REST OF THE BODY FRACTION; UPDATED FOR EPA ASSESSMENT 19 WRE0 = (0.91 - (WLIB0*WLI0 + WFB0*WF0 + WLI0 + WF0))/(1+WREB0) 20 21 ! REST OF THE BODY BLOOD FLOW FRACTION 2 2 QREF = 1.0-(QFF+QLIF) (REST OF BODY BLOOD FLOW (ML/HR) 2 3 (SUMMATION OF BLOOD FLOW FRACTION (SHOULD BE EQUAL TO 1) 2 4 QTTQF = QFF+QREF+QLIF ! SUM MUST EQUAL 1 25 2 6 (COMPARTMENT VOLUME (G) 2 7 WF = WF0 * WT0 I ADIPOSE 2 8 WRE = WRE0 * WT0 I REST OF THE BODY 2 9 WLI = WLI0 * WT0 I LIVER 30 31 ICOMPARTMENT TISSUE BLOOD (G) 3 2 WFB = WFB0 * WF I ADIPOSE 33 WREB = WREB0 * WRE I REST OF THE BODY 3 4 WLIB = WLIB0 * WLI I LIVER 35 3 6 (CARDIAC OUTPUT FOR THE GIVEN BODY WEIGHT 3 7 QC= QCCAR*60*(WT0/1000.0)**0.75 38 3 9 QF = QFF*QC ( ADIPOSE TISSUE BLOOD FLOW RATE (ML/HR) 4 0 QLI = QLIF*QC (LIVER TISSUE BLOODFLOW RATE (ML/HR) 41 QRE = QREF*QC (REST OF THE BODYBLOOD FLOW RATE (ML/HR) 42 4 3 QTTQ = QF+QRE+QLI (TOTAL FLOW RATE (ML/HR) 44 4 5 (PERMEABILITY ORGAN FLOW (ML/HR) ======= 4 6 PAF = PAFF*QF (ADIPOSE TISSUE 4 7 PARE = PAREF*QRE (REST OF THE BODY 4 8 PALI = PALIF*QLI (LIVER TISSUE 49 5 0 (ABSORPTION SECTION 51 (ORAL 5 2 (BACKGROUND EXPOSURE 53 (EXPOSURE FOR STEADY STATE CONSIDERATION 5 4 (REPETITIVE EXPOSURE SCENARIO 55 5 6 MSTOT NMBCKGR = MSTOTBCKGR/322 (AMOUNT IN NMOL/G This document is a draftfor review purposes only and does not constitute Agency policy. C-77 DRAFT--DO NOT CITE OR QUOTE 1 MSTTBCKGR =MSTOT_NMBCKGR *WT0 2 3 IREPETITIVE ORAL BACKGROUND EXPOSURE SCENARIOS 4 DAY_EXPOSURE_BG = PULSE(DAY_LACK_BG,DAY_PERIOD_BG,DAY_FINISH_BG) 5 WEEK_EXPOSURE_BG = PULSE(WEEK_LACK_BG,WEEK_PERIOD_BG,WEEK_FINISH_BG) 6 MONTH_EXPOSURE_BG = PULSE(MONTH_LACK_BG,MONTH_PERIOD_BG,MONTH_FINISH_BG) 7 8 MSTTCH_BG = (DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG)*MSTTBCKGR 9 MSTTFR_BG = MSTTBCKGR/CINT 10 11 totalBG= integ (MSTTCH_BG,0.0) 12 CYCLE BG =DAY EXPOSURE BG*WEEK EXPOSURE BG*MONTH EXPOSURE BG 13 14 15 ICONDITIONAL ORAL EXPOSURE (BACKGROUND EXPOSURE) 16 IF (MSTTCH_BG.EQ.MSTTBCKGR) THEN 17 ABSMSTT_GB= MSTTFR_BG 18 ELSE 19 ABSMSTT_GB = 0.0 2 0 END IF 21 2 2 IEXPOSURE + IREPETITIVE EXPOSURE SCENARIO 2 3 IV= DOSEIV_NM * WT0 IAMOUNT IN NMOL 2 4 MSTT= MSTOT_NM * WT0 IAMOUNT IN NMOL 25 2 6 DAY_EXPOSURE = PULSE(DAY_LACK,DAY_PERIOD,DAY_FINISH) 2 7 WEEK_EXPOSURE = PULSE(WEEK_LACK,WEEK_PERIOD,WEEK_FINISH) 2 8 MONTH_EXPOSURE = PULSE(MONTH_LACK,MONTH_PERIOD,MONTH_FINISH) 29 3 0 MSTTCH = (DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE)*MSTT 31 CYCLE = DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE 32 33 SUMEXPEVENT= integ (CYCLE,0.0)*cint INUMBER OF CYCLE GENERATE DURING 3 4 SIMULATION 35 3 6 MSTTFR = MSTT/CINT 37 3 8 I CONDITIONAL ORAL EXPOSURE 3 9 IF (MSTTCH.EQ.MSTT) THEN 4 0 ABSMSTT= MSTTFR 41 ELSE 4 2 ABSMSTT = 0.0 4 3 END IF 44 4 5 CYCLETOT=INTEG(CYCLE,0.0) 46 47 4 8 !MASS CHANGE IN THE LUMEN 4 9 RMSTT= -(KST+KABS)*MST+ABSMSTT +ABSMSTT_GB ! RATE OF CHANGE (NMOL/H) 5 0 MST = INTEG(RMSTT,0.0) !AMOUNT OF STAY IN DUODENUM (NMOL) 51 5 2 !ABSORPTION IN LYMPH CIRCULATION 53 LYRMLUM = KABS*MST*A 5 4 LYMLUM = INTEG(LYRMLUM,0.0) 55 5 6 !ABSORPTION IN PORTAL CIRCULATION This document is a draftfor review purposes only and does not constitute Agency policy. C-78 DRAFT--DO NOT CITE OR QUOTE 1 LIRMLUM = KABS*MST*B 2 LIMLUM = INTEG(LIRMLUM,0.0) 3 4 IPERCENT OF DOSE REMAINING IN THE GI TRACT 5 PRCT_remain_GIT = (MST/(MSTT+1E-30))*100 6 7 RFECES = KST*MST + REXCLI 8 FECES = INTEG(RFECES,0.0) 9 prctFECES = (FECES/(BDOSE_TOTAL+1E-30))*100 10 11 12 IABSORPTION OF DIOXIN BY IV ROUTE--------- 13 IVR= IV/PFUNC I RATE FOR IV INFUSION IN BLOOD 14 EXPIV= IVR * (1.0-STEP(PFUNC)) 15 IVDOSE = integ(EXPIV,0.0) 16 17 ISYSTEMIC BLOOD CONCENTRATION (NMOL/ML) 18 I MODIFICATION ON OCTOBER 6, 2009 19 CB=(QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM)/(QC+CLURI) I 2 0 CA = CB 21 2 2 IURINARY EXCRETION BY KIDNEY 2 3 I MODIFICATION ON OCTOBER 6, 2009 2 4 RAURI = CLURI *CB 2 5 AURI = INTEG(RAURI,0.0) 26 2 7 prctAURI = (AURI/(BDOSE_TOTAL+1E-30))*100 28 29 3 0 IUNIT CONVERSION POST SIMULATION 31 PRCT_B = (CB/(MSTT+1E-30))*100 I PERCENT OF DOSE/G TISSUE 3 2 CBNGKG=CB*MW*UNITCORR 33 CBSNGKGLIADJ= (CB*MW*UNITCORR*(1.0/B TOTLIP)*(1.0/SERBLO))I[NG of TCDD 3 4 Serum/Kg OF LIPIP] 35 CBPMOL_KG= CB*UNITCORR*UNITCORR ICONCENTRATION IN PMOL/KG 3 6 CBNGG = CB*MW 3 7 IADIPOSE TISSUE COMPARTMENT 3 8 ITISSUE BLOOD SUBCOMPARTMENT 3 9 RAFB = QF*(CA-CFB)-PAF*(CFB-CF/PF) I(NMOL/HR) 4 0 AFB = INTEG(RAFB,0.0) I(NMOL) 41 CFB = AFB/WFB I(NMOL/ML) 4 2 ITISSUE SUBCOMPARTMENT 4 3 RAF = PAF*(CFB-CF/PF) I(NMOL/HR) 4 4 AF = INTEG(RAF,0.0) I(NMOL) 4 5 CF = AF/WF I(NMOL/ML) 46 4 7 IPOST SIMULATION UNIT CONVERSION 4 8 CFTOTAL = (AF + AFB)/(WF + WFB) I TOTAL CONCENTRATION IN FAT(NM/ML) 4 9 PRCT_F = (CFTOTAL/(MSTT+1E-30))*100 I PERCENT OF DOSE IN FAT 5 0 CFNGKG = CFTOTAL*MW*UNITCORR 51 CFUGG=(CFTOTAL*MW)/UNITCORR 5 2 CFPMOL_KG= CFTOTAL*UNITCORR*UNITCORR ICONCENTRATION IN PMOL/KG 53 CFNGG = CFTOTAL*MW 54 55 IREST OF THE BODY COMPARTMENT 5 6 ITISSUE BLOOD SUBCOMPARTMENT This document is a draftfor review purposes only and does not constitute Agency policy. C-79 DRAFT--DO NOT CITE OR QUOTE 1 RAREB= QRE*(CA-CREB)-PARE*(CREB-CRE/PRE) I(NMOL/HR) 2 AREB = INTEG(RAREB,0.0) I(NMOL) 3 CREB = AREB/WREB I(NMOL/ML) 4 ITISSUE SUBCOMPARTMENT 5 RARE = PARE*(CREB - CRE/PRE) I(NMOL/HR) 6 ARE = INTEG(RARE,0.0) I(NMOL) 7 CRE = ARE/WRE I(NMOL/ML) 8 9 IPOST SIMULATION UNIT CONVERSION 10 CRETOTAL= (ARE + AREB)/(WRE + WREB) I CONCENTRATION AT STEADY 11 STATE 12 PRCT RE = (CRETOTAL/(MSTT+1E-30))*100 13 14 15 ILIVER COMPARTMENT 16 ITISSUE BLOOD SUBCOMPARTMENT 17 RALIB = QLI*(CA-CLIB)-PALI*(CLIB-CFLLIR)+LIRMLUM I(NMOL/HR) 18 ALIB = INTeg(RALIB,0.0) I(NMOL) 19 CLIB = ALIB/WLIB 2 0 ITISSUE SUBCOMPARTMENT 21 RALI = PALI*(CLIB-CFLLIR)-REXCLI I(NMOL/HR) 2 2 ALI = integ(RALI,0.0) I(NMOL) 2 3 CLI = ALI/WLI I(NMOL/ML) 24 2 5 IFREE TCCD CONCENTRATION IN LIVER (NMOL/ML) 2 6 PARAMETER (LIVER_1RMN = 1.0E-30) 2 7 CFLLI= IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLI & 2 8 +LIVER_1RMN))+((CYP1A2_1O3*CFLLIR/(KDLI2+CFLLIR & 2 9 +LIVER_1RMN)*PAS_INDUC)))-CFLLI,CFLLI0) 3 0 CFLLIR=DIM(CFLLI,0.0) I FREE CONCENTRATION IN LIVER 31 3 2 CBNDLI= LIBMAX*CFLLIR/(KDLI+CFLLIR+LIVER 1RMN) IBOUND CONCENTRATION 33 3 4 IPOST SIMULATION UNIT CONVERSION 3 5 CLITOTAL= (ALI + ALIB)/(WLI + WLIB)I 3 6 PRCT_LI = (CLITOTAL/(MSTT+1E-30))*100 I PERCENT OF DOSE IN LIVER 3 7 rec_occ_AHR= (CFLLIR/(KDLI+CFLLIR+1E-30))*100.0 I PERCENT OF AhR OCCUPANCY 3 8 PROT_occ_1A2= (CFLLIR/(KDLI2+CFLLIR))*100.0 I PERCENT OF 1A2 OCCUPANCY 3 9 CLINGKG =(CLITOTAL*MW*UNITCORR) 4 0 CBNDLINGKG = CBNDLI*MW*UNITCORR 41 CLIUGG=(CLITOTAL*MW)/UNITCORR 4 2 CLIPMOL_KG= CLITOTAL*UNITCORR*UNITCORR ICONCENTRATION IN PMOL/KG 43 CLINGG = CLITOTAL*MW 44 4 5 IFraction increase of induction of CYP1A2 4 6 fold_ind=(CYP1A2_1OUT/CYP1A2_1A2) 4 7 VARIATIONOfAC =(CYP1A2_1OUT-CYP1A2_1A2)/CYP1A2_1A2 48 4 9 IVARIABLE ELIMINATION BASED ON THE CYP1A2 5 0 KBILE_LI_T =((CYP1A2_1OUT-CYP1A2_1A2)/CYP1A2_1A2)*Kelv IINDUCED BILIARY 51 EXCRETION RATE CONSTANT 52 53 REXCLI= (KBILE_LI_T*CFLLIR*WLI) IDOSE-DEPENDENT EXCRETION RATE 5 4 EXCLI = INTEG(REXCLI,0.0) 55 5 6 ICHEMICAL IN CYP450 (1A2) COMPARTMENT This document is a draftfor review purposes only and does not constitute Agency policy. C-80 DRAFT--DO NOT CITE OR QUOTE 1 IEQUATION FOR INDUCTION OF CYP1A2 2 3 CYP1A2_1KINP = CYP1A2_1KOUT* CYP1A2_1OUTZ 4 5 I MODIFICATION ON OCTOBER 6, 2009 6 CYP1A2_1OUT =INTEG(CYP1A2_1KINP * (1.0 + CYP1A2_1EMAX *(CBNDLI+1.0e-30)**HILL 7& 8 /(CYP1A2_1EC50**HILL + (CBNDLI+1.0e-30)**HILL)) & 9 - CYP1A2_1KOUT*CYP1A2_1OUT, CYP1A2_1OUTZ) 10 I EQUATIONS INCORPORATING DELAY OF CYP1A2 PRODUCTION (NOT USED IN 11 SIMULATIONS) 12 13 CYP1A2_1RO2 = (CYP1A2_1OUT - CYP1A2_1O2)/ CYP1A2_1TAU 14 CYP1A2_1O2 =INTEG(CYP1A2_1RO2, CYP1A2_1A1) 15 CYP1A2_1RO3 = (CYP1A2_1O2 - CYP1A2_1O3)/ CYP1A2_1TAU 16 CYP1A2_1O3 =INTEG(CYP1A2_1RO3, CYP1A2_1A2) 17 18 I MASS BALANCE CONTROL 19 BDOSE= LYMLUM+LIMLUM+IVDOSE 2 0 BMASSE = EXCLI+AURI+AFB+AF+AREB+ARE+ALIB+ALI 21 BDIFF = BDOSE-BMASSE 2 2 I AMOUNT TOTAL PRESENT IN THE GI TRACT 2 3 BDOSE TOTAL =LYMLUM+LIMLUM+FECES 24 2 5 IBODY BURDEN IN NG 2 6 Body burden =(AFB+AF+AREB+ARE+ALIB+ALI)*MW 27 2 8 !BODY BURDEN CONCENTRATION (NG/KG) 2 9 BBNGKG =(((AFB+AF+AREB+ARE+ALIB+ALI)*MW)/(WT0/UNITCORR)) ! 30 31 !COMMAND FOR END OF SIMULATION 3 2 TERMT (T.GE. TimeLimit, 'Time limit has been reached.') 33 3 4 END END OF THE DERIVATIVE SECTION 3 5 END END OF THE DYNAMIC SECTION 3 6 END END OF PROGRAM 37 38 C .2.5.2. Input Files 3 9 C .2.5.2.1. Della Porta (1987) (female) 4 0 output @clear 41 prepare @clear 4 2 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 43 4 4 % Della Porta 1987 for female mice. 4 5 %dose levels: 2.5 and 5 ug/kg/week for 52 weeks 4 6 %dose levels: 2500 and 5000 ng/kg/week for 52 weeks 4 7 %dose levels equivalent to: 357 and 714 ng/kg/d 48 4 9 MAXT = 0.01 5 0 CINT = 0.1 51 EXP_TIME_ON 0. %delay before begin exposure (HOUR) 5 2 EXP_TIME_OFF 8736 %TIME EXPOSURE STOP (HOUR) 53 DAY_CYCLE 168 5 4 BCK TIME ON 0. %DELAY BEFORE BACGROUND EXPOSURE (HOUR) This document is a draftfor review purposes only and does not constitute Agency policy. C-81 DRAFT--DO NOT CITE OR QUOTE 1 BCK_TIME_OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 2 TIMELIMIT = 8736 %SIMULATION LIMIT TIME (HOUR) 3 BW T0 = 20 % Body weight at the beginning of the simulation 4 (g); corresponds to 6 weeks of age and taken from Figure 3 5 6 7 %EXPOSURE DOSE SCENARIOS (UG/KG) 8 %MSTOT = 2.5 % exposure dose ug/kg 9 MSTOT = 5.0 % exposure dose ug/kg 10 11 C .2.5.2.2. Della Porta (1987) (male) 12 output @clear 13 prepare @clear 14 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 15 16 % Della Porta 1987 for male mice. 17 %dose levels: 2.5 and 5 ug/kg/week for 52 weeks 18 %dose levels: 2500 and 5000 ng/kg/week for 52 weeks 19 %dose levels equivalent to: 357 and 714 ng/kg/d 20 21 MAXT = 0.01 2 2 CINT = 0.1 23 EXP_TIME_ON 0. delay before begin exposure (HOUR) 2 4 EXP_TIME_OFF 8736 TIME EXPOSURE STOP (HOUR) 2 5 DAY_CYCLE 168 2 6 BCK_TIME_ON 0. iDELAY BEFORE BACGROUND EXPOSURE (HOUR) 2 7 BCK_TIME_OFF 0. TIME OF BACKGROUND EXPOSURE STOP (HOUR) 2 8 TIMELIMIT 8736 SIMULATION LIMIT TIME (HOUR) 2 9 BW_T0 26 i Body weight at the beginning of the simulation 3 0 (g); corresponds to 6 weeks of age and taken from Figure 3 31 32 33 %EXPOSURE DOSE SCENARIOS (UG/KG) 3 4 %MSTOT = 2.5 % exposure dose ug/kg 35 MSTOT = 5.0 % exposure dose ug/kg 36 3 7 C .2.5.2.3. N TP (1982) (female) (chronic) 3 8 %RAT2.m 3 9 %clear variable 4 0 output @clear 41 prepare @clear 4 2 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 4 3 %output @nciout=168 T SUMEXPEVENT 44 4 5 % NTP 1982. 4 6 %built and check in September 20, 2009 4 7 %protocol: twice weekly gavage for 104 weeks 4 8 %Rat Dioxin 3C June09 2clean 2.csl 4 9 %MICE_NON_GESTAT_ICF_F083109.csl 5 0 %MICE_NON_GESTAT_ICF_F092009.csl (now 09-20-09) 51 %dose levels: 0.02, 0.1, 1 ug/kg/biweekly, ug/kg for 104 weeks 5 2 %dose levels: 20, 100, 1000 ng/kg/biweekly,ng/kg for 104 weeks 53 %dose levels equivalent to: 5.71, 28.57, 285.1 ng/kg/d 54 This document is a draftfor review purposes only and does not constitute Agency policy. C-82 DRAFT--DO NOT CITE OR QUOTE 1 MAXT = 0.01 2 CINT = 0.1 3 EXP TIME ON = 0. %delay before begin exposure (HOUR) 4 EXP TIME OFF = 17472 %TIME EXPOSURE STOP (HOUR) 5 DAY CYCLE = 84 6 BCK TIME ON = 0. %DELAY BEFORE BACGROUND EXPOSURE (HOUR) 7 BCK TIME OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 8 TIMELIMIT = 17472 %SIMULATION LIMIT TIME (HOUR) 9 BW T0 = 23 % Body weight at the beginning of the simulati 10 (g) 11 12 13 %EXPOSURE DOSE SCENARIOS (UG/KG) 14 %MSTOT = 0.02 % exposure dose ug/kg 15 %MSTOT = 0.1 % exposure dose ug/kg 16 MSTOT = 1.0 % exposure dose ug/kg 17 18 C .2.5.2.4. N TP (1982) (male) (chronic). 19 %RAT2.m 2 0 %clear variable 21 output @clear 2 2 prepare @clear 2 3 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 2 4 %output @nciout=168 T SUMEXPEVENT 25 2 6 % NTP 1982. 2 7 %built and check in September 20, 2009 2 8 %protocol: twice weekly gavage for 104 weeks 2 9 %Rat Dioxin 3C June09 2clean 2.csl 3 0 %MICE_NON_GESTAT_ICF_F0831097csl 31 %MICE_NON_GESTAT_ICF_F092009.csl (now 09-20-09) 3 2 %dose levels: 0.005, 0.025, 0.25 ug/kg/biweekly, ug/kg for 104 weeks 33 %dose levels: 5, 25, 250 ng/kg/biweekly,ng/kg for 104 weeks 3 4 %dose levels equivalent to: 1.4, 7.1, 71 ng/kg/d 35 3 6 MAXT = 0.01 3 7 CINT = 0.1 3 8 EXP TIME ON = 0. %delay before begin exposure (HOUR) 3 9 EXP TIME OFF = 17472 %TIME EXPOSURE STOP (HOUR) 4 0 DAY CYCLE = 84 41 BCK TIME ON = 0. %DELAY BEFORE BACGROUND EXPOSURE (HOUR) 4 2 BCK TIME OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 43 TIMELIMIT = 17472 %SIMULATION LIMIT TIME (HOUR) 4 4 BW T0 = 25 % Body weight at the beginning of the simulation 4 5 (g) 46 47 4 8 %EXPOSURE DOSE SCENARIOS (UG/KG) 4 9 %MSTOT = 0.005 % exposure dose ug/kg 5 0 %MSTOT = 0.025 % exposure dose ug/kg 51 MSTOT = 0.25 % exposure dose ug/kg 52 53 C .2.5.2.5. Smialowicz et al. (2008). 5 4 output @clear This document is a draftfor review purposes only and does not constitute Agency policy. C-83 DRAFT--DO NOT CITE OR QUOTE 1 prepare @clear 2 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 3 4 % Smialowicz et al. 2008. 5 %built and check in August 7 2009 6 %protocol: oral gavage 5 days/week for 13 weeks 7 %Mice Dioxin 3C June09 l.csl 8 %MICE_NON_GESTAT_ICF_F083109.csl (now 09-11-09) 9 %dose levels: 0, 0.0015, 0.015, 0.15, 0.45 ug/kg 10 %dose levels: 0, 1.5, 15, 150, 450 nkd (0, 1.07, 10.7, 107, 321 nkd adj) 11 12 MAXT = 0.01 13 CINT = 0.1 14 TIMELIMIT = 2184 %SIMULATION LIMIT TIME (HOUR) 15 EXP_TIME_ON = 0. %delay before begin exposure (HOUR) 16 EXP_TIME_OFF = 2184 %TIME EXPOSURE STOP (HOUR) 17 DAY_CYCLE = 24 18 WEEK_PERIOD = 168 19 WEEK_FINISH = 119 2 0 BCK_TIME_ON = 0. %DELAY BEFORE BACGROUND EXPOSURE (HOUR) 21 BCK_TIME_OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) 2 2 BW_T0 = 28 % Body weight at the beginning of the simulation 2 3 (g) 24 2 5 %EXPOSURE DOSE SCENARIOS (UG/KG) 2 6 %MSTOT = 0.0015 % exposure dose (ug/kg) 2 7 %MSTOT = 0.015 % exposure dose (ug/kg) 2 8 %MSTOT = 0.150 % exposure dose (ug/kg) 2 9 MSTOT = 0.450 % exposure dose (ug/kg) 30 31 C .2.5.2.6. Toth et al. (1979) (1 year). 3 2 output @clear 33 prepare @clear 3 4 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG CBNGKG 35 3 6 % Toth et al. 1979 3 7 %built and check in August 7 2009 3 8 %protocol: weekly gavage for 1 year 3 9 %Mice Dioxin 3C June09 1.csl 4 0 %MICE_NON_GESTAT_ICF_F083109.csl (now 09-11-09) 41 %dose levels: 7, 700, 7000 ng/kg 1/week for 52 weeks (1 year) 4 2 %dose levels: 0.007, 0.7, 7 ug/kg 1/week for 52 weeks (1 year) 4 3 %dose equivalent: 1, 100, 1000 ng/kg/day 44 4 5 MAXT = 0.01 4 6 CINT = 0.1 4 7 TIMELIMIT = 8760 4 8 EXP_TIME_ON = 0. %delay before begin exposure (HOUR) 4 9 EXP_TIME_OFF = 8760 %2208 %TIME EXPOSURE STOP (HOUR) 5 0 DAY_CYCLE = 168 51 WEEK_PERIOD = 8760 5 2 WEEK_FINISH = 8760 53 BCK_TIME_ON = 0. %DELAY BEFORE BACGROUND EXPOSURE (HOUR) 5 4 BCK TIME OFF = 0. %TIME OF BACKGROUND EXPOSURE STOP (HOUR) This document is a draftfor review purposes only and does not constitute Agency policy. C-84 DRAFT--DO NOT CITE OR QUOTE 1 BW T0 = 27 % Body weight at the beginning of the simulation 2 (g) 3 4 5 %EXPOSURE DOSE SCENARIOS (UG/KG) 6 %MSTOT = 0.007 % exposure dose (ug/kg) 7 %MSTOT = 0.7 % exposure dose (ug/kg) 8 MSTOT = 7 % exposure dose (ug/kg) 9 10 C .2.5.2.7. White et al. (1986). 11 output @clear 12 prepare @clear 13 prepare T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CBNDLINGKG 14 15 % White et al 1986 16 %built and check in August 7 2009 17 18 %protocol: oral exposure single dose 19 %dose levels: 0.714, 3.57, 7.14, 35.71, 71.43, 142.86 ng /kg/d ug/kg 1/day 2 0 for 14 consecutive days 21 %dose have been modified following Jeff email on Friday August 21 2009 2 2 %dose levels: 10, 50, 100, 500, 1000, 2000 ng /kg/d ug/kg 1/day for 14 2 3 consecutive days 2 4 %dose levels: 0.010, 0.050, 0.100, 0.500, 1.0, 2.0 ug /kg/d ug/kg 1/day for 2 5 14 consecutive days 26 2 7 MAXT = 0.01 2 8 CINT = 0.1 2 9 TIMELIMIT = 336 3 0 EXP_TIME_ON = 0. %TIME AT WHICH EXPOSURE BEGINS (HOUR) 31 EXP_TIME_OFF = 336 %TIME AT WHICH EXPOSURE ENDS (HOUR) 3 2 DAY_CYCLE = 24 33 WEEK_PERIOD = 336 3 4 WEEK_FINISH = 336 3 5 BCK_TIME_ON = 0. %TIME AT WHICH BACKGROUND EXPOSURE BEGINS (HOUR) 3 6 BCK_TIME_OFF = 0. %TIME AT WHICH BACKGROUND EXPOSURE ENDS (HOUR) 3 7 BW T0 = 23 % BODY WEIGHT AT THE BEGINNING OF THE SIMULATION (G) 38 3 9 )EXPOSURE DOSE SCENARIOS (UG/KG) 4 0 %MSTOT = 0.010 % EXPOSURE DOSE IN UG/KG 41 %MSTOT = 0.050 ) EXPOSURE DOSE IN UG/KG 4 2 %MSTOT = 0.100 ) EXPOSURE DOSE IN UG/KG 4 3 %MSTOT = 0.500 ) EXPOSURE DOSE IN UG/KG 4 4 %MSTOT = 1 ) EXPOSURE DOSE IN UG/KG 4 5 MSTOT = 2 EXPOSURE DOSE IN UG/KG 46 47 48 C.2.6. M ouse G estational M odel 4 9 C.2.6.1. Model Code 50 P R O G R A M : 'T h re e C o m p a rtm e n t P B P K M o d e l fo r T C D D in M ic e (G e sta tio n )' 51 5 2 ! Parameters were change may 16, 2002 53 ! Come from {8MAI_CHR_PRE-EXP_GD} This document is a draftfor review purposes only and does not constitute Agency policy. C -85 D R A FT -- D O N O T C IT E O R Q U O T E 1 ! Come from {12 Mouse GDjfile 2 !******************************************** 3 !{{IMPORTANT-IMPORTANT-IMPORTANT-IMPORTANT}} 4 ! REDUCTION OF MOTHER AND FETUS COMPARTMENT 5 ! 2M_R_TCDD_JULY2002 ////(JULY 18,2002)//// 6 !TCDD_RED_4Species_2003_4 ////(APR 8 ,2003)//// 7 !TCDD_RED_4Species_2003_9 ////(APR 17 ,2003)//// 8 !TCDD_RED_4Species_2003_12 ////(APR 17 ,2003)//// 9 !***************************************************** 10 !APRIL 18 2003 11 !TCDD_4C_4SP_2003 ////(APR 18 ,2003)//// 12 ! was ''Gest 4 species 1.csl'' but update July 2009 13 14 !DevTCDD4Species ICF afterKKfix v3 ratgest.csl 15 !MICE_GESTATIONAL_ICF_F092309.csl _ 16 !MICE_GESTATIONAL_ICF_F100609.csl 17 !***************************************************** 18 19 !Legend/Legend/Legend/Legend/Legend/Legend/Legend/Legend/ 2 0 ILegend for this PBPK model 21 !Mating: control the tenure of exchange between fetus and 2 2 IMother and also control imitated tissue growth 2 3 ICtrl: WTFE, WFO, WPLA0, QPLAF,WT0 2 4 !(for rat, mouse, human, and monkey) 2 5 IControl transfer from mother to fetus and fetus to mother by TRANSTIME ON 2 6 !SWITCH_trans = 0 NO TRANSFER 2 7 !SWITCH_trans = 1 TRANSFER OCCURS 2 8 !Gest off = 1 2 9 !Gest on= 0. 3 0 ! These switches are also controlled by mating parameters 31 3 2 INITIAL ! 33 3 4 !SIMULATION PARAMETERS ==== 3 5 CONSTANT PARA_ZERO = 1E-30 3 6 CONSTANT EXP_TIME_ON = 288. ! TIME AT WHICH EXPOSURE BEGINS (HOURS) 3 7 CONSTANT EXP_TIME_OFF = 504 ! TIME AT WHICH EXPOSURE ENDS (HOURS) 3 8 CONSTANT DAY_CYCLE = 504. ! NUMBER OF HOURS BETWEEN DOSES (HOURS) 3 9 CONSTANT BCK_TIME_ON = 0.0 ! TIME AT WHICH BACKGROUND EXPOSURE 4 0 BEGINS (HOURS) 41 CONSTANT BCK_TIME_OFF = 0.0 ! TIME AT WHICH BACKGROUND EXPOSURE ENDS 4 2 (HOURS) 43 CONSTANT TRANSTIME_ON = 144 !CONTROL TRANSFER FROM MOTHER TO FETUS 4 4 AT GESTATIONAL DAY 6 45 4 6 !UNIT CONVERSION 4 7 CONSTANT MW=322 ! MOLECULAR WEIGHT (NG/NMOL) 4 8 CONSTANT SERBLO = 0.55 4 9 CONSTANT UNITCORR = 1000 50 51 !INTRAVENOUS SEQUENCY 5 2 constant IV_LACK = 0.0 53 constant IV_PERIOD = 0.0 54 55 !PREGNANCY PARAMETER ==== 5 6 CONSTANT MATTING = 0.0 !BEGINNING OF MATING (HOUR) This document is a draftfor review purposes only and does not constitute Agency policy. C-86 DRAFT--DO NOT CITE OR QUOTE 1 CONSTANT N FETUS 10 INUMBER OF FETUS PRESENT 2 3 ICONSTANT EXPOSURE CONTROL =========== 4 !ACUTE, SUBCHRONIC, CHRONIC EXPOSURE ===== 5 IOR BACKGROUND EXPOSURE (IN THIS CASE 3 TIMES A DAY)=== 6 CONSTANT MSTOTBCKGR = 0.0 ! ORAL BACKGROUND EXPOSURE DOSE (UG/KG) 7 CONSTANT MSTOT = 0.0 ! ORAL EXPOSURE DOSE (UG/KG) 8 9 IORAL ABSORPTION 10 MSTOT NM = MSTOT/MW ICONVERTS THE DOSE TO NMOL/G 11 12 ! INTRAVENOUS ABSORPTION 13 CONSTANT DOSEIV = 0.0 ! INJECTED DOSE (UG/KG) 14 DOSEIV_NM = DOSEIV/MW ! CONVERTS THE INJECTED DOSE TO NMOL/G 15 CONSTANT DOSEIVLATE = 0.0 ! INJECTED DOSE LATE (UG/KG) 16 DOSEIVNMlate = DOSEIVLATE/MW !AMOUNT IN NMOL/G 17 18 !INITIAL GUESS OF THE FREE CONCENTRATION IN THE LIGAND (COMPARTMENT 19 INDICATED BELOW)==== 2 0 CONSTANT CFLLI0 = 0.0 !LIVER (NMOL/ML) 21 CONSTANT CFLPLA0 = 0.0 !PLACENTA (NMOL/ML) 22 2 3 !BINDING CAPACITY (AhR) FOR NON LINEAR BINDING (COMPARTMENT INDICATED 2 4 BELOW) (NMOL/ML) === 2 5 CONSTANT LIBMAX = 3.5E-4 ! LIVER (NMOL/ML), WANG ET AL. 1997 2 6 CONSTANT PLABMAX = 2.0E-4 !TEMPORARY PARAMETER 27 2 8 ! PROTEIN AFFINITY CONSTANTS (1A2 OR AhR, COMPARTMENT INDICATED BELOW) 2 9 (NMOL/ML)=== 3 0 CONSTANT KDLI = 1.0E-4 !LIVER (AhR) (NMOL/ML), WANG ET AL. 1997 31 CONSTANT KDLI2 = 4.0E-2 !LIVER (1A2) (NMOL/ML), EMOND ET AL. 2004 3 2 CONSTANT KDPLA = 1.0E-4 !TEMPORARY PARAMETER (AhR) 33 3 4 !EXCRETION AND ABSORPTION CONSTANT 3 5 CONSTANT KST = 0.3 ! GASTRIC RATE CONSTANT (HR-1) 3 6 CONSTANT KABS = 0.48 !INTESTINAL ABSORPTION CONSTANT (HR-1) ), 3 7 WANG ET AL. 1997 38 3 9 ! ELIMINATION CONSTANTS 4 0 CONSTANT CLURI = 0.09 ! URINARY CLEARANCE (ML/HR) 41 4 2 !TEST ELIMINATION VARIABLE 4 3 constant kelv = 0.4 ! INTERSPECIES VARIABLE ELIMINATION 4 4 CONSTANT (1/HOUR) 45 4 6 ! CONSTANT TO DIVIDE THE ABSORPTION INTO LYMPHATIC AND PORTAL FRACTIONS 4 7 CONSTANT A = 0,.7 ! LYMPHATIC FRACTION, WANG ET AL. 1997 48 4 9 !PARTITION COEFFICIENTS 5 0 CONSTANT PF = 400 ! ADIPOSE TISSUE/BLOOD 51 CONSTANT PRE = 3 ! REST OF THE BODY/BLOOD, WANG ET AL. 2000 5 2 CONSTANT PLI = 6 ! LIVER/BLOOD, WANG ET AL. 1997 53 CONSTANT PPLA = 3 ! TEMPORARY PARAMETER NOT CONFIGURED 54 55 !PARAMETER FOR INDUCTION OF CYP 1A2, WANG ET AL. 1997 OR OPTIMIZED 5 6 CONSTANT PAS INDUC =1 ! INCLUDE INDUCTION? (1 = YES, 0 = NO) This document is a draftfor review purposes only and does not constitute Agency policy. C-87 DRAFT--DO NOT CITE OR QUOTE 1 CONSTANT CYP1A2_1OUTZ = 1.6 ! DEGRADATION CONCENTRATION CONSTANT OF 2 1A2 (NMOL/ML) (OPTIMIZED) 3 CONSTANT CYP1A2_1A1 = 1.5 ! BASAL CONCENTRATION OF 1A1 (NMOL/ML), 4 WANG ET AL . (2000) 5 CONSTANT CYP1A2_1EC50 = 0.13 ! DISSOCIATION CONSTANT TCDD-CYP1A2 6 (NMOL/ML) 7 CONSTANT CYP1A2 1A2 = 1.5 !BASAL CONCENTRATION OF 1A2 8 (NMOL/ML),WANG ET AL. (2000) 9 CONSTANT CYP1A2_1KOUT = 0.1 ! FIRST ORDER RATE OF DEGRADATION (H-1) 10 CONSTANT CYP1A2_1TAU = 1.5 !HOLDING TIME (H) (OPTIMIZED), WANG ET AL 11 . (2000) 12 CONSTANT CYP1A2_1EMAX = 600 ! MAXIMUM INDUCTION OVER BASAL EFFECT 13 (UNITLESS) 14 CONSTANT HILL = 0.6 !HILL CONSTANT; COOPERATIVELY LIGAND 15 BINDING EFFECT CONSTANT (UNITLESS) 16 17 IDIFFUSIONAL PERMEABILITY FRACTION, WANG ET AL. 1997 18 CONSTANT PAFF = 0.12 (ADIPOSE (UNITLESS) OPTIMIZED, WANG ET AL. 19 2000 2 0 CONSTANT PAREF = 0.03 IREST OF THE BODY (UNITLESS) 21 CONSTANT PALIF = 0.35 ILIVER (UNITLESS) 2 2 CONSTANT PAPLAF = 0.03 (TEMPORARY PARAMETER NOT CONFIGURED 23 2 4 (FRACTION OF TISSUE WEIGHT ==== 2 5 CONSTANT WLI0 = 0.0549 (LIVER ILSI (1994) 26 2 7 (TISSUE BLOOD FLOW EXPRESSED AS A FRACTION OF CARDIAC OUTPUT CONSTANT QFF 2 8 = 0.070 ( ADIPOSE TISSUE BLOOD FLOW FRACTION (UNITLESS), LEUNG ET AL. 1990 2 9 CONSTANT QLIF = 0.161 (LIVER (UNITLESS), ILSI 1994 30 31 (COMPARTMENT TISSUE BLOOD EXPRESSED AS A FRACTION OF THE TOTAL COMPARTMENT 3 2 VOLUME 33 CONSTANT WFB0 = 0.050 (ADIPOSE TISSUE, WANG ET AL. 1997 3 4 CONSTANT WREB0 = 0.030 (REST OF THE BODY, WANG ET AL. 1997 35 CONSTANT WLIB0 = 0.266 (LIVER, WANG ET AL. 1997 3 6 CONSTANT WPLAB0 = 0.500 (TEMPORARY PARAMETER NOT CONFIGURED 37 3 8 (EXPOSURE SCENARIO FOR UNIQUE OR REPETITIVE WEEKLY OR MONTHLY EXPOSURE 3 9 (NUMBER OF EXPOSURES PER WEEK 4 0 CONSTANT WEEK LACK = 0.0 (DELAY BEFORE EXPOSURE ENDS (WEEK) 41 CONSTANT WEEK PERIOD = 168 ( NUMBER OF HOURS IN THE WEEK (HOURS) 4 2 CONSTANT WEEK FINISH = 168 ( TIME EXPOSURE ENDS (HOURS) 43 4 4 (NUMBER OF EXPOSURES PER MONTH 4 5 CONSTANT MONTH_LACK = 0.0 (DELAY BEFORE EXPOSURE BEGINS (MONTH) 46 4 7 (CONSTANT FOR BACKGROUND EXPOSURE 4 8 CONSTANT Day_LACK_BG = 0.0 ( DELAY BEFORE EXPOSURE BEGINS (HOUR) 4 9 CONSTANT Day_PERIOD_BG = 24 (LENGTH OF EXPOSURE (HOUR) 50 51 (NUMBER OF EXPOSURES PER WEEK 5 2 CONSTANT WEEK_LACK_BG = 0.0 (DELAY BEFORE BACKGROUD EXPOSURE (WEEK) 53 CONSTANT WEEK_PERIOD_BG = 168 ( NUMBER OF HOURS IN THE WEEK (HOURS) 5 4 CONSTANT WEEK FINISH BG = 168 (TIME EXPOSURE ENDS (HOURS) 55 5 6 (INITIAL BODY WEIGHT This document is a draftfor review purposes only and does not constitute Agency policy. C-88 DRAFT--DO NOT CITE OR QUOTE 1 CONSTANT BW_T0 = 30 ! WANG ET AL. 1997 2 CONSTANT RATIO_RATF_MOUSEF = 0.2 !RATIO OF FETUS MOUSE/RAT AT 3 GESTATIONAL DAY 22 4 ! FOR RAT (1) AND FOR MOUSE (0.2) 5 6 !COMPARTMENT LIPID EXPRESSED AS THE FRACTION OF TOTAL LIPID, POULIN ET AL. 7 2000 8 CONSTANT F TOTLIP = 0.855 ! ADIPOSE TISSUE (UNITLESS) 9 CONSTANT B TOTLIP = 0.0033 ! BLOOD (UNITLESS) 10 CONSTANT RE TOTLIP = 0.019 ! REST OF THE BODY 11 (UNITLESS) 12 CONSTANT LI TOTLIP = 0.060 ! LIVER (UNITLESS) 13 CONSTANT PLA TOTLIP 14 CONSTANT FETUS TOTLIP = 0.019 = 0.019 ! PLACENTA (UNITLESS) ! FETUS (UNITLESS) 15 16 END ! END OF THE INITIAL SECTION 17 18 DYNAMIC ! DYNAMIC SIMULATION SECTION 19 ALGORITHM IALG = 2 2 0 CINTERVAL CINT = 0.1 21 MAXTERVAL MAXT = 1.0e+10 2 2 MINTERVAL MINT = 1.0E-10 ! GEAR METHOD ! COMMUNICATION INTERVAL ! MAXIMUM CALCULATION INTERVAL ! MINIMUM CALCULATION INTERVAL o o 23 VARIABLE T 2 4 CONSTANT TIMELIMIT = = 313 !SIMULATION LIMIT TIME (HOUR) 2 5 CINTXY = CINT 2 6 PFUNC = CINT 27 2 8 !TIME CONVERSION 2 9 DAY = T/24 ! TIME IN DAYS 3 0 WEEK = T/168 ! TIME IN WEEKS 31 MONTH = T/730 ! TIME IN MONTHS 3 2 YEAR = T/8760 ! TIME IN YEARS 33 3 4 DERIVATIVE ! PORTION OF CODE THAT SOLVES DIFFERENTIAL EQUATIONS 35 3 6 !CHRONIC OR SUBCHRONIC EXPOSURE SCENARIO 3 7 !NUMBER OF EXPOSURES PER DAY 3 8 DAY_LACK = EXP_TIME_ON DELAY BEFORE EXPOSURE BEGINS (HOURS) 3 9 DAY_PERIOD = DAY_CYCLE EXPOSURE PERIOD (HOURS) 4 0 DAY_FINISH = CINTXY LENGTH OF EXPOSURE (HOURS) 41 MONTH_PERIOD = TIMELIMIT EXPOSURE PERIOD (MONTHS) 4 2 MONTH FINISH = EXP TIME OFF LENGTH OF EXPOSURE (MONTHS) 43 4 4 !NUMBER OF EXPOSURES PER DAY AND MONTH 4 5 DAY_FINISH_BG = CINTXY 4 6 MONTH_LACK_BG = BCK_TIME_ON !DELAY BEFORE BACKGROUD EXPOSURE BEGINS 4 7 (MONTHS) 4 8 MONTH_PERIOD_BG = TIMELIMIT (BACKGROUND EXPOSURE PERIOD (MONTHS) 4 9 MONTH FINISH BG = BCK TIME OFF !LENGTH OF BACKGROUND EXPOSURE (MONTHS) 50 51 !INTRAVENOUS LATE 5 2 IV_FINISH = CINTXY 53 B = 1-A ! FRACTION OF DIOXIN ABSORBED IN THE PORTAL FRACTION OF THE LIVER 54 This document is a draftfor review purposes only and does not constitute Agency policy. C-89 DRAFT--DO NOT CITE OR QUOTE 1 2 !FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUME,FETUS,VOLUM 3E 4 ! FROM OFLAHERTY_1992 5 6 RTESTGEST= T-MATTING 7 TESTGEST=DIM(RTESTGEST,0.0) 8 9 WTFER_RODENT= (2.3d-3*EXP(1.49d-2*(TESTGEST))+1.3d-2)*Gest_on 10 WTFER = (WTFER_RODENT*RATIO_RATF_MOUSEF*N_FETUS) 11 WTFE = DIM(WTFER,0.0) 12 13 ! 14 FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME,FAT,VOLUME 15 ! FAT GROWTH EXPRESSION LINEAR DURING PREGNANCY 16 ! FROM O'FLAHERTY_1992 17 18 WF0= (((9.66d-5*(TESTGEST))*gest_on)+0.069) 19 2 0 ! PLACENTA,VOLUME, PLACENTA,VOLUME, PLACENTA,VOLUME, PLACENTA,VOLUME 21 ! WPLA PLACENTA GROWTH EXPRESSION, SINGLE EXPONENTIAL WITH OFFSET 2 2 ! FROM O'FLAHERTY_1992 ! FOR EACH PUP 23 2 4 WPLA0N_RODENT = (0.6/(1+(5d+3*EXP(-0.0225*(TESTGEST)))))*N_FETUS 2 5 WPLA0R = (WPLA0N_RODENT/WT0)*Gest_on 2 6 WPLA0 = DIM(WPLA0R,0.0) 27 2 8 ! PLACENTA,FLOW RATE, PLACENTA,FLOW RATE, PLACENTA,FLOW RATE, PLACENTA,FLOW 2 9 RATE 3 0 ! QPLA PLACENTA GROWTH EXPRESSION, DOUBLE EXPONENTIAL WITH OFFSET 31 ! FROM O'FLAHERTY_1992 32 33 QPLARF = (1.67d-7 *exp(9.6d-3*(TESTGEST)) & 3 4 +1.6d-3*exp(7.9d-3*(TESTGEST))+0.0)*Gest_on*SWITCH_trans 35 QPLAF=DIM(QPLARF,0.0) !FRACTION OF FLOW RATE IN PLACENTA 36 3 7 ! GESTATION CONTROL 3 8 IF (T.LT.MATTING) THEN 3 9 Gest off = 1 4 0 Gest on= 0.0 41 ELSE 4 2 Gest off = 0.0 4 3 Gest on = 1 4 4 END IF 45 4 6 ! MOTHER BODY WEIGHT GROWTH EQUATION======== 4 7 ! MODIFICATION TO ADAPT THIS MODEL AT HUMAN MODEL 4 8 ! BECAUSE LINEAR DESCRIPTION IS NOT GOOD ENOUGH FOR MOTHER GROWTH 4 9 ! MOTHER BODY WEIGHT GROWTH 50 51 PARAMETER (BW_RMN = 1.0E-30) 5 2 WT0= BW_T0 *(1.0+(0.41*T)/(1402.5+T+BW_RMN)) 53 5 4 ! VARIABILITY OF REST OF THE BODY DEPENDS ON OTHER ORGANS 55 WRE0 = (0.91 - (WLIB0*WLI0 + WFB0*WF0 +WPLAB0*WPLA0 + WLI0 + WF0 + 5 6 WPLA0))/(1.0+WREB0) ! REST OF THE BODY FRACTION; UPDATED FOR EPA ASSESSMENT This document is a draftfor review purposes only and does not constitute Agency policy. C-90 DRAFT--DO NOT CITE OR QUOTE 1 QREF = 1.0-(QFF+QLIF+QPLAF) !REST OF BODY BLOOD FLOW RATE 2 (ML/HR) 3 QTTQF = QFF+QREF+QLIF+QPLAF SUM MUST EQUAL 1 4 5 ! COMPARTMENT VOLUME (ML OR G) 6 WF = WF0 * WT0 ADIPOSE TISSUE 7 WRE = WRE0 * WT0 REST OF THE BODY 8 WLI = WLI0 * WT0 LIVER 9 WPLA= WPLA0* WT0 PLACENTA 10 11 COMPARTMENT TISSUE BLOOD (ML OR G) 12 WFB = WFB0 * WF ! ADIPOSE TISSUE 13 WREB = WREB0 * WRE ! REST OF THE BODY 14 WLIB = WLIB0 * WLI ! LIVER 15 WPLAB = WPLAB0* WPLA ! PLACANTA 16 17 ! CARDIAC OUTPUT FOR THE GIVEN BODY WEIGHT 18 !QC= QCCAR*60*(WT0/1000.0)**0.75 19 CONSTANT QCC=16500 ! EQUIVALENT TO 275 * 60 2 0 QC= QCC*(WT0/UNITCORR)**0.75 21 2 2 (COMPARTMENT BLOOD FLOW RATE (ML/HR) 2 3 QF = QFF*QC !ADIPOSE TISSUE BLOOD FLOW RATE 2 4 QLI = QLIF*QC !LIVER TISSUE BLOOD FLOW RATE 2 5 QRE = QREF*QC !REST OF THE BODY BLOOD FLOW RATE 2 6 QPLA = QPLAF*QC !PLACENTA TISSUE BLOOD FLOW RATE 2 7 QTTQ = QF+QRE+QLI+QPLA !TOTAL FLOW RATE 28 2 9 (PERMEABILITY ORGAN FLOW (ML/HR) : 3 0 PAF = PAFF*QF ADIPOSE TISSUE 31 PARE = PAREF*QRE REST OF THE BODY 3 2 PALI = PALIF*QLI LIVER TISSUE 33 PAPLA = PAPLAF*QPLA PLACENTA 34 35 ************************************** 3 6 ABSORPTION SECTION 3 7 ORAL, 3 8 INTRAPERITONEAL, 3 9 INTRAVENOUS 4 0 ************************************** 41 4 2 !REPETITIVE ORAL BACKGROUND EXPOSURE SCENARIO 43 4 4 MSTOT_NMBCKGR = MSTOTBCKGR/322 !AMOUNT IN NMOL/G 4 5 MSTTBCKGR =MSTOT_NMBCKGR *WT0 46 4 7 DAY_EXPOSURE_BG = PULSE(DAY_LACK_BG,DAY_PERIOD_BG,DAY_FINISH_BG) 4 8 WEEK_EXPOSURE_BG = PULSE(WEEK_LACK_BG,WEEK_PERIOD_BG,WEEK_FINISH_BG) 4 9 MONTH_EXPOSURE_BG = PULSE(MONTH_LACK_BG,MONTH_PERIOD_BG,MONTH_FINISH_BG) 50 51 MSTTCH_BG = (DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG)*MSTTBCKGR 52 MSTTFR_BG = MSTTBCKGR/CINT 53 5 4 CYCLE_BG =DAY_EXPOSURE_BG*WEEK_EXPOSURE_BG*MONTH_EXPOSURE_BG 55 5 6 ! CONDITIONAL ORAL EXPOSURE (BACKGROUND EXPOSURE) This document is a draftfor review purposes only and does not constitute Agency policy. C-91 DRAFT--DO NOT CITE OR QUOTE 1 2 IF (MSTTCH_BG.EQ.MSTTBCKGR) THEN 3 ABSMSTT_GB= MSTTFR_BG 4 ELSE 5 ABSMSTT_GB = 0.0 6 END IF 7 8 CYCLETOTBG=INTEG(CYCLE_BG,0.0) 9 10 IREPETITIVE ORAL EXPOSURE SCENARIO 11 12 MSTT= MSTOT_NM * WT0 IAMOUNT IN NMOL 13 14 DAY_EXPOSURE = PULSE(DAY_LACK,DAY_PERIOD,DAY_FINISH) 15 WEEK_EXPOSURE = PULSE(WEEK_LACK,WEEK_PERIOD,WEEK_FINISH) 16 MONTH_EXPOSURE = PULSE(MONTH_LACK,MONTH_PERIOD,MONTH_FINISH) 17 18 MSTTCH = (DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE)*MSTT 19 MSTTFR = MSTT/CINT 20 21 CYCLE = DAY_EXPOSURE*WEEK_EXPOSURE*MONTH_EXPOSURE 2 2 SUMEXPEVENT= INTEG (CYCLE,0.0)/cint INUMBER OF CYCLES GENERATED DURING 23 SIMULATION 24 2 5 I CONDITIONAL ORAL EXPOSURE 2 6 IF (MSTTCH.EQ.MSTT) THEN 2 7 ABSMSTT= MSTTFR 2 8 ELSE 2 9 ABSMSTT = 0.0 3 0 END IF 31 32 33 CYCLETOT=INTEG(CYCLE,0.0) 34 3 5 ! MASS CHANGE IN THE LUMEN 3 6 RMSTT= -(KST+KABS)*MST +ABSMSTT +ABSMSTT_GB ! RATE OF CHANGE (NMOL/H) 3 7 MST = INTEG(RMSTT,0.0) IAMOUNT REMAINING IN DUODENUM 3 8 (NMOL) 39 4 0 ! ABSORPTION IN LYMPH CIRCULATION 41 LYRMLUM = KABS*MST*A 4 2 LYMLUM = INTEG(LYRMLUM,0.0) 43 4 4 ! ABSORPTION IN PORTAL CIRCULATION 4 5 LIRMLUM = KABS*MST*B 4 6 LIMLUM = INTEG(LIRMLUM,0.0) 47 48 4 9 ---- IV EXPOSURE --------- 50 51 IV= DOSEIV_NM * WT0 IAMOUNT IN NMOL 5 2 IVR= IV/PFUNC I RATE FOR IV INFUSION IN BLOOD 53 EXPIV= IVR * (1.0-STEP(PFUNC)) 5 4 IVDOSE = integ(EXPIV,0.0)I 55 5 6 I------IV late in the cycle This document is a draftfor review purposes only and does not constitute Agency policy. C-92 DRAFT--DO NOT CITE OR QUOTE 1 ! MODIFICATION ON January 13 2004 2 IV_RlateR = DOSEIVNMlate*WT0 3 IV_EXPOSURE=PULSE(IV_LACK,IV_PERIOD,IV_FINISH) 4 5 IV_lateT = IV_EXPOSURE *IV_RlateR 6 IV_late = IV_lateT/CINT 7 8 SUMEXPEVENTIV= integ (IV_EXPOSURE,0.0) !NUMBER OF CYCLE GENERATE DURING 9 SIMULATION 10 11 !SYSTEMIC CONCENTRATION OF TCDD 12 ! MODIFICATION ON OCTOBER 6, 2009 13 CB=(QF*CFB+QRE*CREB+QLI*CLIB+EXPIV+LYRMLUM+QPLA*CPLAB+IV_late)/(QC+CLURI) ! 14 CA = CB ! CONCENTRATION (NMOL/ML) 15 16 !URINARY EXCRETION BY KIDNEY 17 MODIFICATION ON OCTOBER 6, 2009 18 RAURI = CLURI *CB 19 AURI = INTEG(RAURI,0.0) 20 21 !UNIT CONVERSION POST SIMULATION 2 2 CBSNGKGLIADJ=(CB*MW*UNITCORR*(1/B_TOTLIP)*(1/SERBLO))![NG of TCDD Serum/Kg 2 3 OF LIPIP] 2 4 AUCBS_NGKGLIADJ=integ(CBSNGKGLIADJ,0.0) 25 2 6 PRCT_B = (CB/(MSTT+1E-30))*100 ! PERCENT OF ORAL DOSE IN BLOOD 2 7 PRCT_BIV = (CB/(IV_RlateR+1E-30))*100 ! PERCENT OF IV DOSE IN BLOOD 2 8 CBNGKG= CB*MW*UNITCORR 2 9 CBNGG = CB*MW 30 31 !ADIPOSE COMPARTMENT 3 2 !TISSUE BLOOD COMPARTMENT 33 RAFB= QF*(CA-CFB)-PAF*(CFB-CF/PF) !(NMOL/H) 3 4 AFB = INTEG(RAFB,0.0) !(NMOL) 3 5 CFB = AFB/WFB !(NMOL/ML) 3 6 !TISSUE COMPARTMENT 3 7 RAF = PAF*(CFB-CF/PF) !(NMOL/H) 3 8 AF = INTEG(RAF,0.0) !(NMOL) 3 9 CF = AF/WF !(NMOL/ML) 40 41 !UNIT CONVERSION POST SIMULATION 4 2 CFTOTAL= (AF + AFB)/(WF + WFB) ! TOTAL CONCENTRATION IN NMOL/ML 4 3 CFTFREE = CFB + CF !TOTAL FREE CONCENTRATION IN FAT (NM/ML) 4 4 PRCT_F = (CFTOTAL/(MSTT+1E-30))*100 ! PERCENT OF ORAL DOSE IN FAT 4 5 PRCT_FIV = (CFTOTAL/(IV_RlateR+1E-30))*100 ! PERCENT OF IV DOSE IN FAT 4 6 CFNGKG=CFTOTAL*MW*UNITCORR ! FAT CONCENTRATION IN NG/KG 4 7 AUCF_NGKGH=integ(CFNGKG,0.0) 4 8 CFNGG = CFTOTAL*MW 49 5 0 !REST OF THE BODY COMPARTMENT 51 RAREB= QRE *(CA-CREB)-PARE*(CREB-CRE/PRE) !(NMOL/H) 5 2 AREB = INTEG(RAREB,0.0) !(NMOL) 53 CREB = AREB/WREB !(NMOL/H) 5 4 !TISSUE COMPARTMENT 55 RARE = PARE*(CREB - CRE/PRE) !(NMOL/H) 5 6 ARE = INTEG(RARE,0.0) !(NMOL) This document is a draftfor review purposes only and does not constitute Agency policy. C-93 DRAFT--DO NOT CITE OR QUOTE 1 CRE = ARE/WRE !(NMOL/ML) 2 3 !UNIT CONVERSION POST SIMULATION 4 CRETOTAL= (ARE + AREB)/(WRE + WREB) ! TOTAL CONCENTRATION IN 5 NMOL/ML 6 PRCT_RE = (CRETOTAL/(MSTT+1E-30))*100 ! PERCENT OF ORAL DOSE IN REST OF 7 BODY 8 PRCT_REIV = (CRETOTAL/(IV_RlateR+1E-30))*100 ![ PERCENT OF IV DOSE IN 9 REST OF THE BODY ] 10 CRENGKG=CRETOTAL*MW*UNITCORR ! REST OF THE BODY CONCENTRATION IN NG/KG 11 12 13 !LIVER COMPARTMENT 14 !TISSUE BLOOD COMPARTMENT 15 RALIB = QLI*(CA-CLIB)-PALI*(CLIB-CFLLIR)+LIRMLUM ! 16 ALIB = INTEG(RALIB,0.0) !(NMOL) 17 CLIB = ALIB/WLIB !(NMOL/ML) 18 !TISSUE COMPARTMENT 19 RALI = PALI*(CLIB - CFLLIR)-REXCLI ! (NMOL/HR) 2 0 ALI = INTEG(RALI,0.0) !(NMOL) 21 CLI = ALI/WLI !(NMOL/ML) 22 2 3 !FREE TCDD IN LIVER COMPARTMENT 2 4 PARAMETER (LIVER_1RMN = 1.0E-30) 2 5 CFLLI= IMPLC(CLI-(CFLLIR*PLI+(LIBMAX*CFLLIR/(KDLI+CFLLIR & 2 6 +LIVER_1RMN))+((CYP1A2_1O3*CFLLIR/(KDLI2 + CFLLIR & 2 7 +LIVER_1RMN)*PAS_INDUC)))-CFLLI,CFLLI0) 2 8 CFLLIR=DIM(CFLLI,0.0) ! FREE CONCENTRATION IN LIVER 29 3 0 CBNDLI= LIBMAX*CFLLIR/(KDLI+CFLLIR+LIVER_1RMN) !BOUND CONCENTRATION 31 3 2 !VARIABLE ELIMINATION BASED ON THE CYP1A2 33 KBILE_LI_T =((CYP1A2_1OUT-CYP1A2_1A2)/CYP1A2_1A2)*Kelv ! INDUCED BILIARY 3 4 EXCRETION RATE CONSTANT 3 5 REXCLI = KBILE_LI_T*CFLLIR*WLI ! DOSE-DEPENDENT EXCRETION RATE 3 6 EXCLI = INTEG(REXCLI,0.0) 37 3 8 !UNIT CONVERSION POST SIMULATION 3 9 CLITOTAL= (ALI + ALIB)/(WLI + WLIB) ! TOTAL CONCENTRATION IN NMOL/ML 4 0 PRCT_LI = (CLITOTAL/(MSTT+1E-30))*100 ! PERCENT ORAL DOSE IN LIVER 41 PRCT_LIIV = (CLITOTAL/(IV_RlateR+1E-30))*100 ! PERCENT IV DOSE IN LIVER 4 2 Rec_occ= CFLLIR/(KDLI+CFLLIR) 4 3 CLINGKG=CLITOTAL*MW*UNITCORR ! LIVER CONCENTRATION IN NG/KG 4 4 AUCLI_NGKGH=INTEG(CLINGKG,0.0) 4 5 CBNDLINGKG = CBNDLI*MW*UNITCORR 4 6 AUCBNDLI_NGKGH =INTEG(CBNDLINGKG,0.0) 4 7 CLINGG = CLITOTAL*MW 48 4 9 !CHEMICAL IN CYP450 (1A2) COMPARTMENT 5 0 CYP1A2_1KINP = CYP1A2_1KOUT* CYP1A2_1OUTZ ! BASAL RATE OF CYP1A2 PRODUCTION 51 SET EQUAL TO BASAL RATE OF DEGREDATION 52 53 ! MODIFICATION ON OCTOBER 6, 2009 5 4 CYP1A2_1OUT =INTEG(CYP1A2_1KINP * (1.0 + CYP1A2_1EMAX *(CBNDLI+1.0e-30)**HILL 55 & 5 6 /(CYP1A2_1EC50**HILL + (CBNDLI+1.0e-30)**HILL)) & This document is a draftfor review purposes only and does not constitute Agency policy. C-94 DRAFT--DO NOT CITE OR QUOTE 1 - CYP1A2_1KOUT*CYP1A2_1OUT, CYP1A2_1OUTZ) 2 3 ! EQUATIONS INCORPORATING DELAY OF CYP1A2 PRODUCTION (NOT USED IN 4 SIMULATIONS) 5 6 CYP1A2_1RO2 = (CYP1A2_1OUT - CYP1A2_1O2)/ CYP1A2_1TAU 7 CYP1A2_1O2 =INTEG(CYP1A2_1RO2, CYP1A2_1A1) 8 9 CYP1A2_1RO3 = (CYP1A2_1O2 - CYP1A2_1O3)/ CYP1A2_1TAU 10 CYP1A2_1O3 =INTEG(CYP1A2_1RO3, CYP1A2_1A2) 11 12 ! TRANSFER OF DIOXIN FROM PLACENTA TO FETUS 13 ! FETAL EXPOSURE ONLY DURING EXPOSURE 14 15 IF (T.LT.TRANSTIME_ON) THEN 16 SWITCH_trans = 0.0 17 ELSE 18 SWITCH_trans = 1 19 END IF 20 21 !TRANSFER OF DIOXIN FROM PLACENTA TO FETUS 2 2 ! MODIFICATION 26 SEPTEMBER 2003 23 2 4 CONSTANT PFETUS= 4 ! 2 5 CONSTANT CLPLA_FET = 0.17 ! 26 2 7 RAMPF = (CLPLA_FET*CPLA) *SWITCH_trans 2 8 AMPF=INTEG(RAMPF,0.0) 29 3 0 !TRANSFER OF DIOXIN FROM FETUS TO PLACENTA 31 RAFPM = (CLPLA_FET*CFETUS_v)*SWITCH_trans ! 3 2 AFPM = INTEG(RAFPM,0.0) 33 3 4 ! TCDD IN PLACENTA MOTHER COMPARTMENT 3 5 RAPLAB= QPLA*(CA - CPLAB)-PAPLA*(CPLAB -CFLPLAR) ! NMOL/H) 3 6 APLAB = INTEG(RAPLAB,0.0) ! (NMOL) 3 7 CPLAB = APLAB/(WPLAB+1E-30) ! (NMOL/ML) 3 8 RAPLA = PAPLA*(CPLAB-CFLPLAR)-RAMPF + RAFPM ! (NMOL/H) 3 9 APLA = INTEG(RAPLA,0.0) ! (NMOL) 4 0 CPLA = APLA/(WPLA+1e-30) ! (NMOL/ML) 41 4 2 PARAMETER (PARA_ZERO = 1.0E-30) 43 CFLPLA= IMPLC(CPLA-(CFLPLAR*PPLA +(PLABMAX*CFLPLAR/(KDPLA& 4 4 +CFLPLAR+PARA_ZERO)))-CFLPLA,CFLPLA0) 4 5 CFLPLAR=DIM(CFLPLA,0.0) 46 4 7 !UNIT CONVERSION POST SIMULATION 4 8 CPLATOTAL= (APLA + APLAB)/((WPLA + WPLAB)+1e-30)! TOTAL CONCENTRATION IN 4 9 NMOL/ML 5 0 PRCT_PLA = (CPLATOTAL/(MSTT+1E-30))*100 51 PRCT_PLAIV = (CPLATOTAL/(IV_RlateR+1E-30))*100 5 2 CPLANGG = CPLATOTAL*MW 53 5 4 !FETUS COMPARTMENT 55 RAFETUS= RAMPF-RAFPM 5 6 AFETUS=INTEG(RAFETUS,0.0) This document is a draftfor review purposes only and does not constitute Agency policy. C-95 DRAFT--DO NOT CITE OR QUOTE 1 CFETUS=AFETUS/(WTFE+1E-30) 2 CFETOTAL= CFETUS 3 CFETUS_v = CFETUS/PFETUS 4 5 ! UNIT CONVERSION POST SIMULATION 6 CFETUSNGKG = CFETUS*MW*UNITCORR !(NG/KG) 7 AUC_FENGKGH = INTEG(CFETUSNGKG,0.0) 8 PRCT_FE = (CFETOTAL/(MSTT+1E-30))*100 9 PRCT_FEIV = (CFETOTAL/(IV_RlateR+1E-30))*100 10 CFETUSNGG = CFETOTAL*MW 11 12 ! ------------ CONTROL MASS B A L A N C E ---------- 13 BDOSE= IVDOSE +LYMLUM+LIMLUM 14 BMASSE = EXCLI+AURI+AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB+AFETUS 15 BDIFF = BDOSE-BMASSE 16 17 !BODY BURDEN (NG) 18 BODY_BURDEN = AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB ! 19 BBFETUSNG = AFETUS*MW*UNITCORR ! NG 2 0 ! BODY BURDEN IN TERMS OF CONCENTRATION (NG/KG) 21 BBNGKG =(((AFB+AF+AREB+ARE+ALIB+ALI+APLA+APLAB)/WT0)*MW*UNITCORR) ! 2 2 AUC BBNGKGH=INTEG(BBNGKG,0.0) 23 24 2 5 --------COMMAND OF THE END OF SIMULATION -------- 2 6 TERMT (T.GE. TimeLimit, 'Time limit has been reached.') 2 7 END ! END OF THE DERIVATIVE SECTION 2 8 END ! END OF THE DYNAMIC SECTION 2 9 END ! END OF THE PROGRAM 30 31 C .2.6.2. Input Files 3 2 C .2.6.2.1. Keller et al. (2007). 33 %clear variable 3 4 output @clear 3 5 prepare @clear T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH 3 6 AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH 3 7 CBNGKG AUC_CBNGKGH 38 3 9 %output @nciout=10 T SUMEXPEVENT wt0 40 41 %Keller et al. 2007 4 2 %protocol: single oral dose at GD13 43 %DevTCDD4Species.csl 4 4 %MICE_GESTATIONAL_ICF_F092309.csl 4 5 %dose levels: 0.01, 0.100 1 ug/kg at GD13 4 6 %dose levels: 10, 100 1000 ng/kg at GD13 47 4 8 %EXPOSURES SCENARIOS 4 9 MAXT=0.01 5 0 CINT =0.1 51 EXP TIME ON = 312. OO delay before begin exposure (HOUR) 52 EXP "t i m e OFF = 336 OO TIME :EXPOSURE1 STOP (HOUR) 53 DAY_CYCLE = 24 5 4 BCK TIME ON = 0. OO DELAY BEFORE BACGROUND EXPOSURE (HOUR) This document is a draftfor review purposes only and does not constitute Agency policy. C-96 DRAFT--DO NOT CITE OR QUOTE 1 BCK TIME OFF = 0. % TIME OF BACKGROUND EXPOSURE STOP (HOUR) 2 IV LACK = 505 3 IV PERIOD = 505 4 TIMELIMIT = 336 % SIMULATION LIMIT TIME (HOUR) 5 BW T0 = 24 6 MATTING = 0. % BEGINNING MATTING (HOUR) 7 TRANSTIME ON = 144 % SHOULD BE MATTING TIME + 6 DAYS(144 8 HOURS) 9 N_FETUS = 10 10 11 %EXPOSURE DOSE SCENARIOS (UG/KG) 12 13 %MSTOT = 0.01 % ORAL EXPOSURE DOSE (UG/KG) 14 %MSTOT 0.1 % ORAL EXPOSURE DOSE (UG/KG) 15 MSTOT 1 % ORAL EXPOSURE DOSE (UG/KG) 16 17 C .2.6.2.2. L i et al. (2006). 18 %TO BE USED AFTER THE 19 %clear variable 2 0 output @clear 21 prepare @clear T CLINGKG CFNGKG CBSNGKGLIADJ BBNGKG CFETUSNGKG AUCLI_NGKGH 2 2 AUCF_NGKGH AUCBS_NGKGLIADJ AUC_BBNGKGH AUC_FENGKGH CBNDLINGKG AUCBNDLI_NGKGH 23 CBNGKG AUC_CBNGKGH 2 4 %output @nciout=10 T SUMEXPEVENT 2 5 %Li et al.2006 2 6 %protocol: daily oral dose from GDI to GD3 2 7 %DevTCDD4Species.csl 2 8 %MICE_GESTATIONAL_ICF_F092309.csl 2 9 %dose levels: 0.002, 0.050, 0.10 ug/kg/day at GDI to GD3 3 0 %dose levels: 2, 50, 100 ng/kg/day from GDI to GD3 31 3 2 %EXPOSURES SCENARIOS 33 MAXT=0.01 3 4 CINT =0.1 3 5 EXP TIME ON = 0. % delay before begin exposure (HOUR) 3 6 EXP TIME OFF = 72 % TIME EXPOSURE STOP (HOUR) 2 HOURS LESS THAN 3 7 GD3 put 70 to be sure 3 doses will be administrate 3 8 % BECAUSE i STARTED TIME 0 FOR GD1 3 9 DAY CYCLE = 24 4 0 BCK TIME ON = 0. % DELAY BEFORE BACGROUND EXPOSURE (HOUR) 41 BCK TIME OFF = 0. % TIME OF BACKGROUND EXPOSURE STOP (HOUR) 4 2 IV LACK = 505 4 3 IV PERIOD = 505 4 4 TIMELIMIT = 72. % SIMULATION LIMIT TIME (HOUR) Run for 3 4 5 days 4 6 BW T0 = 27 4 7 MATTING = 0. % BEGINNING MATTING (HOUR) 4 8 TRANSTIME ON = 144. % SHOULD BE MATTING TIME + 6 DAYS(144 4 9 HOURS) 5 0 N FETUS = 10 51 5 2 %EXPOSURE DOSE SCENARIOS (UG/KG) 53 5 4 %MSTOT = 0.002 % ORAL EXPOSURE DOSE (UG/KG) 55 %MSTOT = 0.05 % ORAL EXPOSURE DOSE (UG/KG) This document is a draftfor review purposes only and does not constitute Agency policy. C-97 DRAFT--DO NOT CITE OR QUOTE 1 MSTOT 2 = 0.10 % ORAL EXPOSURE DOSE (UG/KG) 3 C.3. TO XICO K INETIC M O D ELIN G RESULTS FO R K EY A N IM A L BIO ASSAY 4 STUDIES 5 T he sim ulated T C D D serum -adjusted lipid concentrations rep o rted in this ap pendix fo r 6 th e ro d en t bio assay s w ere converted to T C D D concentrations in ro d en t w h o le blood. Initially, 7 E P A m u ltip lied th e seru m -ad ju sted lip id co n cen tratio n s by 0.0033, th e ratio o f lip id co n ten t to 8 total serum v o lu m e, th en by 0.55, th e v alu e o f th e hem atocrit. T his p ro d u ct y ield s th e T C D D 9 co ncentration in w h o le ro d en t b lo o d as predicted by th e P B P K m odel. E P A assum ed th at the 10 sam e w h o le b lo o d T C D D co n cen tratio n w o u ld re su lt in th e sam e effects in h u m an s and rodents. 11 T h is c o n v e rs io n a c c o m p lis h e s th e fo llo w in g : 12 1. A llo w s th e h u m a n e q u iv a le n t d o s e ( H E D ) to b e b a s e d o n e q u iv a le n t b lo o d c o n c e n tr a tio n 13 (th a t re p re s e n ts se ru m p lu s e ry th ro c y te T C D D ), w h ic h is p ro p o rtio n a l to tis s u e e x p o su re ; 14 2. A v o id s criticism th a t th e to tal b lo o d co n cen tratio n is n o rm alized to seru m lip id alo n e in 15 an u n b a la n c e d w a y (th u s E P A d o es n o t c o n tra d ic t C e n te rs fo r D ise a se C o n tro l an d 16 P rev e n tio n (C D C ) d ata o r m eth o d s); 17 3. F acto rs o u t any im p act o f th e lip id co n ten t u sed in th e P B P K m o d el; and 18 4. T C D D c o n c en tratio n in w h o le b lo o d is e n c o u rag e d fo r u se in th e a sse ssm e n ts b y th e N A S 19 (N A S , 2 0 0 6 , p. 43); see ad d itio n al in fo rm a tio n in S ectio n 3.3. 20 21 C.3.1. Nongestational Studies 2 2 C .3.1.1. Cantoni et al. (1981) Type: Strain: Rat C D -C O B S rats Dose: Route: 10, 100, 1000 n g /k g /w e e k Oral gavage exposure Body weight: B W set to 125g Regime: 1 dose/w eek for 45 w eeks Sex: Fem ale Simulation time: 7,560 hours (45 w eeks) 23 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 1.43 CADM 1.85 - 3.70 (@ 7,392 hours) - 14.29 Emond 8.84 26.6 (@ 7,392 hours) Terminal 1.82 - 7.97 This document is a draftfor review purposes only and does not constitute Agency policy. C-98 DRAFT--DO NOT CITE OR QUOTE 142.86 Dose (ng/kg-day) Adjusted dose 1.43 14.29 142.86 Dose (ng/kg-day) Adjusted dose 1.43 14.29 142.86 Dose (ng/kg-day) Adjusted dose 1.43 14.29 142.86 CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM -50.0 227 (@ 7,392 hours) -L IV E R C O N C E N T R A T IO N S (ng/kg) Metric Time-weighted Ave Max 247 328 (@ 7,398 hours) 374 431 2,176 2,860 (@ 7,231 hours) 3,884 4,330 20,500 26,978 (@ 7,399 hours) 39,067 43,329 F A T C O N C E N T R A T IO N S (ng/kg) Metric Time-weighted Ave Max 175 200 (@ 7,431 hours) 250 280 837 937 (@ 7,427 hours) 1,209 1,352 4,741 5,374 (@ 7,424 hours) 10,050 11,224 B O D Y B U R D E N (ng/kg) Metric Time-weighted Ave Max 26.1 32.0 170 225 1,337 2,106 31.7 (@ 7,398 hours) 35.0 210 (@ 7,230 hours) 243 1,695 (@ 7,398 hours) 2,266 41.9 - Terminal 242 431 1,928 4,330 17,255 43,329 Terminal 181 244 807 1,167 4,349 9,734 Terminal 26.3 35.0 156 243 1,151 2,266 This document is a draftfor review purposes only and does not constitute Agency policy. C-99 DRAFT--DO NOT CITE OR QUOTE Dose (ng/kg-day) Adjusted dose Model Emond 1.43 CADM 14.29 Emond CADM 142.86 Emond CADM 1 2 3 C .3 .1 .2 . C h u e t al. (2 0 0 7 ) BO U N D L IV E R (ng/kg) Metric Time-weighted Ave Max 6.04 - 23.7 - 66.8 - 7.76 (@ 7,396 hours) - 29.1 (@ 7,228 hours) - 80.0 (@ 1 hours) - Terminal 6.01 - 22.2 - 63.4 - T y p e: S train: Rat Sprague-D aw ley B ody w eight: 200 g Sex: Fem ale D ose: R oute: R egim e: Sim ulation tim e: 2.5, 25, 250, and 1,000 ng/kg-day Oral exposure 1 dose per day for 28 days 672 hours Dose (ng/kg-day) Adjusted dose 2.5 25 250 1,000 Dose (ng/kg-day) Adjusted dose 2.5 25 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Model Time-weighted Ave Metric Max Emond CADM Emond CADM Emond CADM Emond CADM Model 1.26 2.35 (@ 648 hours) -7.66 15.3 (@ 648 hours) -48.8 113 (@ 648 hours) -169 418 (@ 648 hours) -L IV E R C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max Emond CADM Emond CADM 148 - 1,777 - 268 (@ 652 hours) - 2,953 (@ 653 hours) - Terminal 1.88 - 10.4 - 63.7 - 222 - Terminal 255 - 2,806 - This document is a draftfor review purposes only and does not constitute Agency policy. C-100 DRAFT--DO NOT CITE OR QUOTE Emond 250 CADM 19,232 - 30,262 (@ 653 hours) - 28,668 - 1,000 Emond CADM 77,819 - 120,400 (@ 653 hours) - 113,890 - F A T C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.5 CADM 108 - 180 (@ 668 hours) - 180 - Emond 25 CADM 660 - 1,020 (@ 659 hours) - 1,015 - Emond 250 CADM 4,210 - 6,433 (@ 655 hours) - 6,354 - 1,000 Emond CADM 14,576 - 22,610 (@ 655 hours) - 22,280 - B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.5 CADM 16.1 - 27.5 (@ 652 hours) - 26.9 - Emond 25 CADM 138 - 222 (@ 652 hours) - 214 - Emond 250 CADM 1,239 - 1,935 (@ 652 hours) - 1,842 - 1,000 Emond CADM 4,801 - 7,444 (@ 652 hours) - 7,067 - B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.5 CADM 4.15 - 6.51 (@ 652 hours) - 6.21 - Emond 25 CADM 20.5 - 28.5 (@ 652 hours) - 27.4 - Emond 250 CADM 63.3 - 76.0 (@ 652 hours) - 74.7 - This document is a draftfor review purposes only and does not constitute Agency policy. C-101 DRAFT--DO NOT CITE OR QUOTE 1,000 Emond CADM 90.2 - 99.0 (@ 653 hours) - 98.3 - 1 C .3 .1 .3 . C r o fto n e t al. (2 0 0 5 ) Type: R ats Dose: 0, 0.1, 3, 10, 30, 100, 300, 1000, 3000, and 10,000 ng/kg-day Strain: Long Evans Route: Oral exposure Body weight: 4 w eeks old B W set to 190 g Regime: One dose per day for four days Sex: Fem ale Simulation time: 96 h o u rs 2 The C A D M m odel w as not ru n because the dosing duratio n is low er th an the resolution o f the m odel (1 w eek) Dose (ng/kg-day) Adjusted dose 0.1 3 10 30 100 300 1000 3000 10,000 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Model Time-weighted Ave Metric Max Emond 0.0202 0.041 (@ 72 hours) CADM - - Emond 0.488 1.10 (@ 72 hours) CADM - - Emond 1.38 3.40 (@ 72 hours) CADM - - Emond 3.46 9.44 (@ 72 hours) CADM - - Emond 9.26 29.0 (@ 72 hours) CADM - - Emond 23.1 81.8 (@ 72 hours) CADM - - Emond 65.7 260 (@ 72 hours) CADM - - Emond 181 764 (@ 72 hours) CADM - - Emond 583 2,527 (@ 72 hours) CADM - - Terminal 0.0244 - 0.582 - 1.62 - 3.93 - 10.2 24.5 - 68.2 187 607 - This document is a draftfor review purposes only and does not constitute Agency policy. C-102 DRAFT--DO NOT CITE OR QUOTE Dose (ng/kg-day) Adjusted dose 0.1 3 10 30 100 300 1000 3000 10,000 Dose (ng/kg-day) Adjusted dose 0.1 3 10 30 100 L IV E R C O N C E N T R A T IO N S (ng/kg) Model Time-weighted Ave Metric Max Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Model 0.919 1.55 (@ 75 hours) -37.4 62.6 (@ 76 hours) -145 242 (@ 77 hours) -494 818 (@ 78 hours) 1,839 3,025 (@ 78 hours) 5,925 9,692 (@ 78 hours) 20,717 33,738 (@ 79 hours) 63,511 103,140 (@ 79 hours) 212,890 344,910 (@ 79 hours) -F A T C O N C E N T R A T IO N S (ng/kg) Metric Time-weighted Ave Max Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM 1.00 - 24.6 - 70.3 - 177 - 480 - 1.93 (@ 96 hours) - 45.9 (@ 96 hours) - 129 (@ 96 hours) - 317 (@ 96 hours) - 838 (@ 96 hours) - Terminal 1.18 - 53.3 - 214 - 742 - 2,793 - 9,028 - 31,564 - 96,545 - 321,960 - Terminal 1.93 - 45.9 - 129 - 317 - 838 - This document is a draftfor review purposes only and does not constitute Agency policy. C-103 DRAFT--DO NOT CITE OR QUOTE 300 1000 3000 10,000 Dose (ng/kg-day) Adjusted dose 0.1 3 10 30 100 300 1000 3000 10,000 Dose (ng/kg-day) Adjusted dose 0.1 Emond CADM Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Model Emond CADM 1,206 2,065 (@ 96 hours) -- 3,452 5,836 (@ 96 hours) -- 9,522 16,050 (@ 96 hours) -- 30,657 51,918 (@ 96 hours) -- B O D Y B U R D E N (ng/kg) Metric Time-weighted Ave Max 0.138 0.224 (@ 79 hours) -- 4.04 6.56 (@ 78 hours) -- 13.3 21.5 (@ 78 hours) -- 39.3 63.5 (@ 78 hours) -- 129 208 (@ 78 hours) -- 384 618 (@ 77 hours) -- 1,270 2,041 (@ 77 hours) -- 3,793 6,094 (@ 77 hours) -- 12,595 20,226 (@ 77 hours) -- B O U N D L IV E R (ng/kg) Metric Time-weighted Ave Max 0 0.115 (@ 75 hours) -- 2,065 - 5,836 - 16,050 - 51,918 - Terminal 0.223 - 6.44 - 21.0 - 61.5 - 200 - 590 - 1,942 - 5,784 - 19,154 - Terminal 0 - This document is a draftfor review purposes only and does not constitute Agency policy. C-104 DRAFT--DO NOT CITE OR QUOTE Emond 2 3 CADM - 10 Em ond CADM Emond 30 CADM 100 Em ond CADM 4 - 10 - 22 - 300 1000 Emond CADM Emond CADM 41 - 68 - 3000 Emond CADM 90 - 10,000 Emond CADM 104 - 1 2 3 C .3 .1 .4 . D e lla P o r ta e t al. (2 0 0 1 ) (fe m a le ) 2.47 (@ 76 hours) - 6.42 (@ 76 hours) - 14.1 (@ 76 hours) - 29.9 (@ 76 hours) - 51.9 (@ 77 hours) - 80.2 (@ 1 hours) - 98.6 (@ 1 hours) - 108 (@ 1 hours) - 2 5 - 12 27 49 77 96 107 - T y p e: M ouse D ose: 2,500 and 5,000 ng/kg-w eek (equivalent to 357 and 714 ng/kg-day) S train: B6C3 R oute: G avage B ody w eight: 6 w eeks old (BW 20g) R egim e: Once a w eek for 52 weeks Sex: Fem ale Sim ulation tim e: 8,736 hours 4 The CADM m odel was not run because the study duration is longer than the allow ed model duration 5 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 357 CADM 67.0 - 741 (@ 8,568 hours) - 46.8 - Emond 714 CADM 37.6 - 374 (@ 8,568 hours) - 27.2 - This document is a draftfor review purposes only and does not constitute Agency policy. C-105 DRAFT--DO NOT CITE OR QUOTE L IV E R C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 357 CADM 50,269 - 70,070 (@ 8,577 hours) - Emond 714 CADM 25,422 - 35,352 (@ 8,577 hours) - F A T C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 357 CADM 25,235 - 28,559 (@ 8,589 hours) - Emond 714 CADM 14,162 - 15,914 (@ 8,590 hours) - B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 357 CADM 5,473 - 7,247 (@ 8,574 hours) - Emond 714 CADM 2,878 - 3,774 (@ 8,574 hours) - BO U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 357 CADM 71.5 - 99.1 (@ 2 hours) - Emond 714 CADM 56.4 - 88.6 (@ 2 hours) - 1 2 Terminal 37,389 - 19,105 - Terminal 22,498 - 12,810 - Terminal 4,335 - 2,318 - Terminal 65.4 - 50.4 - This document is a draftfor review purposes only and does not constitute Agency policy. C-106 DRAFT--DO NOT CITE OR QUOTE 1 C .3 .1 .5 . D e lla P o r ta e t al. (2 0 0 1 ) (m a le ) Type: M ouse Dose: 2,500 and 5,000 ng/kg-week (equivalent to 357 and 714 ng/kg-day) Strain: B6C3 Route: Gavage Body weight: 6 w eeks old (BW 26g) Regime: Once a week for 52 weeks Sex: M ale Simulation time: 8,736 hours 2 The CADM m odel was not run because the study duration is longer than the allow ed model duration 3 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 357 CADM 67.8 - 787 (@ 8,568 hours) - 47.0 - Emond 714 CADM 38.0 - 398 (@ 8,568 hours) - 27.3 - L IV E R C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 357 CADM 50,397 - 70,052 (@ 8,577 hours) - 37,483 - Emond 714 CADM 25,493 - 35,347 (@ 8,577 hours) - 19,155 - F A T C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 357 CADM 25,516 - 28,851 (@ 8,589 hours) - 22,861 - Emond 714 CADM 14,306 - 16,061 (@ 8,590 hours) - 12,999 - B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 357 CADM 5,504 - 7,282 (@ 8,574 hours) - 4,368 - This document is a draftfor review purposes only and does not constitute Agency policy. C-107 DRAFT--DO NOT CITE OR QUOTE Emond 714 CADM 2,894 - 3,791 (@ 8,574 hours) - B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 357 CADM Emond 714 CADM 1 2 3 C .3 .1 .6 . F a tto r e e t al. (2 0 0 0 ) 71.6 - 56.4 - 99.2 (@ 2 hours) - 88.6 (@ 2 hours) - 2,335 - Terminal 65.4 - 50.4 - T y p e: S train: Rat Sprague D aw ley D ose: R oute: 20, 200, 2,000 ng/kg-day O ral in the diet B ody w eight: 7 w eeks old (BW 150g) R egim e: E v e ry d ay fo r 13 w e ek s Sex: 4 Fem ale and m ale Sim ulation tim e: 2,184 hours W H O L E B L O O D C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose 20 200 Model Emond CADM Emond CADM Time-weighted Ave 9.59 - 57.6 - Metric Max 15.0 (@ 2,160 hours) - 102 (@ 2,160 hours) - Terminal 11.1 63.9 - 2,000 Emond CADM 476 - 903 (@ 2,160 hours) - 522 - This document is a draftfor review purposes only and does not constitute Agency policy. C-108 DRAFT--DO NOT CITE OR QUOTE Dose (ng/kg-day) Adjusted dose 20 200 2,000 Model Emond CADM Emond CADM Emond CADM Dose (ng/kg-day) Adjusted dose 20 200 2,000 Model Emond CADM Emond CADM Emond CADM Dose (ng/kg-day) Adjusted dose 20 200 2,000 Model Emond CADM Emond CADM Emond CADM Dose (ng/kg-day) Adjusted dose 20 200 Model Emond CADM Emond CADM L IV E R C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max 2,448 3,228 (@ 2,164 hours) 4,471 5,639 24,136 30,245 (@ 2,164 hours) 45,337 56,499 234,170 288,020 (@ 2,164 hours) 454,031 565,103 F A T C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max 890 1,113 (@ 2,166 hours) 1,545 1,796 5,355 6,542 (@ 2,165 hours) 13,351 15,604 44,176 54,246 (@ 2,165 hours) 131,259 153,534 B O D Y B U R D E N (ng/kg) Metric Time-weighted Ave Max 187 242 (@ 2,164 hours) 261 324 1,556 1,940 (@ 2,164 hours) 2,496 3,084 14,432 17,797 (@ 2,164 hours) 24,836 30,674 B O U N D L IV E R (ng/kg) Metric Time-weighted Ave Max 24.9 29.8 (@ 2,164 hours) -- 69.4 76.0 (@ 2,164 hours) -- Terminal 3,078 5,639 28,709 56,499 272,590 565,103 Terminal 1,101 1,756 6,430 15,292 53,140 150,516 Terminal 233 324 1,850 3,084 16,891 30,674 Terminal 28.8 - 74.7 - This document is a draftfor review purposes only and does not constitute Agency policy. C-109 DRAFT--DO NOT CITE OR QUOTE 2,000 Emond CADM 104 - 1 2 3 C .3 .1 .7 . F r a n c e t al. (2 0 0 1 ) S p r a g u e D a w le y R a ts 106 (@ 2,164 hours) - 106 - T y p e: R ats D ose: 140, 420, and 1400 ng/kg every tw o w eeks (equivalent to 10, 30, and 100 ng/kg-day) S train: Sprague D aw ley, R oute: Oral gavage B ody w eight: 200 g (10 w eeks old) R egim e: Once every tw o w eeks for 22 w eeks Sex: Fem ale Sim ulation tim e: 3,696 hours 4 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 10 CADM 6.59 - 34.6 (@ 3,360 hours) - 5.52 - Emond 30 CADM 14.5 - 98.1 (@ 3,360 hours) - 11.3 - W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 100 CADM Dose (ng/kg-day) Adjusted dose Model 36.4 315 (@ 3,360 hours) -L IV E R C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max 26.4 - Terminal Emond 10 CADM 1,447 2,616 2,458 (@ 3,368 hours) 3,620 1,150 2,174 Emond 30 CADM 4,228 7,936 7,161 (@ 3,368 hours) 10,899 3,120 6,510 Emond 100 CADM 13,821 26,564 23,417 (@ 3,368 hours) 36,361 9,658 21,703 This document is a draftfor review purposes only and does not constitute Agency policy. C-110 DRAFT--DO NOT CITE OR QUOTE F A T C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max 10 Em ond CADM 619 966 787 (@ 3,417 hours) 1,230 Emond 30 CADM 1,362 2,448 1,741 (@ 3,415 hours) 3,203 100 Em ond CADM 3,430 7,573 4,464 (@ 3,412 hours) 10,052 B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max 10 Em ond CADM 119 159 177 (@ 3,366 hours) 212 Emond 30 CADM 308 450 472 (@ 3,366 hours) 603 100 Em ond CADM 921 1,462 1,445 (@ 3,366 hours) 1,969 B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max 10 Em ond CADM 18.6 - 32.9 (@ 1 hours) - Emond 30 CADM 33.7 - 59.2 (@ 1 hours) - 100 Em ond CADM 57.5 - 1 2 3 C .3 .1 .8 . F r a n c e t al. (2 0 0 1 ) L o n g - E v a n s R a ts 86.9 (@ 1 hours) - Terminal 560 759 1,161 1,849 2,755 5,606 Terminal 99.5 133 240 367 671 1,181 Terminal 16.4 - 29.0 - 50.4 - T y p e: R ats D ose: 140, 420, and 1400 ng/kg every tw o w eeks (equivalent to 10, 30, and 100 ng/kg-day) S train: L o n g -E v an s R oute: Oral gavage B ody w eight: 190 g (10 w eeks old) R egim e: Once every tw o w eeks for 22 w eeks Sex: Fem ale Sim ulation 3,696 hours tim e: This document is a draftfor review purposes only and does not constitute Agency policy. C-111 DRAFT--DO NOT CITE OR QUOTE 1 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 10 CADM 6.58 - 34.2 (@ 3,360 hours) - Emond 30 CADM 14.5 - 97.0 (@ 3,360 hours) - Emond 100 CADM 36.4 - 312 (@ 3,360 hours) - L IV E R C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 10 CADM 1,447 2,616 2,458 (@ 3,368 hours) 3,620 Emond 30 CADM 4,228 7,936 7,161 (@ 3,368 hours) 10,899 Emond 100 CADM 13,821 26,564 23,421 (@ 3,368 hours) 36,361 F A T C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 10 CADM 619 966 788 (@ 3,417 hours) 1,230 Emond 30 CADM 1,362 2,448 1,742 (@ 3,414 hours) 3,203 Emond 100 CADM 3,429 7,573 4,466 (@ 3,412 hours) 10,052 B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 10 CADM 119 159 177 (@ 3,366 hours) 212 Emond 30 CADM 308 450 472 (@ 3,366 hours) 603 Emond 100 CADM 921 1,462 1,445 (@ 3,366 hours) 1,969 Terminal 5.52 - 11.3 - 26.4 - Terminal 1,150 2,174 3,121 6,510 9,659 21,703 Terminal 560 759 1,160 1,849 2,752 5,606 Terminal 99.5 133 240 367 671 1,181 This document is a draftfor review purposes only and does not constitute Agency policy. C-112 DRAFT--DO NOT CITE OR QUOTE Dose (ng/kg-day) Adjusted dose 10 30 Model Emond CADM Emond CADM B O U N D L IV E R (ng/kg) Metric Time-weighted Ave Max 18.6 32.9 (@ 1 hours) -- 33.7 59.2 (@ 1 hours) -- Emond 100 CADM 57.5 - 86.9 (@ 1 hours) - 1 2 3 C .3 .1 .9 . F r a n c e t al. (2 0 0 1 ) H a n s W is ta r R a ts Terminal 16.4 - 29.0 - 50.4 - T y p e: R ats D ose: 140, 420, and 1400 ng/kg every tw o w eeks (equivalent to 10, 30, and 100 ng/kg-day) S train: H ans W istar R oute: Oral gavage B ody w eight: 205 g (10 w eeks old) R egim e: Once every tw o w eeks for 22 w eeks Sex: Fem ale Sim ulation tim e: 3,696 hours 4 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 10 CADM 6.59 - 34.7 (@ 3,360 hours) - 5.52 - Emond 30 CADM 14.5 - 98.7 (@ 3,360 hours) - 11.3 - Emond 100 CADM 36.4 - 317 (@ 3,360 hours) - 26.4 - Dose (ng/kg-day) Adjusted dose Model L IV E R C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max Terminal Emond 10 CADM 1,447 2,616 2,458 (@ 3,368 hours) 3,620 1,150 2,174 Emond 30 CADM 4,228 7,936 7,160 (@ 3,368 hours) 10,899 3,120 6,510 This document is a draftfor review purposes only and does not constitute Agency policy. C-113 DRAFT--DO NOT CITE OR QUOTE 100 Em ond CADM 13,821 26,564 23,416 (@ 3,368 hours) 36,361 F A T C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max 10 Em ond CADM 619 966 787 (@ 3,418 hours) 1,230 Emond 30 CADM 1,363 2,448 1,741 (@ 3,415 hours) 3,203 100 Em ond CADM 3,431 7,573 4,463 (@ 3,412 hours) 10,052 Dose (ng/kg-day) Adjusted dose Model B O D Y B U R D E N (ng/kg) Metric Time-weighted Ave Max 10 Em ond CADM 119 159 177 (@ 3,366 hours) 212 Emond 30 CADM 308 450 472 (@ 3,366 hours) 603 100 Em ond CADM 921 1,462 1,446 (@ 3,366 hours) 1,969 Dose (ng/kg-day) Adjusted dose Model B O U N D L IV E R (ng/kg) Metric Time-weighted Ave Max 10 Em ond CADM Emond 30 CADM 18.6 - 33.7 - 32.9 (@ 1 hours) - 59.2 (@ 1 hours) - Emond 100 CADM 1 57.5 - 86.9 (@ 1 hours) - 9,658 21,703 Terminal 560 759 1,162 1,849 2,757 5,606 Terminal 99.5 133 240 367 671 1,181 Terminal 16.4 - 29.0 - 50.4 - This document is a draftfor review purposes only and does not constitute Agency policy. C-l 14 DRAFT--DO NOT CITE OR QUOTE 1 C .3 .1 .1 0 . H a s s o u n e t al. (2 0 0 0 ) T y p e: Rat D ose: 0, 3, 10, 22, 46, 100 ng/kg/day (2.14, 7.14, 15.7, 32.9, and 71.4 ng/kg/day adjusted doses) S train: Sprague D aw ley R oute: Oral gavage B ody w eight: 8 w eeks old (B W =215g) R egim e: 5 d a y s/w e e k fo r 13 w eek s Sex: Fem ale Sim ulation tim e: 2184 hours 2 W H O L E B L O O D C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 2.14 CADM 1.94 - 3.12 (@ 2,112 hours) - Emond 4.6136 7.71 (@ 2,112 hours) 7.14 CADM - - Emond 8.147 15.7 CADM - 14.2 (@ 2,112 hours) - 32.9 Emond CADM 14.009 - 25.8 (@ 2,112 hours) - 71.4 Emond CADM 25.34 - 49.7 (@ 2,112 hours) - L IV E R C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 266.8 2.14 CADM - 399 (@ 2,116 hours) - Emond 7.14 CADM 888 - 1,259 (@ 2,117 hours) - Emond 1,948.499 2,689 (@ 2,117 hours) 15.7 CADM - - 32.9 Emond CADM 4,055.031 - 5,484 (@ 2,117 hours) - 71.4 Emond 8,774.97 11,692 (@ 2,117 hours) CADM - - Terminal 1,303.17 - 2,901.26 - 4,947.3 - 8,277 - 14,637 - Terminal 349 - 1,079 - 2,278.182 - 4,607.265 - 9,754.31 - This document is a draftfor review purposes only and does not constitute Agency policy. C-115 DRAFT--DO NOT CITE OR QUOTE Dose (ng/kg-day) Adjusted dose 2.14 7.14 15.7 32.9 71.4 Dose (ng/kg-day) Adjusted dose 2.14 7.14 15.7 32.9 71.4 Dose (ng/kg-day) Adjusted dose 2.14 7.14 15.7 32.9 Model Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Emond CADM F A T C O N C E N T R A T IO N S (ng/kg) Metric Time-weighted Ave Max 179.2 243 (@ 2,126 hours) -- 427 553 (@ 2,124 hours) -- 755 958 (@ 2,123 hours) -- 1,299 1,627 (@ 2,122 hours) -- 2,349.892 2,928 (@ 2,121 hours) -- B O D Y B U R D E N (ng/kg) Metric Time-weighted Ave Max 27.425 38.9 (@ 2,116 hours) -- 76.87 105 (@ 2,116 hours) -- 153.1 205 (@ 2,116 hours) -- 295 390 (@ 2,116 hours) -- 600 785 (@ 2,116 hours) -- B O U N D L IV E R (ng/kg) Metric Time-weighted Ave Max 6 13.7242 21.9703 32.817 - 8.48 (@ 2,116 hours) - 17.5 (@ 2,116 hours) - 27.1 (@ 2,116 hours) - 39.2 (@ 2,116 hours) - Terminal 234.9 - 528 - 908 - 1,529 - 2,727.240 - Terminal 35.720 - 93.67 - 180.2 - 339 - 674 - Terminal 8 15.7348 24.4047 35.608 - This document is a draftfor review purposes only and does not constitute Agency policy. C-l 16 DRAFT--DO NOT CITE OR QUOTE 71.4 Emond CADM 1 2 3 C .3 .1 .1 1 . H u t t e t al. (2 0 0 8 ) 47.54 - 55.0 (@ 2,116 hours) - 50.63 - T y p e: S train: Rat Sprague-D aw ley D ose: R oute: 50 ng/kg-w eek Oral gavage B o d y w eig h t: 4.5 g R egim e: 1 /w e e k f o r 13 w e e k s Sex: 4 Fem ale Sim ulation tim e: 2,184 hours (w eekly exposure) W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose 7.14 Model Emond CADM Time-weighted Ave 4.49 - Metric Max 8.86 (@ 2,016 hours) - Terminal 4.71 - L IV E R C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 867.4 7.14 CADM 1,678 1,363 (@ 2,021 hours) 2,007 928.1 2,007 F A T C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 423.6 7.14 CADM 730 555 (@ 2,040 hours) 787.1 459.9 769 B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 7.14 CADM 76 108 108 (@ 2,022 hours) 126 81 126 This document is a draftfor review purposes only and does not constitute Agency policy. C-l 17 DRAFT--DO NOT CITE OR QUOTE B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 7.14 CADM 14 - 19.4 (@ 2,020 hours) - 1 2 3 C .3 .1 .1 2 . K itc h in a n d W o o d s (1 9 7 9 ) Terminal 14 - T y p e: R ats D ose: 0, 0.6, 2, 4, 20, 60, 200, 600, 2000, 5000, 20,000 ng/kg/day S train: Sprague-D aw ley R oute: Oral exposure B ody w eight: 200 to 250 g (B W set R egim e: to 225 g) Single dose Sex: Fem ale Sim ulation tim e: 24 hours* 4 * 1 w eek is the m inim um that can be sim ulated w ith the CADM model, so the CADM m odel was not used. 5 W H O L E B L O O D C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Model Time-weighted Ave Metric Max Terminal Emond 0.0645 0.126 (@ 0 hours) 0.0441 0.6 CADM - -- Emond 0.202 0.421 (@ 0 hours) 0.137 2 CADM - -- Emond 0.384 0.841 (@ 0 hours) 0.258 4 CADM - -- Emond 20 CADM 1.61 4.21 (@ 0 hours) 1.04 - -- Emond 4.15 12.6 (@ 0 hours) 2.55 60 CADM - -- Emond 200 CADM 11.6 42.1 (@ 0 hours) 6.61 - -- Emond 600 CADM 30.3 126 (@ 0 hours) 15.8 - -- 2000 Emond CADM 90.9 422 (@ 0 hours) 42.8 - -- This document is a draftfor review purposes only and does not constitute Agency policy. C-118 DRAFT--DO NOT CITE OR QUOTE 5000 20000 Dose (ng/kg-day) 0.6 2 4 20 60 200 600 2000 5000 20000 Dose (ng/kg-day) 0.6 2 Emond 218 1,056 (@ 0 hours) CADM - - Emond 863 4,233 (@ 0 hours) CADM - - L IV E R C O N C E N T R A T IO N S (ng/kg) Model Time-weighted Ave Metric Max Emond 2.95 3.81 (@ 4 hours) CADM - - Emond 10.5 12.9 (@ 4 hours) CADM - - Emond 22.2 26.3 (@ 4 hours) CADM - - Emond 128 143 (@ 6 hours) CADM - - Emond 420 463 (@ 8 hours) CADM - - Emond 1,523 1,666 (@ 9 hours) CADM - - Emond 4,821 5,258 (@ 10 hours) CADM - - Emond 16,603 18,080 (@ 11 hours) CADM - - Emond 41,971 45,674 (@ 11 hours) CADM - - Emond 167,820 182,580 (@ 11 hours) CADM - - F A T C O N C E N T R A T IO N S (ng/kg) Model Time-weighted Ave Metric Max Emond 1.60 2.47 (@ 24 hours) CADM - - Emond 5.07 7.71 (@ 24 hours) CADM - - 96.9 - 365 - Terminal 2.31 - 8.69 - 18.9 - 118 - 406 - 1,526 - 4,932 - 17,226 - 43,803 - 175,890 - Terminal 2.47 - 7.71 - This document is a draftfor review purposes only and does not constitute Agency policy. C-l 19 DRAFT--DO NOT CITE OR QUOTE 4 20 60 200 600 2000 5000 20000 Dose (ng/kg-day) 0.6 2 4 20 60 200 600 2000 Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM 9.68 14.6 (@ 24 hours) -- 41.7 60.7 (@ 24 hours) -- 110 155 (@ 24 hours) -- 317 427 (@ 24 hours) -- 851 1,102 (@ 24 hours) -- 2,620 3,276 (@ 24 hours) -- 6,361 7,816 (@ 24 hours) -- 25,401 30,827 (@ 24 hours) -- B O D Y B U R D E N (ng/kg) Metric Time-weighted Ave Max 0.322 0.341 (@ 9 hours) -- 1.07 1.14 (@ 8 hours) -- 2.14 2.27 (@ 8 hours) -- 10.6 11.3 (@ 8 hours) -- 31.7 33.8 (@ 7 hours) -- 105 112 (@ 7 hours) -- 315 337 (@ 7 hours) -- 1,049 1,123 (@ 7 hours) -- 14.6 - 60.7 - 155 - 427 - 1,102 - 3,276 - 7,816 - 30,827 - Terminal 0.338 - 1.12 - 2.23 - 11.0 - 32.8 - 108 - 324 - 1,074 - This document is a draftfor review purposes only and does not constitute Agency policy. C-120 DRAFT--DO NOT CITE OR QUOTE 5000 20000 Emond CADM Emond CADM Dose (ng/kg-day) Model Emond 0.6 CADM Emond 2 CADM Emond 4 CADM Emond 20 CADM Emond 60 CADM Emond 200 CADM Emond 600 CADM 2000 Emond CADM 5000 Emond CADM 20000 Emond CADM 1 2 3 C .3 .1 .1 3 . K o c ib a e t al. (1 9 7 6 ) 2,621 2,806 (@ 7 hours) -- 10,468 11,215 (@ 7 hours) -- B O U N D L IV E R (ng/kg) Time-weighted Ave Metric Max 0.216 0.309 (@ 3 hours) -- 0 .6 6 8 0.975 (@ 3 hours) -- 1.25 1.86 (@ 3 hours) -- 4.87 7.67 (@ 2 hours) -- 11.2 18.3 (@ 2 hours) -- 25.1 40.8 (@ 1 hours) -- 45.8 68.2 (@ 1 hours) -- 73.3 93.1 (@ 1 hours) -- 90.9 104 (@ 1 hours) -- 106 110 (@ 1 hours) -- T y p e: S train: R ats Sprague-D aw ley (S p a rta n ) D ose: R oute: B ody w eight: 170-190 g (bw =180g) R egim e: Sex: Fem ale Sim ulation tim e: 1, 10, 100, 1000 ng/kg-day D iet exposure 5 d a y s/w e e k fo r 13 w eek s 2,184 hours (13w k exposed) 2,680 - 10,693 - Terminal 0.159 - 0.494 - 0.927 - 3.66 - 8.55 - 19.7 - 37.6 - 64.7 - 84.7 - 104 - This document is a draftfor review purposes only and does not constitute Agency policy. C-121 DRAFT--DO NOT CITE OR QUOTE 1 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal 0.714 Emond CADM 0.859 - 1.38 (@ 2,112 hours) - 1.13 - 7.143 Emond CADM 4.61 - 7.62 (@ 2,112 hours) - 5.27 - W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal 71.43 Emond CADM 25.3 - 48.8 (@ 2,112 hours) - 26.6 - 714.3 Emond CADM 181 - 403 (@ 2,112 hours) - 184 - L IV E R C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal 0.714 Emond CADM 88.3 89.0 140 (@ 2,116 hours) 192 126 12.1 7.143 Emond CADM 888 970 1,259 (@ 2,117 hours) 2,007 1,079 29.0 71.43 Emond CADM 8,776 9,841 11,693 (@ 2,117 hours) 20,170 9,756 88.0 714.3 Emond CADM 86,329 98,617 112,580 (@ 2,117 hours) 201,814 92,835 455 F A T C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal 0.714 Emond CADM 79.4 120 114 (@ 2,129 hours) 190 111 43.0 7.143 Emond CADM 427 456 553 (@ 2,124 hours) 787 528 67.0 71.43 Emond CADM 2,348 3,036 2,925 (@ 2,121 hours) 5,748 2,720 117 This document is a draftfor review purposes only and does not constitute Agency policy. C-122 DRAFT--DO NOT CITE OR QUOTE 714.3 Emond CADM 16,815 28,382 21,126 (@ 2,120 hours) 55,013 B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose 0.714 Model Emond CADM Time-weighted Ave 10.8 11.5 Metric Max 16.1 (@ 2,116 hours) 20.0 7.143 Emond CADM 76.9 65.3 105 (@ 2,116 hours) 126 71.43 Emond CADM 600 553 785 (@ 2,116 hours) 1,113 714.3 Emond CADM 5,366 5,401 6,960 (@ 2,116 hours) 10,967 Dose (ng/kg-day) Adjusted dose Model B O U N D L IV E R (ng/kg) Time-weighted Ave Metric Max 0.714 Emond CADM 2.89 - 4.17 (@ 2,116 hours) - 7.143 Emond CADM 13.7 - 17.5 (@ 2,116 hours) - 71.43 Emond CADM 47.5 - 55.0 (@ 2,116 hours) - 714.3 Emond CADM 93.4 - 1 2 3 C .3 .1 .1 4 . K o c ib a e t al. (1 9 7 8 ) F e m a le 98.2 (@ 2,117 hours) - 19,233 274 Terminal 15.1 3.75 93.6 6.22 673 12.0 5,842 37.0 Terminal 3.81 - 15.7 - 50.6 - 95.7 - T y p e: S train: R ats Sprague-D aw ley (S p a rta n ) D ose: R oute: B ody w eight: 170-190 g (bw =180) R egim e: Sex: 4 Fem ale Sim ulation tim e: 0 , 1, 10, 100 ng/kg-day D iet exposure 7 days/w eek for 104 w eeks 17,472 hours This document is a draftfor review purposes only and does not constitute Agency policy. C-123 DRAFT--DO NOT CITE OR QUOTE Dose (ng/kg-day) Adjusted dose 1 10 100 Dose (ng/kg-day) Adjusted dose 1 10 100 Dose (ng/kg-day) Adjusted dose 1 10 100 Dose (ng/kg-day) Adjusted dose 1 10 Model W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max Emond CADM Emond CADM Emond CADM Model 1.55 1.92 (@ 17,448 hours) -7.15 9.25 (@ 17,448 hours) -38.6 57.5 (@ 17,448 hours) -L IV E R C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max Emond CADM Emond CADM Emond CADM Model 192 226 (@ 17,452 hours) 292 333 1,618 1,742 (@ 17,452 hours) 2,981 3,342 14,892 15,673 (@ 17,452 hours) 29,917 33,432 F A T C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max Emond CADM Emond CADM Emond CADM Model 147 165 (@ 17,457 hours) 196 229 680 713 (@ 17,454 hours) 861 1,015 3,663 3,788 (@ 17,454 hours) 6,756 7,939 B O D Y B U R D E N (ng/kg) Metric Time-weighted Ave Max Emond CADM Emond CADM 21.2 26.0 131 169 24.3 (@ 17,452 hours) 27.0 140 (@ 17,452 hours) 176 Terminal 1.69 - 7.16 - 37.1 - Terminal 218 333 1,665 3,342 14,907 33,432 Terminal 164 181 706 789 3,731 6,203 Terminal 23.8 27.0 136 176 This document is a draftfor review purposes only and does not constitute Agency policy. C-124 DRAFT--DO NOT CITE OR QUOTE Emond 100 CADM 989 1,546 1,039 (@ 17,452 hours) 1,601 B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 1 CADM 5.11 - 5.77 (@ 17,452 hours) - Emond 10 CADM 20.0 - 21.1 (@ 17,452 hours) - Emond 100 CADM 59.9 - 61.5 (@ 17,452 hours) - 1 2 3 C .3 .1 .1 5 . K o c ib a e t al. (1 9 7 8 ) M a le 994 1,601 Terminal 5.59 - 20.4 - 60.1 - T y p e: S train: R ats Sprague-D aw ley (S p a rta n ) D ose: R oute: 0 , 1, 10, 100 ng/kg-day D iet exposure B ody w eight: B ody w eight approxim ated to be 250 g R egim e: 7 days/w eek for 104 w eeks Sex: 4 M ale Sim ulation tim e: 17,472 hours W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 1 CADM Emond 10 CADM Emond 100 CADM Dose (ng/kg-day) Adjusted dose Model 1.56 1.96 (@ 17,448 hours) 7.16 9.35 (@ 17,448 hours) -38.7 59.3 (@ 17,448 hours) -L IV E R C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max 1.70 - 7.11 - 37.1 - Terminal Emond 1 CADM 194 - 229 (@ 17,452 hours) - 221 - This document is a draftfor review purposes only and does not constitute Agency policy. C-125 DRAFT--DO NOT CITE OR QUOTE 10 Em ond CADM 100 Em ond CADM Dose (ng/kg-day) Adjusted dose 1 10 100 Model Emond CADM Emond CADM Emond CADM Dose (ng/kg-day) Adjusted dose 1 10 100 Model Emond CADM Emond CADM Emond CADM Dose (ng/kg-day) Adjusted dose 1 10 100 Model Emond CADM Emond CADM Emond CADM 1,616 1,723 (@ 17,452 hours) -- 14,898 15,671 (@ 17,452 hours) -- F A T C O N C E N T R A T IO N S (ng/kg) Metric Time-weighted Ave Max 148 167 (@ 17,456 hours) -- 680 709 (@ 17,454 hours) -- 3,677 3,803 (@ 17,453 hours) -- B O D Y B U R D E N (ng/kg) Metric Time-weighted Ave Max 21.4 24.6 (@ 17,452 hours) -- 131 139 (@ 17,452 hours) -- 991 1,041 (@ 17,452 hours) -- B O U N D L IV E R (ng/kg) Metric Time-weighted Ave Max 5.15 - 20.0 60.0 - 5.83 (@ 17,452 hours) - 21.0 (@ 17,452 hours) - 61.5 (@ 17,452 hours) - 1,649 - 14,912 - Terminal 166 - 703 - 3,747 - Terminal 24.1 - 134 - 995 - Terminal 5.64 - 20.3 - 60.1 - This document is a draftfor review purposes only and does not constitute Agency policy. C-126 DRAFT--DO NOT CITE OR QUOTE 1 C .3 .1 .1 6 . L a tc h o u m y c a n d a n e a n d M a th u r (2 0 0 2 ) T y p e: S train: Rat W istar D ose: R oute: 0 , 1, 10, 100 ng/kg-day Oral gavage B ody w eight: 45 days old (BW set to 200g) R egim e: 1/day fo r 45 days Sex: 2 M ale Sim ulation tim e: 1,080 hours (daily exposure) W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 0.785 1.37 (@ 1,056 hours) 1 CADM - - Emond 4.65 10 CADM - 8.18 (@ 1,056 hours) - Emond 100 CADM 27.3 - 53.9 (@ 1,056 hours) - L IV E R C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 78.5 1 CADM 116 138 (@ 1,060 hours) 217 Emond 902 10 1,423 (@ 1,060 hours) CADM 1,669 2,550 Emond 9,579 14,015 (@ 1,061 hours) 100 CADM 17,681 25,915 Dose (ng/kg-day) Adjusted dose Model F A T C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max Emond 69.8 1 CADM 150 113 (@ 1,072 hours) 220 Emond 416 10 CADM 744 608 (@ 1,065 hours) 1,009 Emond 2,448 100 CADM 5,719 3,425 (@ 1,062 hours) 7,866 Terminal 1.18 - 6.18 - 33.8 - Terminal 133 217 1,358 2,550 13,306 25,915 Terminal 113 220 604 1,009 3,380 7,866 This document is a draftfor review purposes only and does not constitute Agency policy. C-127 DRAFT--DO NOT CITE OR QUOTE Dose (ng/kg-day) Adjusted dose 1 10 100 Model Emond CADM Emond CADM Emond CADM Dose (ng/kg-day) Adjusted dose Model Emond 1 CADM Emond 10 CADM Emond 100 CADM 1 2 3 C .3 .1 .1 7 . L i e t al. (1 9 9 7 ) B O D Y B U R D E N (ng/kg) Metric Time-weighted Ave Max 9.56 15.9 (@ 1,060 hours) 14.0 2 2 .2 76.7 117 (@ 1,060 hours) 106 157 646 933 (@ 1,060 hours) 988 1,439 B O U N D L IV E R (ng/kg) Metric Time-weighted Ave Max 2.64 - 13.7 - 48.6 - 4.12 (@ 1,060 hours) - 18.8 (@ 1,060 hours) - 59.0 (@ 1,060 hours) - Terminal 15.6 22.2 113 157 891 1,439 Terminal 3.96 - 18.1 - 57.5 - T y p e: R ats D ose: 0, 3, 10, 30, 100, 300, 1000, 3000, 10000, 30000 ng/kg/day S train: Sprague-D aw ley R oute: G astric intubation B ody w eight: 22 day old, 55 to 58 g (B W set to 56.5 g) R egim e: One dose for one day Sex: Fem ale Sim ulation tim e: 24 hours 4 The C A D M m odel w as not ru n because the dosing duratio n is low er th an the resolution o f the m odel (1 w eek) Dose (ng/kg-day) 3 10 W H O L E B L O O D C O N C E N T R A T IO N S (ng/kg) Model Time-weighted Ave Metric Max Emond 0.266 0.470 (@ 1 hours) CADM - - Emond 0.799 1.57 (@ 1 hours) CADM - - Terminal 0.180 - 0.535 - This document is a draftfor review purposes only and does not constitute Agency policy. C-128 DRAFT--DO NOT CITE OR QUOTE 30 100 300 1,000 3,000 10,000 30,000 Dose (ng/kg-day) 3 10 30 100 300 1,000 3,000 10,000 30,000 Emond 2.10 4.68 (@ 1 hours) CADM - - Emond 5.87 15.6 (@ 1 hours) CADM - - Emond 15.0 46.8 (@ 0 hours) CADM - - Emond 43.3 156 (@ 0 hours) CADM Emond -- 120 469 (@ 0 hours) CADM - - Emond 386 1,570 (@ 0 hours) CADM - - Emond 1,172 4,762 (@ 0 hours) CADM - - L IV E R C O N C E N T R A T IO N S (ng/kg) Model Time-weighted Ave Metric Max Emond 14.7 18.6 (@ 4 hours) CADM - - Emond 55.0 65.2 (@ 5 hours) CADM Emond -- 185 210 (@ 6 hours) CADM - - Emond 690 768 (@ 7 hours) CADM - - Emond 2,248 2,473 (@ 8 hours) CADM - - Emond 7,938 8,671 (@ 9 hours) CADM - - Emond 24,474 26,639 (@ 9 hours) CADM - - Emond 82,349 89,464 (@ 9 hours) CADM - - Emond 245,610 265,670 (@ 10 hours) CADM - - 1.37 - 3.68 - 8.83 - 23.4 - 59.9 - 182 - 535 - Terminal 11.9 - 47.6 - 170 - 666 2,240 8,094 25,267 85,597 255,390 - This document is a draftfor review purposes only and does not constitute Agency policy. C-129 DRAFT--DO NOT CITE OR QUOTE Dose (ng/kg-day) 3 10 30 100 300 1,000 3,000 10,000 30,000 Dose (ng/kg-day) 3 10 30 100 300 F A T C O N C E N T R A T IO N S (ng/kg) Model Time-weighted Ave Metric Max Emond 8.75 12.7 (@ 24 hours) CADM - - Emond 26.6 38.0 (@ 24 hours) CADM - - Emond 70.8 98.9 (@ 24 hours) CADM - - Emond 202 273 (@ 24 hours) CADM - - Emond 530 689 (@ 24 hours) CADM - - Emond 1,573 1,958 (@ 24 hours) CADM - - Emond 4,433 5,358 (@ 24 hours) CADM - - Emond 14,428 17,119 (@ 24 hours) CADM - - Emond 44,361 51,948 (@ 22 hours) CADM - - B O D Y B U R D E N (ng/kg) Model Time-weighted Ave Metric Max Emond 1.60 1.70 (@ 8 hours) CADM - - Emond 5.33 5.66 (@ 8 hours) CADM - - Emond 15.9 16.9 (@ 8 hours) CADM - - Emond 52.8 56.2 (@ 7 hours) CADM - - Emond 158 169 (@ 7 hours) CADM - - Terminal 12.7 - 38.0 - 98.9 - 273 - 689 - 1,958 - 5,358 - 17,119 - 51,898 - Terminal 1.68 - 5.56 - 16.5 - 54.5 - 163 - This document is a draftfor review purposes only and does not constitute Agency policy. C-130 DRAFT--DO NOT CITE OR QUOTE 1,000 3,000 10,000 30,000 Dose (ng/kg-day) 3 10 30 100 300 1,000 3,000 10,000 30,000 Emond CADM Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM 525 561 (@ 7 hours) -- 1,574 1,684 (@ 7 hours) -- 5,240 5,610 (@ 7 hours) -- 15,758 16,815 (@ 7 hours) -- B O U N D L IV E R (ng/kg) Time-weighted Ave Metric Max 0.89 1.37 (@ 3 hours) -- 2.58 4.10 (@ 2 hours) -- 6.37 10.5 (@ 2 hours) -- 15.54 25.9 (@ 2 hours) -- 31.25 50.1 (@ 1 hours) -- 56.75 79.8 (@ 1 hours) -- 81.28 98.4 (@ 1 hours) -- 99.77 108 (@ 1 hours) -- 107.69 111 (@ 1 hours) -- 539 - 1,611 - 5,360 - 16,041 - Terminal 0.64 - 1.88 - 4.71 - 11.77 - 24.57 - 47.62 - 73.32 - 95.68 - 106.24 - This document is a draftfor review purposes only and does not constitute Agency policy. C-131 DRAFT--DO NOT CITE OR QUOTE 1 C.3.1.18. M urray et al. (1979) A d u lt Portion T y p e: S train: Rat Sprague D aw ley D ose: R oute: 1, 10, and 100 ng/kg-day D iet oral dose B o d y w eig h t: B W set to 4.5 g R egim e: O nce per day for 120 days Sex: Fem ale Sim ulation tim e: 2880 hours 2 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose 1 Model Emond CADM Time-weighted Ave 1.12 - Metric Max 1.51 (@ 2,856 hours) - 10 Em ond CADM 5.88 7.59 (@ 2,856 hours) -- 100 Em ond CADM 32.7 44.3 (@ 2,856 hours) -- L IV E R C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max 1 10 100 Dose (ng/kg-day) Adjusted dose Emond CADM Emond CADM Emond CADM Model 128 180 (@ 2,859 hours) -- 1,273 1,618 (@ 2,860 hours) -- 12,601 15,281 (@ 2,860 hours) -- F A T C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max 1 Emond CADM 10 Em ond CADM 100 Em ond CADM 106 - 556 - 3,095 - 139 (@ 2,865 hours) - 665 (@ 2,864 hours) - 3,604 (@ 2,862 hours) - Terminal 1.42 - 6.75 - 36.0 - Terminal 173 - 1,540 - 14,460 - Terminal 138 - 657 - 3,534 - This document is a draftfor review purposes only and does not constitute Agency policy. C-132 DRAFT--DO NOT CITE OR QUOTE B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 1 CADM 14.8 2 0 .0 (@ 2,860 hours) -- Emond 10 CADM 105 130 (@ 2,860 hours) -- Emond 100 CADM 837 1,003 (@ 2,860 hours) -- B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 1 CADM 3.77 - 4.95 (@ 2,859 hours) - E m ond 17.1 20.3 (@ 2,859 hours) 10 CADM - - Emond 100 CADM 55.3 - 60.9 (@ 2,860 hours) - 1 2 3 C .3 .1 .1 9 . N T P (1 9 8 2 )-- F e m a le R a ts , C h r o n ic Terminal 19.6 - 126 - 957 - Terminal 4.77 - 19.5 - 59.4 - T y p e: R at D ose: S train: B ody w eight Sex: O sborne-M endel R oute: 6 w eeks old R egim e: (B W set to 250g) Fem ale S im ulation tim e 10, 50 and 500 ng/kg/w k, tw o doses per w eek Oral exposure B iw eekly (Sim ulation has been perform using fem ale B W 17,472 hours (104 w eeks o f exposure) W H O L E B L O O D C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 1.4 CADM 1.96 3.11 (@ 17,220 hours) 1.94 - -- Emond 5.69 11.0 (@ 17,388 hours) 5.40 7.1 CADM - -- This document is a draftfor review purposes only and does not constitute Agency policy. C-133 DRAFT--DO NOT CITE OR QUOTE 71 Dose (ng/kg-day) Adjusted dose 1.4 7.1 71 Dose (ng/kg-day) Adjusted dose 1.4 7.1 71 Dose (ng/kg-day) Adjusted dose 1.4 7.1 71 Dose (ng/kg-day) Adjusted dose 1.4 Emond 29.8 82.2 (@ 17,388 hours) CADM - - L IV E R C O N C EN TR ATIO N S (ng/kg) Model Time-weighted Ave Metric Max Emond CADM Emond CADM Emond CADM Model 265 308 (@ 17,226 hours) 15,318 20,170 1,175 1,338 (@ 17,394 hours) 30,700 40,353 10,734 12,182 (@ 17,395 hours) 30,700 40,353 F A T C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max Emond CADM Emond CADM Emond CADM Model 186 200 (@ 17,328 hours) 4,655 5,748 541 569 (@ 17,409 hours) 9,064 11,224 2,826 2,973 (@ 17,404 hours) 17,879 B O D Y B U R D E N (ng/kg) 22,172 Metric Time-weighted Ave Max Emond CADM Emond CADM Emond CADM Model 27.9 31.1 (@ 17,225 hours) 855 1,113 99.4 110 (@ 17,393 hours) 1,695 2,208 729 814 (@ 17,393 hours) 3,375 4,395 B O U N D L IV E R (ng/kg) Metric Time-weighted Ave Max Emond CADM 6.37 7.26 (@ 17,224 hours) -- 26.9 - Terminal 265 7,102 1,117 14,200 9,882 14,200 Terminal 193 2,107 544 3,964 2,769 7,671 Terminal 28.4 403 96.7 787 683 1,556 Terminal 6.38 - This document is a draftfor review purposes only and does not constitute Agency policy. C-134 DRAFT--DO NOT CITE OR QUOTE Emond 7.1 CADM 16.6 - Emond 71 CADM 52.7 - 1 2 3 C .3 .1 .2 0 . N T P (1 9 8 2 )-- M a le R a ts , C h r o n ic 18.5 (@ 17,392 hours) - 56.4 (@ 17,393 hours) - 16.1 - 50.9 - T y p e: R at D ose: 10, 50 and 500 ng/kg/w k, tw o doses per w eek S train: O sbom e-M endel R oute: Oral exposure B ody w eight 6 w eeks old R egim e: (B W set to 350g) B iw eekly (Sim ulation has been perform using fem ale B W Sex: 4 M ale S im u la tio n tim e 17,472 hours (104 w eeks o f exposure) W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 1.4 CADM 1.96 3.18 (@ 17,388 hours) 1.93 - -- Emond 7.1 CADM 5.70 11.4 (@ 17,388 hours) 5.39 - -- Emond 71 CADM 29.9 - 87.0 (@ 17,388 hours) - 26.9 - L IV E R C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 1.4 CADM 265 306 (@ 17,394 hours) 263 - -- L IV E R C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 7.1 CADM 1,174 - 1,334 (@ 17,394 hours) - 1,114 - Emond 71 CADM 10,736 - 12,170 (@ 17,395 hours) - 9,881 - This document is a draftfor review purposes only and does not constitute Agency policy. C-135 DRAFT--DO NOT CITE OR QUOTE Dose (ng/kg-day) Adjusted dose 1.4 7.1 71 Dose (ng/kg-day) Adjusted dose 1.4 7.1 71 Dose (ng/kg-day) Adjusted dose 1.4 7.1 71 1 Model Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM F A T C O N C E N T R A T IO N S (ng/kg) Metric Time-weighted Ave Max 186 199 (@ 17,412 hours) -- 541 569 (@ 17,409 hours) -- 2,836 2,983 (@ 17,404 hours) -- B O D Y B U R D E N (ng/kg) Metric Time-weighted Ave Max 27.8 30.9 (@ 17,393 hours) -- 99.5 110 (@ 17,393 hours) -- 730 816 (@ 17,393 hours) -- B O U N D L IV E R (ng/kg) Metric Time-weighted Ave Max 6.36 - 16.6 - 52.7 - 7.22 (@ 17,392 hours) - 18.4 (@ 17,392 hours) - 56.3 (@ 17,393 hours) - Terminal 193 - 544 - 2,784 - Terminal 28.2 - 96.6 - 684 - Terminal 6.35 - 16.0 - 50.9 - This document is a draftfor review purposes only and does not constitute Agency policy. C-136 DRAFT--DO NOT CITE OR QUOTE 1 C.3.1.21. N T P (1982)-- F em ale M ice, Chronic T y p e: M ice D ose: 40, 200 and 2000 ng/kg/w k, tw o doses during the w eek S train: B6C3F1 R oute: Oral exposure B ody w eight 6 w eeks old R egim e: (B W set to 23g) B iw eekly (Sim ulation has been perform using fem ale BW ) Sex: Fem ale S im u la tio n tim e 17,472 hours (104 w eeks o f exposure) 2 * The mice chronic exposure could not be sim ulated w ith the CADM model because this model simulates for only 3 123 days. 4 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 5.7 CADM 1.95 4.86 (@ 16,800 hours) 1.82 - -- Emond 28.6 CADM 5.84 19.8 (@ 17,388 hours) 5.17 - -- Emond 286 32.1 171 (@ 16,884 hours) 26.0 CADM - -- L IV E R C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 5.7 CADM 490 582 (@ 16,807 hours) 463 - -- Emond 28.6 CADM 2,236 - 2,629 (@ 17,395 hours) - 2,025 - Emond 286 CADM 20,841 - 24,353 (@ 17,396 hours) - 18,182 - F A T C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 5.7 CADM 737 785 (@ 17,408 hours) 757 - -- Emond 28.6 CADM 2,213 - 2,337 (@ 17,404 hours) - 2,216 - Emond 286 CADM 12,138 - 12,861 (@ 17,400 hours) - 11,775 - This document is a draftfor review purposes only and does not constitute Agency policy. C-137 DRAFT--DO NOT CITE OR QUOTE B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 5.7 CADM 91.9 103 (@ 17,393 hours) -- Emond 28.6 CADM 329 370 (@ 17,393 hours) -- Emond 286 CADM 2,400 - 2,740 (@ 17,393 hours) - B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 5.7 CADM 6.18 - 7.29 (@ 16,805 hours) - 28.6 Emond CADM 16.3 - 18.9 (@ 17,393 hours) - Emond 286 CADM 52.3 - 67.8 (@ 2 hours) - 1 2 3 C .3 .1 .2 2 . N T P (1 9 8 2 )-- M a le M ic e , C h r o n ic Terminal 91.2 - 313 - 2,176 - Terminal 5.93 - 15.3 - 49.3 - T y p e: M ice D ose: 10, 50 and 500ng/kg/w k, tw o doses during the w eek S train: B6C3F1 R oute: Oral exposure B ody w eight 6 w eeks old (B W set to 25g) R egim e: B iw eekly Sex: M ale S im u la tio n tim e 17,472 hours (104 w eeks o f exposure) 4 * The mice chronic exposure could not be sim ulated w ith the CADM model because this model simulates for only 5 123 days. 6 W H O L E B L O O D C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 1.4 CADM 0.767 - 1.53 (@ 17,304 hours) - 0.749 - Emond 7.1 CADM 2.27 5.99 (@ 17,052 hours) 2.11 - -- This document is a draftfor review purposes only and does not constitute Agency policy. C-138 DRAFT--DO NOT CITE OR QUOTE 71 Dose (ng/kg-day) Adjusted dose 1.4 7.1 71 Dose (ng/kg-day) Adjusted dose 1.4 7.1 71 Dose (ng/kg-day) Adjusted dose 1.4 7.1 71 Dose (ng/kg-day) Adjusted dose 1.4 Emond CADM Model Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Model Emond CADM 11.2 46.7 (@ 17,388 hours) -- L IV E R C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max 138 165 (@ 17,310 hours) -- 606 722 (@ 17,059 hours) -- 5,409 6,328 (@ 17,395 hours) -- F A T C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max 290 314 (@ 17,411 hours) -- 860 918 (@ 17,155 hours) -- 4,257 4,490 (@ 17,402 hours) -- B O D Y B U R D E N (ng/kg) Metric Time-weighted Ave Max 32.3 - 110 - 710 - BO U N D L IV E R (ng/kg) 36.2 (@ 17,309 hours) - 123 (@ 17,057 hours) - 802 (@ 17,393 hours) - Metric Time-weighted Ave Max 2.56 3.03 (@ 17,309 hours) -- 9.59 - Terminal 136 - 571 - 4,805 - Terminal 306 - 883 - 4,204 - Terminal 33.3 - 108 - 660 - Terminal 2.53 - This document is a draftfor review purposes only and does not constitute Agency policy. C-139 DRAFT--DO NOT CITE OR QUOTE 7.1 71 1 2 Emond CADM Emond CADM 3 C .3 .1 .2 3 . N T P (2 0 0 6 ) 1 4 W e e k s 7.12 - 27.1 - 8.40 (@ 17,057 hours) - 32.4 (@ 2 hours) - 6.82 - 25.3 - T y p e: S train: Rat Sprague D aw ley D ose: R oute: 0, 3, 10, 22, 46, 100 ng/kg-day Oral gavage Body w eight: 8 w eeks old (BW =215g) R egim e: 5 days/w eeks fo r 14 w eeks Sex: 4 Fem ale and m ale S im u latio n tim e: 2,352 hours (14 w eeks) W H O L E B L O O D C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal 2.14 Em ond 1.98 CADM - 7.14 Em ond 4.69 3.15 (@ 2,280 hours) - 7.75 (@ 2,280 hours) 2.39 - 5.30 CADM - - - 15.7 Em ond 8.27 14.3 (@ 2,280 hours) 9.02 CADM - - - 32.9 Emond 14.2 25.9 (@ 2,280 hours) 15.1 CADM - - - 71.4 Emond 25.7 49.8 (@ 2,280 hours) 26.6 CADM - - - Dose (ng/kg-day) Adjusted dose Model L IV E R C O N C E N T R A T IO N S (ng/kg) Metric Time-weighted Ave Max Terminal 2.14 Em ond 275 404 (@ 2,284 hours) 354 CADM - - - 7.14 Em ond 909 1,270 (@ 2,285 hours) 1,089 CADM - - - 15.7 Em ond 1,988 2,703 (@ 2,285 hours) 2,291 CADM - - - This document is a draftfor review purposes only and does not constitute Agency policy. C-140 DRAFT--DO NOT CITE OR QUOTE 32.9 Emond 4,129 5,508 (@ 2,285 hours) 4,628 CADM - - - 71.4 Emond 8,921 11,734 (@ 2,285 hours) 9,792 CADM - - - F A T C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal 2.14 Em ond 184 246 (@ 2,294 hours) 237 CADM - - - 7.14 Em ond 436 557 (@ 2,292 hours) 532 CADM - - - 15.7 Em ond 768 962 (@ 2,291 hours) 912 CADM - - - 32.9 Emond 1,319 1,633 (@ 2,289 hours) 1,535 CADM - - - 71.4 Emond 2,385 2,938 (@ 2,289 hours) 2,736 CADM - - - B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal 2.14 Em ond 28.2 39.4 (@ 2,284 hours) 36.1 CADM - - - 7.14 Em ond 78.5 106 (@ 2,284 hours) 94.4 CADM - - - 15.7 Em ond 156 206 (@ 2,284 hours) 181 CADM - - - 32.9 Emond 300 391 (@ 2,284 hours) 340 CADM - - - 71.4 Emond 610 788 (@ 2,284 hours) 676 CADM - - - B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal 2.14 Em ond 6.41 8.55 (@ 2,284 hours) 7.74 CADM - - - This document is a draftfor review purposes only and does not constitute Agency policy. C-141 DRAFT--DO NOT CITE OR QUOTE 7.14 Em ond CADM 15.7 Em ond CADM 32.9 Emond CADM 71.4 Emond CADM 1 2 3 C .3 .1 .2 4 . N T P (2 0 0 6 ) 3 1 W e e k s 13.9 - 22.2 - 33.2 - 47.9 - 17.6 (@ 2,284 hours) - 27.2 (@ 2,284 hours) - 39.3 (@ 2,284 hours) - 55.1 (@ 2,284 hours) - 15.8 - 24.5 - 35.7 - 50.7 - T y p e: S train: Rat Sprague D aw ley D ose: R oute: 0, 3, 10, 22, 46, 100 ng/kg-day Oral gavage Body w eight: 8 w eeks old (BW =215g) R egim e: 5 days/w eeks for 31 w eeks Sex: Fem ale and m ale S im u la tio n tim e : 5,208 hours (31 w eeks) 4 W H O L E B L O O D C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.33 2.14 CADM - 3.25 (@ 3,960 hours) - 2.48 - Emond 5.32 7.14 CADM - 7.89 (@ 3,960 hours) - 5.40 - Emond 15.7 CADM 9.21 - 14.5 (@ 3,960 hours) - 9.15 - 32.9 71.4 Emond CADM Emond CADM 15.7 - 28.1 - 26.2 (@ 5,136 hours) - 50.4 (@ 5,136 hours) - 15.3 - 27.0 - Dose (ng/kg-day) Adjusted dose Model L IV E R C O N C E N T R A T IO N S (ng/kg) Metric Time-weighted Ave Max Terminal Emond 2.14 CADM 341 - 425 (@ 5,140 hours) - 373 - Emond 1,075 1,308 (@ 3,965 hours) 1,117 7.14 CADM - - - This document is a draftfor review purposes only and does not constitute Agency policy. C-142 DRAFT--DO NOT CITE OR QUOTE Emond 2,296 2,756 (@ 3,965 hours) 2,336 15.7 CADM - - - 32.9 Emond CADM 4,696 - 5,597 (@ 5,141 hours) - 4,712 - 71.4 Emond 10,033 11,905 (@ 5,141 hours) 9,953 CADM - - - F A T C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.14 CADM 220 - 256 (@ 5,149 hours) - 246 - Emond 7.14 CADM 501 - 570 (@ 4,139 hours) - 542 - Emond 15.7 CADM 868 - 978 (@ 4,138 hours) - 926 - 32.9 Emond CADM 1,476 - 1,657 (@ 5,145 hours) - 1,558 - 71.4 Emond CADM 2,652 - 2,978 (@ 5,144 hours) - 2,775 - B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.14 CADM 34.2 - 41.2 (@ 5,140 hours) - 37.8 - Emond 7.14 CADM 91.6 - 108 (@ 3,964 hours) - 96.6 - Emond 15.7 CADM 178 - 209 (@ 3,964 hours) - 184 - 32.9 Emond CADM 339 - 398 (@ 5,140 hours) - 346 - 71.4 Emond CADM 682 - 799 (@ 5,140 hours) - 687 - B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.14 CADM 7.48 - 8.83 (@ 5,140 hours) - 8.01 - 7.14 Em ond 15.6 17.9 (@ 3,964 hours) 16.1 This document is a draftfor review purposes only and does not constitute Agency policy. C-143 DRAFT--DO NOT CITE OR QUOTE CADM Emond 15.7 CADM 32.9 Emond CADM 71.4 Emond CADM 1 2 3 C .3 .1 .2 5 . N T P (2 0 0 6 ) 5 3 W e e k s 24.3 35.7 50.9 - 27.4 (@ 3,964 hours) 39.6 (@ 5,140 hours) 55.4 (@ 5,140 hours) - 24.8 36.0 51.1 - T y p e: S train: Rat Sprague D aw ley D ose: R oute: 0, 3, 10, 22, 46, 100 ng/kg-day Oral gavage Body w eight: 8 w eeks old (BW =215g) R egim e: 5 days/w eeks fo r 105 w eeks Sex: Fem ale and m ale Sim ulation tim e: 8,904 hours (53 w eeks) 4 W H O L E B L O O D C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.46 2.14 CADM - 3.25 (@ 6,312 hours) - 2.48 - Emond 5.53 7.14 CADM - 7.89 (@ 3,960 hours) - 5.41 - Emond 15.7 CADM 9.54 - 14.5 (@ 8,832 hours) - 9.17 - 32.9 Emond CADM 16.2 - 26.3 (@ 8,832 hours) - 15.3 - 71.4 Emond CADM 29.0 - 50.6 (@ 8,832 hours) - 27.1 - L IV E R C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.14 CADM 366 - 426 (@ 6,316 hours) - 373 - Emond 1,134 1,308 (@ 3,965 hours) 1,121 7.14 CADM - - - Emond 2,406 2,759 (@ 8,837 hours) 2,345 15.7 CADM - - - This document is a draftfor review purposes only and does not constitute Agency policy. C-144 DRAFT--DO NOT CITE OR QUOTE 32.9 Emond CADM 4,902 - 5,612 (@ 8,837 hours) - 4,727 - 71.4 Emond CADM 10,439 - 11,938 (@ 8,837 hours) - 9,985 - F A T C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.14 CADM 233 - 256 (@ 6,325 hours) - 247 - Emond 7.14 CADM 524 - 570 (@ 4,139 hours) - 544 - Emond 15.7 CADM 904 - 980 (@ 8,842 hours) - 929 - 32.9 Emond CADM 1,533 - 1,661 (@ 8,841 hours) - 1,562 - 71.4 Emond CADM 2,749 - 2,986 (@ 8,840 hours) - 2,784 - B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.14 CADM 36.4 - 41.2 (@ 6,316 hours) - 37.8 - Emond 7.14 CADM 96.1 - 108 (@ 3,964 hours) - 96.9 - Emond 15.7 CADM 186 - 210 (@ 8,836 hours) - 185 - 32.9 Emond CADM 353 - 399 (@ 8,836 hours) - 347 - 71.4 Emond CADM 709 - 801 (@ 8,836 hours) - 689 - B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.14 CADM 7.87 - 8.84 (@ 6,316 hours) - 8.01 - Emond 7.14 CADM 16.2 - 17.9 (@ 3,964 hours) - 16.1 - 15.7 Em ond 25.1 27.5 (@ 8,836 hours) 24.8 This document is a draftfor review purposes only and does not constitute Agency policy. C-145 DRAFT--DO NOT CITE OR QUOTE CADM 32.9 Emond CADM 71.4 Emond CADM 1 2 3 C .3 .1 .2 6 . N T P (2 0 0 6 ) 2 Y e a rs 36.6 51.9 - 39.7 (@ 8,836 hours) 55.4 (@ 8,836 hours) - 36.1 51.1 - T y p e: S train: Rat Sprague D aw ley D ose: R oute: 0, 3, 10, 22, 46, 100 ng/kg-day Oral gavage Body w eight: 8 w eeks old (BW =215g) R egim e: 5 days/w eeks fo r 105 w eeks Sex: Fem ale and m ale Sim ulation tim e: 17,640 hours* (105 w eeks) 4 *The CADM model simulates for 104 weeks only (17,472 hours). As a result, the term inal values from the CADM 5 m odel may be underestim ated com pared to the Em ond model, w hich considers the full 105 w eeks o f exposure. W H O L E B L O O D C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.14 CADM 2.56 - 3.47 (@ 17,568 hours) - 2.62 - Emond 7.14 CADM 5.69 - 7.97 (@ 17,568 hours) - 5.46 - Emond 15.7 CADM 9.79 - 14.6 (@ 17,568 hours) - 9.22 - Emond 32.9 CADM 16.6 - 26.4 (@ 17,568 hours) - 15.4 - Emond 71.4 CADM 29.7 - 50.8 (@ 17,568 hours) - 27.1 - L IV E R C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.14 CADM 385 632 460 (@ 17,572 hours) 715 403 715 Emond 7.14 CADM 1,177 2,127 1,320 (@ 17,573 hours) 2,387 1,135 2,387 15.7 Em ond 2,487 2,779 (@ 17,573 hours) 2,361 This document is a draftfor review purposes only and does not constitute Agency policy. C-146 DRAFT--DO NOT CITE OR QUOTE CADM 4,691 5,252 5,252 Emond 32.9 CADM 5,051 9,822 5,637 (@ 17,573 hours) 10,984 4,749 10,984 Emond 71.4 CADM 10,734 21,366 11,976 (@ 17,573 hours) 23,880 10,018 23,880 F A T C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.14 CADM 243 302 271 (@ 17,581 hours) 355 261 277 Emond 7.14 CADM 541 667 575 (@ 17,579 hours) 787 549 611 Emond 15.7 CADM 930 1,242 985 (@ 17,578 hours) 1,463 934 1,138 Emond 32.9 CADM 1,574 2,369 1,667 (@ 17,577 hours) 2,787 1,568 2,173 Emond 71.4 CADM 2,821 4,890 2,995 (@ 17,576 hours) 5,748 2,792 4,489 B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 2.14 CADM 38.1 46.0 44.0 (@ 17,572 hours) 48.0 40.4 48.0 Emond 7.14 CADM 99.5 125 109 (@ 17,572 hours) 130 97.9 130 Emond 15.7 CADM 192 257 211 (@ 17,572 hours) 267 186 267 Emond 32.9 CADM 364 520 400 (@ 17,572 hours) 538 348 538 Emond 71.4 CADM 729 1,110 804 (@ 17,572 hours) 1,149 691 1,149 B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal 2.14 Em ond 8.17 9.30 (@ 17,572 hours) 8.43 This document is a draftfor review purposes only and does not constitute Agency policy. C-147 DRAFT--DO NOT CITE OR QUOTE CADM Emond 7.14 CADM Emond 15.7 CADM Emond 32.9 CADM Emond 71.4 CADM 1 2 3 C .3 .1 .2 7 . S e w a ll e t al. (1 9 9 5 ) 16.6 25.6 37.3 52.7 - 18.0 (@ 17,572 hours) 27.6 (@ 17,572 hours) 39.7 (@ 17,572 hours) 55.5 (@ 17,572 hours) - 16.2 24.9 36.2 51.2 - T y p e: Rat D ose: 49, 149.8, 490, and 1750 ng/kg every tw o w eeks o r 3.5, 10.7, 35, and 125 ng/kg-day S train: Sprauge-D aw ley R oute: Oral gavage B ody w eight: 12 w k old (BW set to 250g) R egim e: Once every 2 w eeks for 30 w eeks Sex: Fem ale Sim ulation tim e: 5040 hours 4 W H O L E B L O O D C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 3.5 CADM 3.29 - 13.7 (@ 4,704 hours) - 2.88 - Emond 10.7 CADM 7.11 - 38.7 (@ 4,704 hours) - 5.79 - Emond 35 CADM 16.6 - 120 (@ 4,704 hours) - 12.6 - Emond 125 CADM 44.7 - 414 (@ 4,704 hours) - 31.4 - Dose (ng/kg-day) Adjusted dose Model L IV E R C O N C E N T R A T IO N S (ng/kg) Metric Time-weighted Ave Max Terminal Emond 3.5 CADM 550 - 901 (@ 4,711 hours) - 459 - Emond 10.7 CADM 1,605 - 2,632 (@ 4,712 hours) - 1,229 - 35 E m ond 5,072 8,350 (@ 4,712 hours) 3,618 This document is a draftfor review purposes only and does not constitute Agency policy. C-148 DRAFT--DO NOT CITE OR QUOTE CADM - -- Emond 125 CADM 17,683 - 29,256 (@ 4,713 hours) - 12,011 - F A T C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 3.5 CADM 310 - 383 (@ 4,765 hours) - 290 - Emond 10.7 CADM 670 - 827 (@ 4,763 hours) - 590 - Emond 35 CADM 1,569 - 1,957 (@ 4,760 hours) - 1,304 - Emond 125 CADM 4,217 - 5,376 (@ 4,757 hours) - 3,303 - B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 3.5 CADM 51.4 - 72.5 (@ 4,710 hours) - 45.3 - Emond 10.7 CADM 130 - 189 (@ 4,710 hours) - 106 - Emond 35 CADM 364 - 546 (@ 4,710 hours) - 274 - Emond 125 CADM 1,164 - 1,793 (@ 4,710 hours) - 824 - B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 3.5 CADM 10.2 - 15.8 (@ 2 hours) - 9.18 - Emond 10.7 CADM 19.8 - 34.4 (@ 1 hours) - 17.0 - Emond 35 CADM 37.0 - 63.2 (@ 1 hours) - 31.4 - Emond 125 CADM 63.1 - 90.9 (@ 1 hours) - 55.2 - 1 2 This document is a draftfor review purposes only and does not constitute Agency policy. C-149 DRAFT--DO NOT CITE OR QUOTE 1 C .3 .1 .2 8 . S h i e t al. (2 0 0 7 ) A d u lt P o r tio n T y p e: S train: Rat Sprague D aw ley D ose: R oute: 1, 5, 5 0 a n d 2 0 0 n g /k g Oral exposure B ody w eight: B W set to 4.5 g R egim e: W e e k ly d o s e s f o r 11 m o n th s Sex: 2 Fem ale S im ulation tim e: 8040 hours W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal 0.143 0.714 Emond CADM Emond CADM 0.342 - 1.07 - 0.475 (@ 7,561 hours) - 1.53 (@ 7,560 hours) - 0.380 - 1.09 - Emond 7.14 CADM 5.23 - 9.12 (@ 7,560 hours) - 4.86 - 28.6 Dose (ng/kg-day) Adjusted dose Emond CADM Model 13.9 29.2 (@ 7,560 hours) L IV E R C O N C EN TR ATIO N S (ng/kg) Metric - Time-weighted Ave Max 12.4 - Terminal 0.143 0.714 Emond CADM Emond CADM 26.1 - 118 - 36.5 (@ 7,564 hours) - 159 (@ 7,564 hours) - 29.6 - 120 - Emond 7.14 CADM 1,068 - 1,415 (@ 7,565 hours) - 970 - Emond 28.6 CADM Dose (ng/kg-day) Adjusted dose Model 4,119 5,450 (@ 7,565 hours) F A T C O N C EN TR ATIO N S (ng/kg) - Metric Time-weighted Ave Max 3,574 - Terminal 0.143 Emond CADM 32.5 - 40.0 (@ 7,583 hours) - 36.7 - 0.714 Emond CADM 102 - 120 (@ 7,584 hours) - 106 - This document is a draftfor review purposes only and does not constitute Agency policy. C-150 DRAFT--DO NOT CITE OR QUOTE Emond 7.14 CADM 497 - 571 (@ 7,584 hours) - Emond 28.6 CADM 1,322 - 1,527 (@ 7,584 hours) - B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max 0.143 Emond CADM 3.94 - 4.99 (@ 7,566 hours) - 0.714 Emond CADM 14.0 - 17.2 (@ 7,566 hours) - Emond 7.14 CADM 90.8 - 112 (@ 7,566 hours) - Emond 28.6 CADM 300 - 374 (@ 7,566 hours) - B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max 0.143 Emond CADM 1.18 - 1.60 (@ 7,563 hours) - 0.714 Emond CADM 3.62 - 4.75 (@ 7,563 hours) - Emond 7.14 CADM 15.6 - 19.7 (@ 7,564 hours) - Emond 28.6 CADM 33.5 - 40.7 (@ 7,564 hours) - 1 2 3 C .3 .1 .2 9 . S m ia lo w ic z e t al. (2 0 0 8 ) 475 - 1,217 - Terminal 4.45 - 14.5 - 84.4 - 266 - Terminal 1.31 - 3.70 - 14.7 - 31.2 - T y p e: S train: M ice B6C3F1 B o d y w e ig h t: 13 w k o ld (BW set to 28g) Sex: 4 Fem ale D ose: R oute: R egim e: 0, 1.5, 15, 150, 450 n g /k g -d ay Oral gavage 5 d a y s/w e e k fo r 13 w e e k s Sim ulation tim e: 2184 This document is a draftfor review purposes only and does not constitute Agency policy. C-151 DRAFT--DO NOT CITE OR QUOTE Dose (ng/kg-day) Adjusted dose 1.07 10.7 107 321 Dose (ng/kg-day) Adjusted dose 1.07 10.7 107 321 Dose (ng/kg-day) Adjusted dose 1.07 10.7 107 321 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Model Time-weighted Ave Metric Max Emond CADM Emond CADM Emond CADM Emond CADM Model 0.438 0.815 (@ 2,112 hours) 2.46 5.12 (@ 2,112 hours) -13.4 36.4 (@ 2,112 hours) -31.6 98.6 (@ 2,112 hours) -L IV E R C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max Emond CADM Emond CADM Emond CADM Emond CADM Model 67.1 107 (@ 2,116 hours) 59.0 92.0 683 971 (@ 2,117 hours) 767 1,000 6,784 9,010 (@ 2,117 hours) 8,349 10,306 20,218 26,379 (@ 2,117 hours) 25,344 31,006 F A T C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max Emond CADM Emond CADM Emond CADM Emond CADM 156 151 885 689 4,831 2,771 11,420 6,337 229 (@ 2,130 hours) 210 1,155 (@ 2,124 hours) 815 5,979 (@ 2,120 hours) 3,224 14,037 (@ 2,119 hours) 7,509 Terminal 0.557 - 2.65 - 12.7 - 28.4 - Terminal 91.5 88.0 787 907 7,043 8,998 20,405 26,967 Terminal 225 204 1,111 774 5,591 2,937 12,920 6 ,6 8 8 This document is a draftfor review purposes only and does not constitute Agency policy. C-l 52 DRAFT--DO NOT CITE OR QUOTE B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 1.07 CADM 17.0 21.0 25.5 (@ 2,116 hours) 29.0 Emond 10.7 CADM 117 119 159 (@ 2,116 hours) 145 Emond 107 CADM 852 727 1,103 (@ 2,116 hours) 875 Emond 321 CADM 2,304 1,961 2,958 (@ 2,116 hours) 2,370 B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 1.07 CADM 1.48 - 2.17 (@ 2,116 hours) - Emond 10.7 CADM 7.60 - 9.86 (@ 2,116 hours) - Emond 107 CADM 30.3 - 36.0 (@ 2,117 hours) - Emond 321 CADM 51.1 - 58.1 (@ 2,117 hours) - 1 2 3 C .3 .1 .3 0 . T o th e t a l., 1 Y e a r (1 9 7 9 ) Terminal 23.9 29.0 141 135 923 778 2,419 2,080 Terminal 1.90 - 8.42 - 31.1 - 51.8 - T y p e: S train: M ice Sw iss/H /R iop D ose: R oute: 7, 700, 7000 ng/kg/w eek Oral gavage In gastric tube B o d y w e ig h t: 10 w eeks old (B W = 27g) R egim e: 1/w eek for 1 y ear (365 days) Sex: Fem ale and m ale S im u latio n tim e: 8,760 hours W e did not sim ulate the scenario using the C A D M m odel because this m odel can only b e run fo r a m axim um o f 123 days. 4 This document is a draftfor review purposes only and does not constitute Agency policy. C-153 DRAFT--DO NOT CITE OR QUOTE Dose (ng/kg-day) Adjusted dose 1 100 1,000 Dose (ng/kg-day) Adjusted dose 1 100 1,000 Dose (ng/kg-day) Adjusted dose 1 100 1,000 Dose (ng/kg-day) Adjusted dose 1 100 Model Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max 0.573 - 14.2 - 91.2 1.61 (@ 8,736 hours) - 116 (@ 8,736 hours) - 1,108 (@ 8,736 hours) -L IV E R C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max 94.2 - 7,343 131 (@ 8,743 hours) - 10,134 (@ 8,745 hours) 70,243 97,658 (@ 8,745 hours) -F A T C O N C EN TR ATIO N S (ng/kg) Metric Time-weighted Ave Max 215 - 5,339 247 (@ 8,613 hours) - 5,914 (@ 8,760 hours) 34,249 38,828 (@ 8,756 hours) -B O D Y B U R D E N (ng/kg) Metric Time-weighted Ave Max 23.4 - 929 - 28.4 (@ 8,742 hours) - 1,189 (@ 8,742 hours) - Terminal 0.682 - 15.7 - 99.3 - Terminal 123 - 9,604 - 92,506 - Terminal 245 - 5,914 - 38,807 - Terminal 27.9 - 1,132 - This document is a draftfor review purposes only and does not constitute Agency policy. C-154 DRAFT--DO NOT CITE OR QUOTE 1,000 Emond CADM 7,569 - 10,045 (@ 8,742 hours) - B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 1 CADM 1.93 - 2.65 (@ 8,741 hours) - Emond 100 CADM 31.8 - 58.4 (@ 2 hours) - 1,000 Emond CADM 78.6 - 103 (@ 2 hours) - 1 2 3 C .3 .1 .3 1 . V a n B ir g e le n e t a l (1 9 9 5 ) 9,471 - Terminal 2.35 - 36.7 - 84.8 - T y p e: Rat S train: B ody w eight: Sex: Sprague D aw ley 150 g Fem ale D ose: R oute: R egim e: S im ulation tim e: 0, 13.5, 26.4, 46.9, 320, 1024 ng/kgday Oral gavage O n c e p e r d a y fo r 13 w e e k s 2184 hours (13 w eeks) Dose (ng/kg-day) Adjusted dose 13.5 26.4 46.9 320 1024 W H O L E B L O O D C O N C E N T R A T IO N S (ng/kg) Model Time-weighted Ave Metric Max Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM 7.20 - 11.8 - 18.1 - 86.4 - 250 - 11.1 (@ 2,160 hours) - 18.6 (@ 2,160 hours) - 29.6 (@ 2,160 hours) - 156 (@ 2,160 hours) - 470 (@ 2,160 hours) - Terminal 8.47 - 13.5 - 20.5 - 95.4 - 275 - This document is a draftfor review purposes only and does not constitute Agency policy. C-155 DRAFT--DO NOT CITE OR QUOTE L IV E R C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 1,655 2,208 (@ 2,164 hours) 2,107 13.5 CADM - - - 26.4 Emond CADM 3,228 - 4,216 (@ 2,164 hours) - 4,017 - 46.9 Emond CADM 5,719 - 7,366 (@ 2,164 hours) - 7,008 - Emond 38,484 47,999 (@ 2,164 hours) 45,537 320 CADM - - - 1024 Emond CADM 121,640 - 150,410 (@ 2,164 hours) - 142,510 - F A T C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 13.5 CADM 669 - 843 (@ 2,167 hours) - 835 - 26.4 Emond CADM 1,092 - 1,357 (@ 2,166 hours) - 1,342 - 46.9 Emond CADM 1,680 - 2,071 (@ 2,166 hours) - 2,045 - Emond 8,027 9,816 (@ 2,165 hours) 9,639 320 CADM - - - 1024 Emond CADM 23,234 - 28,519 (@ 2,165 hours) - 27,954 - B O D Y B U R D E N (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal Emond 13.5 CADM 132 - 173 (@ 2,164 hours) - 167 - 26.4 Emond CADM 240 - 308 (@ 2,164 hours) - 296 - 46.9 Emond CADM 404 - 513 (@ 2,164 hours) - 492 - This document is a draftfor review purposes only and does not constitute Agency policy. C-156 DRAFT--DO NOT CITE OR QUOTE 320 1024 Emond CADM Emond CADM 2,437 - 7,521 - 3,031 (@ 2,164 hours) - 9,310 (@ 2,164 hours) - 2,887 - 8,846 - B O U N D L IV E R (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Emond 13.5 CADM 19.9 - 24.2 (@ 2,164 hours) - 26.4 Emond CADM 29.0 - 34.3 (@ 2,164 hours) - 46.9 Emond CADM 38.8 - 45.0 (@ 2,164 hours) - Emond 79.1 320 CADM - 85.2 (@ 2,164 hours) - 1024 Emond CADM 97.5 - 101 (@ 2,164 hours) - 1 2 3 C .3 .1 .3 2 . V a n d e n H e u v e l e t al. (1 9 9 4 ) Terminal 23.4 - 33.2 - 43.7 - 84.1 - 101 - T y p e: S train: Rat Sprague D aw ley D ose: R oute: 0 .0 5 , 0 .1 , 1, 10, 100, 1 000, 1 0 0 0 0 n g /k g /d Oral gavage Body 10 w eeks old R egim e: Single dose w eight: (BW 225 to 275g, set to 250g) Sex: Fem ale Sim ulation 24 hours * tim e: 4 * 1 w eek is the m inim um that can be sim ulated w ith the CADM model, so the CADM m odel was not used. 5 W H O L E B L O O D C O N C E N T R A T IO N S (ng/kg) Dose (ng/kg-day) Adjusted dose Model Time-weighted Ave Metric Max Terminal 0.05 Emond 0.01 CADM - Emond 0.0113 0.1 CADM - 0.011 (@ 0 hours) - 0.022 (@ 0 hours) - 0.0039 - 0.008 - Emond 0.106 1 CADM - 0.215 (@ 0 hours) - 0.0723 - This document is a draftfor review purposes only and does not constitute Agency policy. C-157 DRAFT--DO NOT CITE OR QUOTE 10 100 1000 10000 Dose (ng/kg-day) Adjusted dose 0.05 0.1 1 10 100 1000 10000 Dose (ng/kg-day) Adjusted dose 0.05 0.1 1 Emond CADM Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM 0.883 2.15 (@ 0 hours) -- 6.45 21.5 (@ 0 hours) -- 48.3 216 (@ 0 hours) -- 435 2,166 (@ 0 hours) -- L IV E R C O N C E N T R A T IO N S (ng/kg) Metric Time-weighted Ave Max 0.232 0.315 (@ 3 hours) -- 0.469 0.631 (@ 3 hours) -- 5.08 6.42 (@ 4 hours) -- 60.2 68.7 (@ 5 hours) -- 730 800 (@ 9 hours) -- 8,186 8,919 (@ 11 hours) -- 84,254 91,675 (@ 11 hours) -- F A T C O N C E N T R A T IO N S (ng/kg) Metric Time-weighted Ave Max 0.138 - 0.274 - 2.58 - 0.215 (@ 24 hours) - 0.427 (@ 24 hours) - 3.97 (@ 24 hours) - 0.583 - 3.85 - 23.9 - 186 - Terminal 0.173 0.0140 0.353 0.0320 4.08 0.950 54.1 52.7 719 1,342 8,442 15,967 88,230 162,773 Terminal 0.215 0.780 0.427 1.57 3.97 15.3 This document is a draftfor review purposes only and does not constitute Agency policy. C-158 DRAFT--DO NOT CITE OR QUOTE 10 100 1000 10000 Dose (ng/kg-day) Adjusted dose 0.05 0.1 1 10 100 1000 10000 Dose (ng/kg-day) Adjusted dose 0.05 0.1 1 Emond CADM Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM 22.1 32.8 (@ 24 hours) -- 170 235 (@ 24 hours) -- 1,348 1,720 (@ 24 hours) -- 12,500 15,265 (@ 24 hours) -- B O D Y B U R D E N (ng/kg) Metric Time-weighted Ave Max 0.0269 0.028 (@ 9 hours) -- 0.0538 0.057 (@ 9 hours) -- 0.536 0.568 (@ 9 hours) -- 5.32 5.65 (@ 8 hours) -- 52.8 56.3 (@ 7 hours) -- 525 562 (@ 7 hours) -- 5,238 5,610 (@ 7 hours) -- B O U N D L IV E R (ng/kg) Metric Time-weighted Ave Max 0.0194 - 0.0383 - 0.353 - 0.027 (@ 3 hours) - 0.054 (@ 3 hours) - 0.506 (@ 3 hours) - 32.8 125 235 739 1,720 5,779 15,265 55,825 Terminal 0.0283 0.0450 0.0565 0.0900 0.562 0.900 5.55 9.00 54.4 90.0 538 900 5,353 9,000 Terminal 0.0142 - 0.0281 - 0.261 - This document is a draftfor review purposes only and does not constitute Agency policy. C-159 DRAFT--DO NOT CITE OR QUOTE Emond 10 CADM Emond 100 CADM 1000 Emond CADM 10000 Emond CADM 1 2 3 C .3 .1 .3 3 . W h ite e t al. (1 9 8 6 ) 2.77 - 16.1 - 57.4 - 100 - 4.24 (@ 2 hours) - 26.4 (@ 2 hours) - 80.2 (@ 1 hours) - 108 (@ 1 hours) - 2.08 - 12.4 - 48.5 - 96.1 - T y p e: S train: M ice B6C3F1 B ody w eight: 7 w eeks old (B W set to 23g) Sex: Fem ale D ose: R oute: R egim e: Sim ulation tim e: 10, 50, 100, 500, 1000, 2000 ng/kg-day Oral gavage 1/day fo r 14 days 336 hours Dose (ng/kg-day) Adjusted dose 10 50 100 500 1,000 2 ,0 0 0 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) Model Time-weighted Ave Metric Max Emond 1.09 2.73 (@ 312 hours) CADM - - Emond 4.08 11.6 (@ 312 hours) CADM - - Emond 7.14 21.7 (@ 312 hours) CADM - - Emond 26.8 96.5 (@ 312 hours) CADM - - Emond 48.7 187 (@ 312 hours) CADM - - Emond 90.6 365 (@ 312 hours) CADM - - Terminal 1.42 - 4.98 - 8.44 - 29.8 - 53.1 - 97.5 - This document is a draftfor review purposes only and does not constitute Agency policy. C-160 DRAFT--DO NOT CITE OR QUOTE Dose (ng/kg-day) Adjusted dose 10 50 100 500 1,000 2 ,0 0 0 Dose (ng/kg-day) Adjusted dose 10 50 100 500 1,000 2 ,0 0 0 Model Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM L IV E R C O N C E N T R A T IO N S (ng/kg) Metric Time-weighted Ave Max 216 375 (@ 317 hours) 217 468 (336h) 1,279 2,164 (@ 317 hours) 1,775 3,261 (336h) 2,707 4,525 (@ 317 hours) 3,999 6,923 (336h) 14,802 24,165 (@ 317 hours) 22,705 36,362 (336h) 30,278 49,034 (@ 317 hours) 46,309 73,145 (336h) 61,381 98,703 (@ 317 hours) 93,577 146,695 (336h) F A T C O N C E N T R A T IO N S (ng/kg) Metric Time-weighted Ave Max 279 507 (@ 336 hours) 316 537 (336h) 1,056 1,846 (@ 336 hours) 1,029 1,564 (336h) 1,854 3,195 (@ 333 hours) 1,662 2,470 (336h) 7,008 11,868 (@ 324 hours) 5,711 8,594 (336h) 12,746 21,566 (@ 323 hours) 10,498 15,993 (336h) 23,691 40,177 (@ 322 hours) 19,990 30,726 (336h) Terminal 343 463 1,997 3,261 4,184 6,923 22,383 36,362 45,414 73,145 91,363 146,695 Terminal 507 537 1,846 1,564 3,195 2,470 11,816 8,594 21,424 15,993 39,843 30,726 This document is a draftfor review purposes only and does not constitute Agency policy. C-161 DRAFT--DO NOT CITE OR QUOTE Dose (ng/kg-day) Adjusted dose 10 50 100 500 1,000 2,000 Dose (ng/kg-day) Adjusted dose 10 50 100 500 1,000 2,000 1 Model Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Model Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM Emond CADM B O D Y B U R D E N (ng/kg) Metric Time-weighted Ave Max 37.7 65.9 (@ 317 hours) 47.9 85.9 (336h) 175 297 (@ 317 hours) 207 342 (336h) 338 570 (@ 316 hours) 388 624 (336h) 1,597 2,637 (@ 316 hours) 1,761 2,754 (336h) 3,137 5,153 (@ 316 hours) 3,455 5,387 (336h) 6,186 10,118 (@ 316 hours) 6,836 10,643 (336h) B O U N D L IV E R (ng/kg) Metric Time-weighted Ave Max 3.49 5.32 (@ 316 hours) -- 11.4 16.4 (@ 317 hours) -- 18.1 25.1 (@ 317 hours) -- 44.2 56.2 (@ 317 hours) -- 59.3 71.9 (@ 317 hours) -- 74.4 86.1 (@ 317 hours) -- Terminal 63.8 85.9 284 342 542 624 2,480 2,754 4,830 5,387 9,459 10,643 Terminal 4.82 - 15.1 - 23.4 - 53.8 - 69.7 - 84.3 - This document is a draftfor review purposes only and does not constitute Agency policy. C-162 DRAFT--DO NOT CITE OR QUOTE 1 C .3 .2 . G e sta tio n a l S tu d ies 2 C .3 .2 .1 . B e ll e t al. (2 0 0 7 ) T y p e: Rat D ose: 2.4, 8 , and 46 ng/kg-day w ith a 0.03 ng/kg-day background S train: H an/W istar R oute: D iet oral dose B ody w eight: 6 w eeks (B W = 85g) R egim e: O nce p e r day fo r 12 w eeks p rio r to m ating, during the tw o w eek m ating period, and during gestation Sex: Fem ale Sim ulation tim e: 2,352 h r (98 days) p rio r to gestation + 504 h r (21 days) during gestation for a total sim ulation o f 2,856 hours 3 * Time averages are com puted during the gestation period only. W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 2.43 2.20 6,295 3.10 (@ 2,352 hours) 8.03 5.14 14,674 7.31 (@ 2,352 hours) 46.03 18.4 52,584 28.1 (@ 2,352 hours) L IV E R C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 2.43 320 914,290 437 (@ 2,356 hours) 1,349 (@ 2,356 8.03 1,040 2,969,800 hours) 46.03 5,892 16,829,000 7,289 (@ 2,356 hours) F A T C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 2.43 205 585,530 263 (@ 2,336 hours) 8.03 478 1,365,100 589 (@ 2,335 hours) 46.03 1,713 4,891,500 2,045 (@ 2,334 hours) Terminal 2 .2 0 5.08 18.1 Terminal 321 1,042 6,007 Terminal 211 486 1,745 This document is a draftfor review purposes only and does not constitute Agency policy. C-163 DRAFT--DO NOT CITE OR QUOTE B O D Y B U R D E N (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 2.43 33.0 94,390 44.4 (@ 2,836 hours) 8.03 90.4 258,110 117 (@ 2,836 hours) 46.03 422 1,206,500 531 (@ 2,836 hours) FE T U S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 2.43 3.03 8,648 39.6 (@ 2,530 hours) 8.03 6.65 18,999 86.7 (@ 2,529 hours) 46.03 20.9 59,794 272 (@ 2,527 hours) B O U N D L IV E R (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 2.43 7.10 20,289 8.98 (@ 2,356 hours) 8.03 15.1 43,242 18.2 (@ 2,356 hours) 46.03 39.6 113,070 44.8 (@ 2,356 hours) 1 2 3 C .3 .2 .2 . H a a v is to e t al. (2 0 0 6 ) Terminal 43.4 114 511 Terminal 6.48 14.4 46.0 Terminal 7.23 15.4 40.6 T y p e: S train: R at Sprague D aw ley D ose: R oute: 20, 400, and 1,000 ng/kg Oral exposure B ody w eight B W = 190 g R egim e: Single dose on G D 13 Sex: Fem ale S im ulation tim e 336 hours 4 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 20 2.8 6 68.9 8.01 (@ 312 hours) Terminal 1.73 400 11.3 273 40.1 (@ 312 hours) 6.28 1000 46.9 1,129 202 (@ 312 hours) 2 2 .8 This document is a draftfor review purposes only and does not constitute Agency policy. C-164 DRAFT--DO NOT CITE OR QUOTE Dose (ng/kg-day) Adjusted dose 20 400 1000 Dose (ng/kg-day) Adjusted dose 20 400 1000 Dose (ng/kg-day) Adjusted dose 20 400 1000 Dose (ng/kg-day) Adjusted dose 20 400 1000 Dose (ng/kg-day) Adjusted dose 20 400 1000 L IV E R C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max 265 6,371 298 (@ 319 hours) 1,497 36,005 1,653 (@ 320 hours) 8,061 193,860 8,832 (@ 321 hours) F A T C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max 56.3 1,354 81.9 (@ 336 hours) 232 1,002 5,584 24,084 321 (@ 336 hours) 1,313 (@ 336 hours) B O D Y B U R D E N (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave 21.1 Area Under the Curve 508 Max 22.5 (@ 319 hours) 105 2,528 112 (@ 319 hours) 524 12,612 561 (@ 319 hours) FE T U S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max 8.47 203 11.3 (@ 336 hours) 31.2 751 40.3 (@ 336 hours) 112 2,689 139 (@ 336 hours) B O U N D L IV E R (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave 8.20 Area Under the Curve 197 Max 13.5 (@ 314 hours) 24.9 598 40.8 (@ 313 hours) 57.1 1,373 80.1 (@ 313 hours) Terminal 244 1,462 8,147 Terminal 81.9 321 1,313 Terminal 21.9 108 538 Terminal 11.3 40.3 139 Terminal 6.03 19.1 47.7 This document is a draftfor review purposes only and does not constitute Agency policy. C-165 DRAFT--DO NOT CITE OR QUOTE 1 C .3 .2 .3 . H o jo e t al. (2 0 0 2 ) T y p e: S train: R at Sprague D aw ley D ose: R oute: 2 0 , 60 and 180 ng/kg Oral exposure B ody w eight 20 ng/kg B W = 271g 60 ng/kg B W = 275g 180 ng/kg B W = 262g R egim e: Single dose on G D 8 Sex: 2 Fem ale S im ulation tim e 216 hours W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max Terminal 20 1.62 39.1 4.47 (@ 192 hours) 1.02 60 4.17 100 13.3 (@ 192 hours) 2.50 180 10.7 258 40.3 (@ 192 hours) 5.96 Dose (ng/kg-day) Adjusted dose L IV E R C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max Terminal 20 128 20,554 144 (@ 198 hours) 43.2 60 420 72,340 465 (@ 200 hours) 147 180 1,364 250,820 1,497 (@ 201 hours) 497 F A T C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max Terminal 20 32.5 17,253 63.0 (@ 281 hours) 49.4 60 86.4 44,093 161 (@ 284 hours) 124 180 226 108,730 398 (@ 286 hours) 301 B O D Y B U R D E N (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max Terminal 20 10.6 3,054 11.3 (@ 200 hours) 8.67 60 31.8 8,702 33.8 (@ 199 hours) 23.6 180 95.0 24,747 101 (@ 199 hours) 63.4 This document is a draftfor review purposes only and does not constitute Agency policy. C-166 DRAFT--DO NOT CITE OR QUOTE FE T U S (ng/kg) a n d A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 20 15.9 2,334 18.4 (@ 206 hours) 60 39.8 5,829 45.7 (@ 205 hours) 180 96.3 13,866 110 (@ 203 hours) B O U N D L IV E R (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 20 4.88 759 7.74 (@ 194 hours) 60 11.2 1,848 18.5 (@ 194 hours) 180 23.6 4,157 38.5 (@ 193 hours) 1 2 3 C .3 .2 .4 . I k e d a e t al. (2 0 0 5 ) Terminal 1.64 4.10 9.72 Terminal 1.75 4.26 9.65 T y p e: Rat D ose: 400 ng/kg single dose and 80 ng/kg w eekly m aintenance dose S train: Sprague D aw ley R oute: Oral gavage B ody w eight: 10 w eeks (B W = 250g) R egim e: 400 ng/kg single dose, tw o w eekly m aintenance doses prior to gestation and w eekly m aintenance doses during gestation Sex: 4 Fem ale S im u la tio n 504 h r (21 days) p rio r to gestation + 504 h r (21 days) tim e: during g estation fo r a to tal sim ulation o f 1,008 hours W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max Terminal 16.5 22.9 23,086 101 (@ 144 hours) 10.1 Dose (ng/kg-day) Adjusted dose L IV E R C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max Terminal 16.5 7,755 7,817,300 17,016 (@ 150 hours) 2,698 This document is a draftfor review purposes only and does not constitute Agency policy. C-167 DRAFT--DO NOT CITE OR QUOTE F A T C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 16.5 2,087 2,103,900 3,663 (@ 184 hours) B O D Y B U R D E N (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 16.5 548 552,590 1,085 (@ 149 hours) FE T U S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 16.5 45.9 46,290 245 (@ 679 hours) B O U N D L IV E R (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 16.5 44.0 44,361 63.8 (@ 149 hours) 1 2 3 C .3 .2 .5 . K a tta in e n e t al. (2 0 0 1 ) T ype: Rat D ose: 30, 100, 300, and 1,000 ng/kg S train: H an/W istar (K uopio) and Long/Evans (Turku/A B ) crossing. R oute: Oral exposure B ody w eight: B W no specify (B W set to 190g)* R egim e: Single dose in the G D 15 Sex: Fem ale Sim ulation tim e: 360 hours 4 *Derelanko and Hollinger (1995). 5 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 30 2.23 100 6.25 53.7 5.95 (@ 336 hours) 150 19.8 (@ 336 hours) Terminal 1,028 Terminal 262 Terminal 30.2 Terminal 26.8 Terminal 1.36 3.62 This document is a draftfor review purposes only and does not constitute Agency policy. C-168 DRAFT--DO NOT CITE OR QUOTE 300 1,000 Dose (ng/kg-day) Adjusted dose 30 100 300 1,000 Dose (ng/kg-day) Adjusted dose 30 100 300 1,000 Dose (ng/kg-day) Adjusted dose 30 100 300 1,000 Dose (ng/kg-day) Adjusted dose 30 100 300 1,000 16.1 387 59.8 (@ 336 hours) 46.9 1,128 200 (@ 336 hours) L IV E R C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max 193 4,648 219 (@ 342 hours) 713 17,141 793 (@ 344 hours) 2,298 55,266 2,533 (@ 345 hours) 8,055 193,720 8,831 (@ 345 hours) F A T C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max 42.8 1,027 62.8 (@ 360 hours) 123 2,964 175 (@ 360 hours) 327 7,853 446 (@ 360 hours) 981 23,588 1,289 (@ 360 hours) B O D Y B U R D E N (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max 15.9 382 16.9 (@ 343 hours) 52.7 1,266 56.2 (@ 343 hours) 158 3,791 168 (@ 343 hours) 524 12,612 561 (@ 343 hours) FE T U S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max 4.86 117 6 .6 6 (@ 360 hours) 13.2 317 17.6 (@ 360 hours) 31.5 758 41.2 (@ 360 hours) 82.2 1,975 104 (@ 360 hours) 8.62 22.7 Terminal 175 680 2,267 8,134 Terminal 62.8 175 446 1,289 Terminal 16.4 54.3 162 538 Terminal 6.66 17.6 41.2 104 This document is a draftfor review purposes only and does not constitute Agency policy. C-169 DRAFT--DO NOT CITE OR QUOTE B O U N D L IV E R (ng/kg) a n d A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 30 6.57 158 10.7 (@ 338 hours) 100 15.8 381 26.3 (@ 338 hours) 300 31.6 760 50.6 (@ 337 hours) 1,000 57.1 1,373 80.1 (@ 337 hours) 1 2 3 C .3 .2 .6 . K e lle r e t al. (2 0 0 7 ) Terminal 4.80 11.9 24.7 47.7 T y p e: S train: M ouse C B A /J and C 3H /H eJ D ose: R oute: 10, 100, and 1000 ng/kg O ral B ody w eight: N ot specified (24 g R egim e: used in the sim ulation) S in g le d o se a t g e s ta tio n d a y 13 Sex: 4 Fem ale Sim ulation tim e: 336 hours W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max Terminal 10 0.537 12.9 1.43 (@ 312 hours) 0.269 100 4.29 103 14.3 (@ 312 hours) 1.95 1,000 34.1 820 143 (@ 312 hours) 12.3 Dose (ng/kg-day) Adjusted dose L IV E R C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max Terminal 10 30.6 737 39.8 (@ 316 hours) 2 2 .2 100 1,000 371 4,214 8,922 101,360 421 (@ 319 hours) 4,697 (@ 321 hours) 317 3,940 Dose (ng/kg-day) Adjusted dose F A T C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max Terminal 10 22.4 538 33.3 (@ 336 hours) 33.3 100 188 4,523 264 (@ 336 hours) 264 1,000 1,591 38,233 2,080 (@ 336 hours) 2,080 This document is a draftfor review purposes only and does not constitute Agency policy. C-170 DRAFT--DO NOT CITE OR QUOTE B O D Y B U R D E N (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose 10 100 1,000 Time-weighted Ave 5.57 54.3 530 Metric Area Under the Curve Max 134 5.99 (@ 319 hours) 1,306 59.0 (@ 318 hours) 12,747 581 (@ 318 hours) FE T U S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose 10 100 1,000 Time-weighted Ave 2.57 21.7 179 Metric Area Under the Curve Max 61.7 3.80 (@ 336 hours) 522 30.0 (@ 334 hours) 4,312 233 (@ 329 hours) B O U N D L IV E R (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose 10 100 1,000 Time-weighted Ave 1.74 11.5 46.7 Metric Area Under the Curve Max 41.8 3.14 (@ 315 hours) 276 23.5 (@ 314 hours) 1,123 79.8 (@ 314 hours) 1 2 3 C .3 .2 .7 . L i e t a l (2 0 0 6 ) 3 -D a y Terminal 5.72 54.7 524 Terminal 3.80 29.9 225 Terminal 1.01 6.99 32.9 T y p e: S train: M ouse N IH D ose: R oute: 2, 50, and 100 ng/kg-day O ral B ody w eig h t: 25-28 g (used 27 g in R egim e: the sim ulation) D aily exposure from gestation day 1 to gestation day 8 Sex: 4 Fem ale Sim ulation tim e: 72 hours W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose 2 Time-weighted Ave 0.159 Metric Area Under the Curve Max 11.4 0.392 (@ 48 hours) Terminal 0.136 50 2.84 100 5.12 205 8.90 (@ 48 hours) 2.38 369 17.3 (@ 48 hours) 4.20 This document is a draftfor review purposes only and does not constitute Agency policy. C-171 DRAFT--DO NOT CITE OR QUOTE L IV E R C O N C EN TR ATIO N S (ng/kg) a n d A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 2 8.98 647 15.1 (@ 52 hours) 50 333 23,971 539 (@ 53 hours) 100 718 51,738 1,156 (@ 53 hours) F A T C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 2 17.0 1,227 31.1 (@ 72 hours) 50 315 22,704 548 (@ 72 hours) 100 576 41,460 984 (@ 72 hours) B O D Y B U R D E N (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 2 2.29 165 3.51 (@ 55 hours) 50 53.6 3,863 82.2 (@ 54 hours) 100 105 7,598 162 (@ 53 hours) FE T U S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 2 0.00 0 0.000 (@ 72 hours) 50 0.0 0 0.000 (@ 72 hours) 100 0.0 0 0.000 (@ 72 hours) B O U N D L IV E R (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 2 0.538 38.8 0.864 (@ 51 hours) 50 8.24 594 13.5 (@ 2 hours) 100 13.6 981 23.7 (@ 2 hours) 1 2 Terminal 9.10 402 888 Terminal 31.1 548 984 Terminal 3.43 77.1 150 Terminal 0.00 0.00 0.00 Terminal 0.498 8.16 13.6 This document is a draftfor review purposes only and does not constitute Agency policy. C-172 DRAFT--DO NOT CITE OR QUOTE 1 C .3 .2 .8 . M a r k o w s k i e t al. (2 0 0 1 ) T y p e: S train: Rat H oltzm an rats D ose: R oute: 2 0 , 60 and 180 ng/kg Oral exposure B ody w eight: B W no specify (B W set to 190g)* R egim e: Single dose in the G D 18 Sex: Fem ale Sim ulation tim e: 432 hours 2 *Derelanko and Hollinger (1995). 3 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 20 1.56 37.5 3.82 (@ 408 hours) 60 4.03 97.0 11.5 (@ 408 hours) 180 10.3 248 34.8 (@ 408 hours) Dose (ng/kg-day) Adjusted dose L IV E R C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max 20 123 2,959 141 (@ 414 hours) 60 409 9,843 459 (@ 415 hours) 180 1,334 32,086 1,479 (@ 416 hours) Dose (ng/kg-day) Adjusted dose F A T C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max 20 27.9 670 41.6 (@ 432 hours) 60 74.0 1,778 107 (@ 432 hours) 180 195 4,685 273 (@ 432 hours) Dose (ng/kg-day) Adjusted dose B O D Y B U R D E N (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max 20 10.6 254 11.2 (@ 415 hours) 60 31.7 762 33.8 (@ 415 hours) 180 94.7 2,278 101 (@ 415 hours) Terminal 0.958 2.38 5.72 Terminal 109 382 1,295 Terminal 41.6 107 273 Terminal 10.9 32.7 97.5 This document is a draftfor review purposes only and does not constitute Agency policy. C-173 DRAFT--DO NOT CITE OR QUOTE FE T U S (ng/kg) a n d A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose 20 Time-weighted Ave 1.26 Metric Area Under the Curve Max 30.2 1.80 (@ 432 hours) 60 3.21 77.2 4.49 (@ 432 hours) 180 7.81 188 10.7 (@ 432 hours) B O U N D L IV E R (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose 20 60 Time-weighted Ave 4.74 11.0 Metric Area Under the Curve Max 114 7.59 (@ 410 hours) 265 18.2 (@ 410 hours) 180 23.2 559 38.1 (@ 409 hours) 1 2 3 C .3 .2 .9 . M ie tin n e n e t al. (2 0 0 6 ) T y p e: S train: Rat Dose: cross-breeding o f H an/W istar and LongE vans rats R oute: 30, 100, 300 and 1000 ng/kg Oral exposure B o d y w e ig h t: B W 11 w e e k s (BW set to 180g) R egim e: Single dose in the G D 15 Sex: 4 Fem ale Sim ulation tim e: 360 hours W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 30 2.22 53.4 5.87 (@ 336 hours) 100 6.23 150 19.6 (@ 336 hours) 300 1,000 16.0 46.6 386 1,123 59.0 (@ 336 hours) 198 (@ 336 hours) Terminal 1.80 4.49 10.7 Terminal 3.43 8.16 17.7 Terminal 1.36 3.61 8.61 22.7 This document is a draftfor review purposes only and does not constitute Agency policy. C-174 DRAFT--DO NOT CITE OR QUOTE Dose (ng/kg-day) Adjusted dose 30 100 300 1,000 Dose (ng/kg-day) Adjusted dose 30 100 300 1,000 Dose (ng/kg-day) Adjusted dose 30 100 300 1,000 Dose (ng/kg-day) Adjusted dose 30 100 300 1,000 Dose (ng/kg-day) Adjusted dose 30 L IV E R C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max 193 4,631 219 (@ 342 hours) 711 17,096 791 (@ 344 hours) 2,294 55,166 2,530 (@ 345 hours) 8,042 193,410 8,820 (@ 345 hours) F A T C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max 43.0 1,034 63.2 (@ 360 hours) 124 2,984 176 (@ 360 hours) 329 7,905 449 (@ 360 hours) 987 23,729 1,296 (@ 360 hours) B O D Y B U R D E N (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max 15.9 381 16.9 (@ 343 hours) 52.6 1,266 56.1 (@ 343 hours) 158 3,791 168 (@ 343 hours) 524 12,609 561 (@ 343 hours) FE T U S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max 4.83 116 6.62 (@ 360 hours) 13.1 315 17.5 (@ 360 hours) 31.3 753 41.0 (@ 360 hours) 81.7 1,963 104 (@ 360 hours) B O U N D L IV E R (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max 6.56 158 10.7 (@ 338 hours) Terminal 174 677 2,260 8,114 Terminal 63.2 176 449 1,296 Terminal 16.4 54.3 162 538 Terminal 6.62 17.5 41.0 104 Terminal 4.78 This document is a draftfor review purposes only and does not constitute Agency policy. C-175 DRAFT--DO NOT CITE OR QUOTE 100 15.8 300 31.6 1,000 57.0 1 2 3 C .3 .2 .1 0 . N o h a r a e t al. (2 0 0 0 ) 381 760 1,372 26.3 (@ 338 hours) 50.5 (@ 337 hours) 80.1 (@ 337 hours) 11.9 24.6 47.6 T y p e: S train: Rat H oltzm an rats D ose: R oute: 12.5, 50, 200 o r 800 ng T C D D /kg Oral exposure B ody w eight: B W no specify (B W set to 190g)* R egim e: Single dose in the G D 15 Sex: Fem ale Sim ulation tim e: 360 hours 4 *Derelanko and Hollinger (1995). 5 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max Terminal 12.5 1.03 24.8 2.44 (@ 336 hours) 0.645 50 3.45 82.9 9.78 (@ 336 hours) 2.07 200 11.3 271 39.2 (@ 336 hours) 6.25 800 38.1 918 158 (@ 336 hours) 18.9 L IV E R C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max Terminal 12.5 73.8 1,776 86.1 (@ 341 hours) 63.6 50 336 8,084 378 (@ 343 hours) 311 200 1,492 35,890 1,651 (@ 344 hours) 1,454 800 6,389 153,640 7,012 (@ 345 hours) 6,423 F A T C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max Terminal 12.5 19.7 473 29.5 (@ 360 hours) 29.5 50 67.6 1,624 97.8 (@ 360 hours) 97.8 200 229 5,504 317 (@ 360 hours) 317 800 803 19,292 1,061 (@ 360 hours) 1,061 This document is a draftfor review purposes only and does not constitute Agency policy. C-176 DRAFT--DO NOT CITE OR QUOTE B O D Y B U R D E N (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 12.5 6.62 159 7.04 (@ 343 hours) 50 26.4 635 28.1 (@ 343 hours) 200 105 2,528 112 (@ 343 hours) 800 420 10,092 449 (@ 343 hours) FE T U S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 12.5 2.25 54.0 3.14 (@ 360 hours) 50 7.43 179 10.1 (@ 360 hours) 200 22.8 548 30.1 (@ 360 hours) 800 68.1 1,638 87.0 (@ 360 hours) B O U N D L IV E R (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 12.5 3.24 77.9 5.12 (@ 338 hours) 50 9.66 232 16.0 (@ 338 hours) 200 24.8 597 40.7 (@ 337 hours) 800 51.9 1,248 75.0 (@ 337 hours) 1 2 3 C .3 .2 .1 1 . O h s a k o e t al. (2 0 0 1 ) Terminal 6.88 27.3 108 430 Terminal 3.14 10.1 30.1 87.0 Terminal 2.32 7.12 19.0 42.7 T ype: R at D ose: 12.5, 50, 200, and 800 ng/kg-day S train: H oltzm ann B ody w eight 10 w eeks (2 0 0 g ) Sex: 4 Fem ale R oute: R egim e: S im ulation tim e Oral exposure on GD15 Single dose 384 hours This document is a draftfor review purposes only and does not constitute Agency policy. C-177 DRAFT--DO NOT CITE OR QUOTE W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 12.5 1.04 25.0 2.48 (@ 360 hours) 50 3.47 83.6 9.93 (@ 360 hours) 2 0 0 11.4 273 39.9 (@ 360 hours) 800 38.4 925 161 (@ 360 hours) L IV E R C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 12.5 74.3 1,788 86.5 (@ 365 hours) 50 338 8,126 379 (@ 367 hours) 200 1,497 36,006 1,655 (@ 368 hours) 800 6,402 153,960 7,025 (@ 369 hours) F A T C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 12.5 19.0 457 28.6 (@ 384 hours) 50 65.3 1,569 94.7 (@ 384 hours) 200 221 5,321 307 (@ 384 hours) 800 777 18,671 1,029 (@ 384 hours) B O D Y B U R D E N (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 12.5 6.63 159 7.05 (@ 367 hours) 50 26.4 635 28.2 (@ 367 hours) 200 105 2,529 112 (@ 367 hours) 800 420 10,093 449 (@ 367 hours) FE T U S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 12.5 1.65 39.5 2.33 (@ 384 hours) Terminal 0.649 2.07 6.26 18.9 Terminal 64.2 314 1,461 6,443 Terminal 28.6 94.7 307 1,029 Terminal 6.89 27.3 108 430 Terminal 2.33 This document is a draftfor review purposes only and does not constitute Agency policy. C-178 DRAFT--DO NOT CITE OR QUOTE 50 5.44 131 7.48 (@ 384 hours) 2 0 0 16.7 401 22.3 (@ 384 hours) 800 49.9 1,200 64.6 (@ 384 hours) B O U N D L IV E R (ng/kg) a n d A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 12.5 3.25 78.3 5.13 (@ 362 hours) 50 9.69 233 16.0 (@ 362 hours) 200 24.9 598 40.7 (@ 361 hours) 800 51.9 1,249 75.0 (@ 361 hours) 1 2 3 C .3 .2 .1 2 . S c h a n tz e t al. (1 9 9 6 ) a n d A m i n e t al. (2 0 0 0 ) 7.48 22.3 64.6 Terminal 2.34 7.16 19.1 42.8 T y p e: S train: Rat Sprague D aw ley D ose: R oute: 25 and 100 ng/kg-day Oral exposure B ody w eight: B W not specified (BW set to 250g) R egim e: D aily doses from G D 10 - 16 Sex: Fem ale S im ulation tim e: 384 hours; tim e averages are calculated from the beginning o f the dosing 4 W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max Terminal 25 3.38 487 8.63 (@ 360 hours) 4.03 100 10.6 1,522 31.1 (@ 360 hours) 12.3 L IV E R C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max Terminal 25 512 73,686 871 (@ 365 hours) 778 100 2,374 341,960 4,012 (@ 366 hours) 3,665 Dose (ng/kg-day) Adjusted dose F A T C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Metric Time-weighted Ave Area Under the Curve Max Terminal 25 169 24,323 306 (@ 384 hours) 306 This document is a draftfor review purposes only and does not constitute Agency policy. C-179 DRAFT--DO NOT CITE OR QUOTE 100 532 76,675 950 (@ 384 hours) B O D Y B U R D E N (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 25 45.1 6,490 76.6 (@ 365 hours) 100 177 25,438 298 (@ 365 hours) FE T U S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 25 25.2 3,627 30.4 (@ 343 hours) 100 74.1 10,672 88.1 (@ 342 hours) B O U N D L IV E R (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 25 9.99 1,439 14.4 (@ 364 hours) 100 25.2 3,632 34.2 (@ 364 hours) 1 2 3 C .3 .2 .1 3 . S e o e t al. (1 9 9 5 ) 950 Terminal 74.3 287 Terminal 27.3 77.9 Terminal 12.8 31.6 T ype: Rat D ose: 25 and 100 ng/kg-day S train: Sprague D aw ley R oute: Oral exposure B ody w eight: B W not specified (BW set to 190g) R egim e: D aily doses from G D 10 - 16 Sex: 4 Fem ale S im ulation tim e: 384 hours; tim e averages are calculated from the beginning o f the dosing W H O LE B LO O D C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max Terminal 25 3.33 479 8.25 (@ 360 hours) 4.00 100 10.4 1,498 29.6 (@ 360 hours) 12.2 This document is a draftfor review purposes only and does not constitute Agency policy. C-180 DRAFT--DO NOT CITE OR QUOTE L IV E R C O N C EN TR ATIO N S (ng/kg) a n d A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 25 504 72,592 861 (@ 365 hours) 100 2,347 337,970 3,978 (@ 365 hours) F A T C O N C EN TR ATIO N S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 25 172 24,807 310 (@ 384 hours) 100 542 78,097 962 (@ 384 hours) B O D Y B U R D E N (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 25 45.0 6,486 76.5 (@ 365 hours) 100 176 25,387 298 (@ 365 hours) FE T U S (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 25 24.7 3,551 29.8 (@ 343 hours) 100 72.6 10,456 86.6 (@ 342 hours) B O U N D L IV E R (ng/kg) and A U C ((ng/kg) h r) Dose (ng/kg-day) Adjusted dose Time-weighted Ave Metric Area Under the Curve Max 25 9.90 1,426 14.3 (@ 364 hours) 100 25.0 3,607 34.1 (@ 364 hours) 1 Terminal 767 3,627 Terminal 310 962 Terminal 74.2 287 Terminal 26.8 76.8 Terminal 12.7 31.4 This document is a draftfor review purposes only and does not constitute Agency policy. C-181 DRAFT--DO NOT CITE OR QUOTE 1 T ab le C -1. M o d el in p u t p a ra m e te rs p o ten tially a d d re sse d by selected 2 articles 3 M odel input param eters potentially addressed Absorption Desorption Distribution Elimination Kinetics Induction CYP1A1 Interspecies differences Age Differences Aryl hydrocarbon receptor (AhR) Mode of action Partition coefficient A rticles Aylward et al., 2004 Aylward et al., 2005a, b Aylward et al., 2009 Bohonowych and Denison, 2007 B overhof et al., 2005 Connor and Aylward, 2006 Heinzl et al., 2007 Irigaray et al., 2005 K erger et al., 2006 K erger et al., 2007 K im et al., 2003 K orenaga et al., 2007 K orkalainen et al., 2004 K ransler et al., 2007 M aruyam a et al., 2002 M aruyam a et al., 2003 M aruyam a and Aoki, 2006 M illbrath et al., 2009 M oser and M cLachlan, 2002 M ullerova and Kopecky, 2007 N adal et al., 2009 N ohara et al., 2006 O lsm an et al., 2007 Saghir et al., 2005 Schecter et al., 2003 Staskal et al., 2005 Toyoshiba et al., 2004 W ilkes et al., 2008 4 Partition coefficient estim ates and CYP param eter value estim ates were derived from W ang et al. (1997, 2000) and 5 Santostefano et al. (1998). This document is a draftfor review purposes only and does not constitute Agency policy. C-182 DRAFT--DO NOT CITE OR QUOTE 1 C .4 . R E S P O N S E S U R F A C E T A B L E S 2 In o rd er to calculate h u m an eq u iv alen t doses, the h u m an m odel m u st b e ru n w ith a daily 3 intake w h ich gives average b lo o d concentrations w h ich m atch th e average concentrations in the 4 ro d en t m odels. H ow ever, such calculation can req u ire num erous hum an m odel runs w ith 5 repeated intake adjustm ents in order to reach th e targ et b lo o d concentrations. T o facilitate this 6 process, a response surface w as created fo r th e hum an m odel. In th e response surface, num erous 7 in tak es w ere ru n and th e b lo o d , fat, and b o d y b u rd en av erag e co n cen tratio n s w ere recorded. 8 T hese tables can then be u sed to estim ate the intake w hich w ould give a targ et blood 9 concentration. T he tw o clo sest in tak es are found and th e in tak e is estim ated by linearly 10 in terp o latin g b etw een th e tw o doses. T hen, th is in tak e is ru n th ro u g h th e h u m an m odel to 11 co n firm th at th e av erag e b lo o d co n cen tratio n is w ith in a sp ecified to leran ce o f th e targ et b lo o d 12 concentration. 13 F o r th e c u rre n t a n a ly sis, th re e d iffe re n t re s p o n s e s u rfa c e s w e re c re a te d : n o n -g e s ta tio n a l 14 lifetim e to b e u sed w ith lo n g -term anim al b io assay s, n o n g estatio n al fiv e y e a r av erag e ru n s to be 15 u se d w ith sh o rte r te rm a n im al b io a ssa y s, a n d g e sta tio n sl to b e u se d w ith g e sta tio n a l an im al 16 b io assay s. A ll th ree resp o n se su fraces are sh o w n in th e fo llo w in g tables. This document is a draftfor review purposes only and does not constitute Agency policy. C-183 DRAFT--DO NOT CITE OR QUOTE D R A FT: D O N O T C IT E O R Q U O T E C -184 This document is a draftfor review purposes only and does not constitute Agency policy. C .4 .1 . N o n g e sta tio n a l L ife tim e Nongestational Lifetime Intake (ng/kg/ day) 1.00E-09 1.33E-09 1.67E-09 2.00E-09 2.33E-09 2.67E-09 3.00E-09 3.33E-09 3.67E-09 4.00E-09 4.33E-09 4.67E-09 5.00E-09 5.33E-09 5.67E-09 6.00E-09 6.33E-09 6.67E-09 7.00E-09 7.33E-09 7.67E-09 8.00E-09 8.33E-09 8.67E-09 9.00E-09 9.33E-09 9.67E-09 1.00E-08 Fat (ng/kg) 2.39E-05 3.18E-05 3.98E-05 4.77E-05 5.57E-05 6.36E-05 7.16E-05 7.95E-05 8.74E-05 9.54E-05 1.03E-04 1.11E-04 1.19E-04 1.27E-04 1.35E-04 1.43E-04 1.51E-04 1.59E-04 1.67E-04 1.75E-04 1.83E-04 1.91E-04 1.99E-04 2.07E-04 2.14E-04 2.22E-04 2.30E-04 2.38E-04 Body Burden (ng/kg) 8.58E-06 1.14E-05 1.43E-05 1.72E-05 2.00E-05 2.29E-05 2.57E-05 2.86E-05 3.14E-05 3.43E-05 3.72E-05 4.00E-05 4.29E-05 4.57E-05 4.86E-05 5.14E-05 5.43E-05 5.71E-05 6.00E-05 6.29E-05 6.57E-05 6.86E-05 7.14E-05 7.43E-05 7.71E-05 8.00E-05 8.28E-05 8.57E-05 Blood (ng/kg) 2.52E-07 3.35E-07 4.19E-07 5.03E-07 5.87E-07 6.70E-07 7.54E-07 8.38E-07 9.22E-07 1.01E-06 1.09E-06 1.17E-06 1.26E-06 1.34E-06 1.42E-06 1.51E-06 1.59E-06 1.68E-06 1.76E-06 1.84E-06 1.93E-06 2.01E-06 2.09E-06 2.18E-06 2.26E-06 2.34E-06 2.43E-06 2.51E-06 Nongestational Lifetime Intake (ng/kg/ day) 1.33E-08 1.67E-08 2.00E-08 2.33E-08 2.67E-08 3.00E-08 3.33E-08 3.67E-08 4.00E-08 4.33E-08 4.67E-08 5.00E-08 5.33E-08 5.67E-08 6.00E-08 6.33E-08 6.67E-08 7.00E-08 7.33E-08 7.67E-08 8.00E-08 8.33E-08 8.67E-08 9.00E-08 9.33E-08 9.67E-08 1.00E-07 1.33E-07 1.67E-07 Fat (ng/kg) 3.17E-04 3.96E-04 4.75E-04 5.54E-04 6.33E-04 7.12E-04 7.91E-04 8.70E-04 9.49E-04 1.03E-03 1.11E-03 1.19E-03 1.26E-03 1.34E-03 1.42E-03 1.50E-03 1.58E-03 1.66E-03 1.73E-03 1.81E-03 1.89E-03 1.97E-03 2.05E-03 2.13E-03 2.21E-03 2.28E-03 2.36E-03 3.14E-03 3.92E-03 Body Burden (ng/kg) 1.14E-04 1.43E-04 1.71E-04 1.99E-04 2.28E-04 2.56E-04 2.85E-04 3.13E-04 3.41E-04 3.70E-04 3.98E-04 4.27E-04 4.55E-04 4.83E-04 5.12E-04 5.40E-04 5.68E-04 5.96E-04 6.25E-04 6.53E-04 6.81E-04 7.10E-04 7.38E-04 7.66E-04 7.94E-04 8.23E-04 8.51E-04 1.13E-03 1.41E-03 Blood (ng/kg) 3.34E-06 4.18E-06 5.01E-06 5.84E-06 6.67E-06 7.50E-06 8.34E-06 9.17E-06 1.00E-05 1.08E-05 1.17E-05 1.25E-05 1.33E-05 1.41E-05 1.50E-05 1.58E-05 1.66E-05 1.75E-05 1.83E-05 1.91E-05 1.99E-05 2.08E-05 2.16E-05 2.24E-05 2.32E-05 2.41E-05 2.49E-05 3.31E-05 4.13E-05 Nongestational Lifetime Intake (ng/kg/ day) 2.00E-07 2.33E-07 2.67E-07 3.00E-07 3.33E-07 3.67E-07 4.00E-07 4.33E-07 4.67E-07 5.00E-07 5.33E-07 5.66E-07 5.99E-07 6.33E-07 6.66E-07 6.99E-07 7.32E-07 7.65E-07 7.98E-07 8.32E-07 8.65E-07 8.98E-07 9.31E-07 9.64E-07 9.97E-07 1.01E-06 1.03E-06 1.04E-06 1.06E-06 Fat (ng/kg) 4.70E-03 5.48E-03 6.26E-03 7.04E-03 7.82E-03 8.60E-03 9.38E-03 1.02E-02 1.09E-02 1.17E-02 1.25E-02 1.32E-02 1.40E-02 1.47E-02 1.55E-02 1.63E-02 1.70E-02 1.78E-02 1.85E-02 1.93E-02 2.00E-02 2.08E-02 2.16E-02 2.23E-02 2.31E-02 2.34E-02 2.37E-02 2.41E-02 2.44E-02 Body Burden (ng/kg) 1.70E-03 1.98E-03 2.26E-03 2.54E-03 2.82E-03 3.10E-03 3.38E-03 3.66E-03 3.95E-03 4.23E-03 4.50E-03 4.78E-03 5.05E-03 5.32E-03 5.60E-03 5.87E-03 6.15E-03 6.42E-03 6.69E-03 6.97E-03 7.24E-03 7.52E-03 7.79E-03 8.07E-03 8.34E-03 8.46E-03 8.59E-03 8.71E-03 8.84E-03 Blood (ng/kg) 4.96E-05 5.78E-05 6.60E-05 7.42E-05 8.24E-05 9.06E-05 9.89E-05 1.07E-04 1.15E-04 1.24E-04 1.31E-04 1.39E-04 1.47E-04 1.55E-04 1.63E-04 1.71E-04 1.79E-04 1.87E-04 1.95E-04 2.03E-04 2.11E-04 2.19E-04 2.27E-04 2.35E-04 2.43E-04 2.47E-04 2.50E-04 2.54E-04 2.58E-04 D R A FT: D O N O T C IT E O R Q U O T E C -185 This document is a draftfor review purposes only and does not constitute Agency policy. Nongestational Lifetime Intake (ng/kg/ day) 1.07E-06 1.09E-06 1.11E-06 1.12E-06 1.14E-06 1.16E-06 1.17E-06 1.19E-06 1.21E-06 1.23E-06 1.24E-06 1.26E-06 1.28E-06 1.30E-06 1.32E-06 1.34E-06 1.36E-06 1.38E-06 1.40E-06 1.42E-06 1.44E-06 1.46E-06 1.49E-06 1.53E-06 1.58E-06 1.62E-06 1.67E-06 1.72E-06 1.77E-06 Fat (ng/kg) 2.48E-02 2.52E-02 2.55E-02 2.59E-02 2.63E-02 2.67E-02 2.70E-02 2.74E-02 2.78E-02 2.82E-02 2.87E-02 2.91E-02 2.95E-02 2.99E-02 3.04E-02 3.08E-02 3.13E-02 3.17E-02 3.22E-02 3.26E-02 3.31E-02 3.36E-02 3.41E-02 3.51E-02 3.61E-02 3.72E-02 3.83E-02 3.94E-02 4.05E-02 Body Burden (ng/kg) 8.97E-03 9.10E-03 9.23E-03 9.37E-03 9.51E-03 9.65E-03 9.79E-03 9.93E-03 1.01E-02 1.02E-02 1.04E-02 1.05E-02 1.07E-02 1.08E-02 1.10E-02 1.12E-02 1.13E-02 1.15E-02 1.16E-02 1.18E-02 1.20E-02 1.22E-02 1.24E-02 1.27E-02 1.31E-02 1.35E-02 1.39E-02 1.43E-02 1.47E-02 Blood (ng/kg) 2.61E-04 2.65E-04 2.69E-04 2.73E-04 2.77E-04 2.81E-04 2.85E-04 2.89E-04 2.93E-04 2.98E-04 3.02E-04 3.06E-04 3.11E-04 3.15E-04 3.20E-04 3.25E-04 3.29E-04 3.34E-04 3.39E-04 3.44E-04 3.49E-04 3.54E-04 3.59E-04 3.70E-04 3.81E-04 3.92E-04 4.03E-04 4.15E-04 4.27E-04 Nongestational Lifetime Intake (ng/kg/ day) 1.83E-06 1.88E-06 1.94E-06 2.00E-06 2.06E-06 2.12E-06 2.18E-06 2.25E-06 2.32E-06 2.39E-06 2.46E-06 2.53E-06 2.61E-06 2.68E-06 2.76E-06 2.85E-06 2.93E-06 3.02E-06 3.11E-06 3.21E-06 3.30E-06 3.40E-06 3.50E-06 3.61E-06 3.72E-06 3.83E-06 3.94E-06 4.06E-06 4.18E-06 Fat (ng/kg) 4.17E-02 4.29E-02 4.41E-02 4.54E-02 4.67E-02 4.81E-02 4.95E-02 5.09E-02 5.24E-02 5.39E-02 5.55E-02 5.71E-02 5.87E-02 6.04E-02 6.22E-02 6.40E-02 6.58E-02 6.77E-02 6.96E-02 7.16E-02 7.37E-02 7.58E-02 7.80E-02 8.02E-02 8.25E-02 8.48E-02 8.73E-02 8.98E-02 9.23E-02 Body Burden (ng/kg) 1.51E-02 1.56E-02 1.60E-02 1.65E-02 1.70E-02 1.75E-02 1.80E-02 1.85E-02 1.90E-02 1.96E-02 2.02E-02 2.07E-02 2.13E-02 2.20E-02 2.26E-02 2.33E-02 2.39E-02 2.46E-02 2.53E-02 2.61E-02 2.68E-02 2.76E-02 2.84E-02 2.92E-02 3.01E-02 3.09E-02 3.18E-02 3.27E-02 3.37E-02 Blood (ng/kg) 4.39E-04 4.52E-04 4.65E-04 4.79E-04 4.93E-04 5.07E-04 5.22E-04 5.37E-04 5.52E-04 5.68E-04 5.85E-04 6.02E-04 6.19E-04 6.37E-04 6.55E-04 6.74E-04 6.93E-04 7.13E-04 7.34E-04 7.55E-04 7.76E-04 7.99E-04 8.22E-04 8.45E-04 8.69E-04 8.94E-04 9.20E-04 9.46E-04 9.73E-04 Nongestational Lifetime Intake (ng/kg/ day) 4.31E-06 4.44E-06 4.57E-06 4.71E-06 4.85E-06 4.99E-06 5.14E-06 5.30E-06 5.46E-06 5.62E-06 5.79E-06 5.96E-06 6.14E-06 6.33E-06 6.52E-06 6.71E-06 6.91E-06 7.12E-06 7.33E-06 7.55E-06 7.78E-06 8.01E-06 8.25E-06 8.50E-06 8.76E-06 9.02E-06 9.29E-06 9.57E-06 9.86E-06 Fat (ng/kg) 9.49E-02 9.76E-02 1.00E-01 1.03E-01 1.06E-01 1.09E-01 1.12E-01 1.15E-01 1.19E-01 1.22E-01 1.25E-01 1.29E-01 1.33E-01 1.36E-01 1.40E-01 1.44E-01 1.48E-01 1.52E-01 1.56E-01 1.61E-01 1.65E-01 1.70E-01 1.74E-01 1.79E-01 1.84E-01 1.89E-01 1.94E-01 2.00E-01 2.05E-01 Body Burden (ng/kg) 3.47E-02 3.57E-02 3.67E-02 3.77E-02 3.88E-02 3.99E-02 4.11E-02 4.22E-02 4.34E-02 4.47E-02 4.59E-02 4.73E-02 4.86E-02 5.00E-02 5.14E-02 5.28E-02 5.43E-02 5.58E-02 5.74E-02 5.90E-02 6.06E-02 6.23E-02 6.41E-02 6.59E-02 6.77E-02 6.96E-02 7.15E-02 7.35E-02 7.56E-02 Blood (ng/kg) 1.00E-03 1.03E-03 1.06E-03 1.09E-03 1.12E-03 1.15E-03 1.18E-03 1.22E-03 1.25E-03 1.29E-03 1.32E-03 1.36E-03 1.40E-03 1.44E-03 1.48E-03 1.52E-03 1.56E-03 1.60E-03 1.65E-03 1.69E-03 1.74E-03 1.79E-03 1.84E-03 1.89E-03 1.94E-03 1.99E-03 2.05E-03 2.10E-03 2.16E-03 D R A FT: D O N O T C IT E O R Q U O T E C -186 This document is a draftfor review purposes only and does not constitute Agency policy. Nongestational Lifetime Intake (ng/kg/ day) 1.02E-05 1.05E-05 1.08E-05 1.11E-05 1.14E-05 1.18E-05 1.21E-05 1.25E-05 1.29E-05 1.32E-05 1.36E-05 1.41E-05 1.45E-05 1.49E-05 1.54E-05 1.58E-05 1.63E-05 1.68E-05 1.73E-05 1.78E-05 1.83E-05 1.89E-05 1.95E-05 2.00E-05 2.06E-05 2.13E-05 2.19E-05 2.25E-05 2.32E-05 Fat (ng/kg) 2.11E-01 2.16E-01 2.22E-01 2.28E-01 2.34E-01 2.41E-01 2.47E-01 2.54E-01 2.61E-01 2.68E-01 2.75E-01 2.83E-01 2.90E-01 2.98E-01 3.06E-01 3.14E-01 3.23E-01 3.31E-01 3.40E-01 3.49E-01 3.58E-01 3.68E-01 3.78E-01 3.88E-01 3.98E-01 4.09E-01 4.20E-01 4.31E-01 4.42E-01 Body Burden (ng/kg) 7.77E-02 7.98E-02 8.20E-02 8.43E-02 8.67E-02 8.91E-02 9.15E-02 9.41E-02 9.67E-02 9.93E-02 1.02E-01 1.05E-01 1.08E-01 1.11E-01 1.14E-01 1.17E-01 1.20E-01 1.23E-01 1.27E-01 1.30E-01 1.34E-01 1.37E-01 1.41E-01 1.45E-01 1.49E-01 1.53E-01 1.57E-01 1.61E-01 1.66E-01 Blood (ng/kg) 2.22E-03 2.28E-03 2.34E-03 2.41E-03 2.47E-03 2.54E-03 2.61E-03 2.68E-03 2.75E-03 2.82E-03 2.90E-03 2.98E-03 3.06E-03 3.14E-03 3.22E-03 3.31E-03 3.40E-03 3.49E-03 3.58E-03 3.68E-03 3.78E-03 3.88E-03 3.98E-03 4.09E-03 4.20E-03 4.31E-03 4.42E-03 4.54E-03 4.66E-03 Nongestational Lifetime Intake (ng/kg/ day) 2.39E-05 2.46E-05 2.54E-05 2.61E-05 2.69E-05 2.77E-05 2.86E-05 2.94E-05 3.03E-05 3.12E-05 3.21E-05 3.31E-05 3.41E-05 3.51E-05 3.62E-05 3.73E-05 3.84E-05 3.95E-05 4.07E-05 4.19E-05 4.32E-05 4.45E-05 4.58E-05 4.72E-05 4.86E-05 5.01E-05 5.16E-05 5.31E-05 5.47E-05 Fat (ng/kg) 4.54E-01 4.66E-01 4.78E-01 4.91E-01 5.04E-01 5.17E-01 5.31E-01 5.45E-01 5.59E-01 5.74E-01 5.89E-01 6.06E-01 6.22E-01 6.38E-01 6.54E-01 6.71E-01 6.89E-01 7.07E-01 7.23E-01 7.41E-01 7.60E-01 7.80E-01 8.00E-01 8.20E-01 8.41E-01 8.63E-01 8.84E-01 9.07E-01 9.30E-01 Body Burden (ng/kg) 1.70E-01 1.75E-01 1.80E-01 1.84E-01 1.89E-01 1.95E-01 2.00E-01 2.05E-01 2.11E-01 2.16E-01 2.22E-01 2.29E-01 2.35E-01 2.41E-01 2.48E-01 2.54E-01 2.61E-01 2.68E-01 2.74E-01 2.82E-01 2.89E-01 2.97E-01 3.05E-01 3.13E-01 3.21E-01 3.29E-01 3.38E-01 3.47E-01 3.56E-01 Blood (ng/kg) 4.78E-03 4.91E-03 5.04E-03 5.17E-03 5.31E-03 5.45E-03 5.59E-03 5.74E-03 5.89E-03 6.05E-03 6.20E-03 6.38E-03 6.54E-03 6.72E-03 6.89E-03 7.08E-03 7.25E-03 7.45E-03 7.62E-03 7.82E-03 8.01E-03 8.22E-03 8.43E-03 8.64E-03 8.86E-03 9.09E-03 9.32E-03 9.55E-03 9.80E-03 Nongestational Lifetime Intake (ng/kg/ day) 5.64E-05 5.81E-05 5.98E-05 6.16E-05 6.34E-05 6.54E-05 6.73E-05 6.93E-05 7.14E-05 7.36E-05 7.58E-05 7.80E-05 8.04E-05 8.28E-05 8.53E-05 8.78E-05 9.05E-05 9.32E-05 9.60E-05 9.89E-05 1.02E-04 1.05E-04 1.08E-04 1.11E-04 1.15E-04 1.18E-04 1.22E-04 1.25E-04 1.29E-04 Fat (ng/kg) 9.53E-01 9.77E-01 1.00E+00 1.03E+00 1.05E+00 1.08E+00 1.11E+00 1.13E+00 1.16E+00 1.19E+00 1.22E+00 1.25E+00 1.28E+00 1.31E+00 1.34E+00 1.37E+00 1.41E+00 1.44E+00 1.48E+00 1.51E+00 1.55E+00 1.59E+00 1.62E+00 1.66E+00 1.70E+00 1.75E+00 1.79E+00 1.83E+00 1.87E+00 Body Burden (ng/kg) 3.65E-01 3.75E-01 3.84E-01 3.95E-01 4.05E-01 4.15E-01 4.26E-01 4.37E-01 4.48E-01 4.58E-01 4.70E-01 4.82E-01 4.94E-01 5.07E-01 5.20E-01 5.33E-01 5.47E-01 5.61E-01 5.75E-01 5.90E-01 6.05E-01 6.20E-01 6.36E-01 6.52E-01 6.69E-01 6.86E-01 7.03E-01 7.20E-01 7.39E-01 Blood (ng/kg) 1.00E-02 1.03E-02 1.06E-02 1.08E-02 1.11E-02 1.14E-02 1.17E-02 1.19E-02 1.22E-02 1.25E-02 1.28E-02 1.31E-02 1.34E-02 1.38E-02 1.41E-02 1.45E-02 1.48E-02 1.52E-02 1.55E-02 1.59E-02 1.63E-02 1.67E-02 1.71E-02 1.75E-02 1.79E-02 1.84E-02 1.88E-02 1.93E-02 1.97E-02 D R A FT: D O N O T C IT E O R Q U O T E C -187 This document is a draftfor review purposes only and does not constitute Agency policy. Nongestational Lifetime Intake (ng/kg/ day) 1.33E-04 1.37E-04 1.41E-04 1.45E-04 1.50E-04 1.54E-04 1.59E-04 1.63E-04 1.68E-04 1.73E-04 1.79E-04 1.84E-04 1.89E-04 1.95E-04 2.01E-04 2.07E-04 2.13E-04 2.20E-04 2.26E-04 2.33E-04 2.40E-04 2.47E-04 2.55E-04 2.62E-04 2.70E-04 2.78E-04 2.86E-04 2.95E-04 3.04E-04 Fat (ng/kg) 1.92E+00 1.97E+00 2.01E+00 2.08E+00 2.11E+00 2.16E+00 2.23E+00 2.29E+00 2.32E+00 2.37E+00 2.43E+00 2.49E+00 2.55E+00 2.61E+00 2.67E+00 2.76E+00 2.80E+00 2.86E+00 2.95E+00 3.02E+00 3.09E+00 3.14E+00 3.21E+00 3.29E+00 3.39E+00 3.47E+00 3.55E+00 3.61E+00 3.72E+00 Body Burden (ng/kg) 7.57E-01 7.76E-01 7.96E-01 8.23E-01 8.36E-01 8.57E-01 8.88E-01 9.10E-01 9.24E-01 9.47E-01 9.71E-01 9.96E-01 1.02E+00 1.05E+00 1.07E+00 1.11E+00 1.13E+00 1.16E+00 1.19E+00 1.22E+00 1.25E+00 1.27E+00 1.31E+00 1.34E+00 1.38E+00 1.42E+00 1.45E+00 1.48E+00 1.53E+00 Blood (ng/kg) 2.02E-02 2.07E-02 2.12E-02 2.19E-02 2.22E-02 2.27E-02 2.35E-02 2.41E-02 2.44E-02 2.50E-02 2.56E-02 2.62E-02 2.68E-02 2.75E-02 2.81E-02 2.91E-02 2.94E-02 3.01E-02 3.11E-02 3.18E-02 3.26E-02 3.30E-02 3.38E-02 3.46E-02 3.57E-02 3.65E-02 3.74E-02 3.80E-02 3.91E-02 Nongestational Lifetime Intake (ng/kg/ day) 3.13E-04 3.22E-04 3.32E-04 3.42E-04 3.52E-04 3.63E-04 3.74E-04 3.85E-04 3.97E-04 4.08E-04 4.21E-04 4.33E-04 4.46E-04 4.60E-04 4.74E-04 4.88E-04 5.02E-04 5.17E-04 5.33E-04 5.49E-04 5.65E-04 5.82E-04 6.00E-04 6.18E-04 6.36E-04 6.55E-04 6.75E-04 6.95E-04 7.16E-04 Fat (ng/kg) 3.80E+00 3.89E+00 3.98E+00 4.07E+00 4.16E+00 4.26E+00 4.35E+00 4.45E+00 4.55E+00 4.66E+00 4.76E+00 4.87E+00 4.98E+00 5.09E+00 5.20E+00 5.32E+00 5.43E+00 5.55E+00 5.68E+00 5.80E+00 5.93E+00 6.06E+00 6.19E+00 6.33E+00 6.46E+00 6.60E+00 6.75E+00 6.89E+00 7.04E+00 Body Burden (ng/kg) 1.56E+00 1.60E+00 1.64E+00 1.68E+00 1.72E+00 1.77E+00 1.81E+00 1.85E+00 1.90E+00 1.94E+00 1.99E+00 2.04E+00 2.09E+00 2.14E+00 2.19E+00 2.24E+00 2.30E+00 2.35E+00 2.41E+00 2.47E+00 2.53E+00 2.59E+00 2.65E+00 2.71E+00 2.78E+00 2.84E+00 2.91E+00 2.98E+00 3.05E+00 Blood (ng/kg) 4.00E-02 4.10E-02 4.19E-02 4.29E-02 4.38E-02 4.48E-02 4.58E-02 4.69E-02 4.80E-02 4.90E-02 5.01E-02 5.13E-02 5.24E-02 5.36E-02 5.48E-02 5.60E-02 5.72E-02 5.85E-02 5.98E-02 6.11E-02 6.24E-02 6.38E-02 6.52E-02 6.66E-02 6.80E-02 6.95E-02 7.10E-02 7.26E-02 7.41E-02 Nongestational Lifetime Intake (ng/kg/ day) 7.38E-04 7.60E-04 7.83E-04 8.06E-04 8.30E-04 8.55E-04 8.81E-04 9.07E-04 9.21E-04 9.35E-04 9.49E-04 9.63E-04 9.69E-04 9.77E-04 1.17E-03 1.18E-03 1.20E-03 1.22E-03 1.24E-03 1.26E-03 1.27E-03 1.29E-03 1.31E-03 1.33E-03 1.35E-03 1.37E-03 1.39E-03 1.41E-03 1.43E-03 Fat (ng/kg) 7.20E+00 7.35E+00 7.51E+00 7.61E+00 7.77E+00 7.94E+00 8.11E+00 8.30E+00 8.37E+00 8.46E+00 9.14E+00 9.54E+00 9.70E+00 9.87E+00 1.01E+01 1.02E+01 1.03E+01 1.04E+01 1.05E+01 1.06E+01 1.07E+01 1.08E+01 1.09E+01 1.10E+01 1.11E+01 1.12E+01 1.13E+01 1.14E+01 1.15E+01 Body Burden (ng/kg) 3.13E+00 3.20E+00 3.28E+00 3.33E+00 3.41E+00 3.49E+00 3.58E+00 3.67E+00 3.70E+00 3.75E+00 4.12E+00 4.33E+00 4.42E+00 4.51E+00 4.58E+00 4.63E+00 4.68E+00 4.73E+00 4.75E+00 4.81E+00 4.86E+00 4.92E+00 4.97E+00 5.03E+00 5.08E+00 5.13E+00 5.18E+00 5.23E+00 5.29E+00 Blood (ng/kg) 7.58E-02 7.74E-02 7.91E-02 8.01E-02 8.19E-02 8.36E-02 8.54E-02 8.74E-02 8.81E-02 8.90E-02 9.62E-02 1.00E-01 1.02E-01 1.04E-01 1.07E-01 1.08E-01 1.09E-01 1.10E-01 1.10E-01 1.11E-01 1.12E-01 1.14E-01 1.15E-01 1.16E-01 1.17E-01 1.18E-01 1.19E-01 1.20E-01 1.21E-01 D R A FT: D O N O T C IT E O R Q U O T E C -188 This document is a draftfor review purposes only and does not constitute Agency policy. Nongestational Lifetime Intake (ng/kg/ day) 1.46E-03 1.48E-03 1.50E-03 1.52E-03 1.54E-03 1.57E-03 1.59E-03 1.61E-03 1.64E-03 1.66E-03 1.69E-03 1.71E-03 1.74E-03 1.76E-03 1.79E-03 1.82E-03 1.84E-03 1.87E-03 1.90E-03 2.02E-03 2.08E-03 2.14E-03 2.20E-03 2.27E-03 2.34E-03 2.41E-03 2.48E-03 2.55E-03 2.63E-03 Fat (ng/kg) 1.16E+01 1.17E+01 1.18E+01 1.20E+01 1.21E+01 1.22E+01 1.24E+01 1.25E+01 1.27E+01 1.28E+01 1.29E+01 1.31E+01 1.32E+01 1.33E+01 1.35E+01 1.36E+01 1.38E+01 1.40E+01 1.46E+01 1.50E+01 1.51E+01 1.53E+01 1.56E+01 1.59E+01 1.62E+01 1.66E+01 1.69E+01 1.72E+01 1.76E+01 Body Burden (ng/kg) 5.34E+00 5.40E+00 5.47E+00 5.54E+00 5.61E+00 5.66E+00 5.73E+00 5.82E+00 5.88E+00 5.95E+00 6.02E+00 6.10E+00 6.17E+00 6.24E+00 6.32E+00 6.39E+00 6.46E+00 6.59E+00 6.95E+00 7.16E+00 7.23E+00 7.31E+00 7.47E+00 7.65E+00 7.82E+00 8.00E+00 8.19E+00 8.38E+00 8.57E+00 Blood (ng/kg) 1.22E-01 1.23E-01 1.25E-01 1.26E-01 1.28E-01 1.29E-01 1.30E-01 1.32E-01 1.33E-01 1.35E-01 1.36E-01 1.37E-01 1.39E-01 1.40E-01 1.42E-01 1.43E-01 1.45E-01 1.47E-01 1.54E-01 1.58E-01 1.59E-01 1.61E-01 1.64E-01 1.68E-01 1.71E-01 1.74E-01 1.78E-01 1.81E-01 1.85E-01 Nongestational Lifetime Intake (ng/kg/ day) 2.71E-03 2.79E-03 2.87E-03 2.96E-03 3.05E-03 3.14E-03 3.23E-03 3.33E-03 3.43E-03 3.53E-03 3.64E-03 3.75E-03 3.98E-03 4.10E-03 4.22E-03 4.35E-03 4.48E-03 4.61E-03 4.75E-03 4.89E-03 5.04E-03 5.19E-03 5.35E-03 5.51E-03 5.67E-03 5.84E-03 5.93E-03 6.02E-03 6.20E-03 Fat (ng/kg) 1.79E+01 1.83E+01 1.87E+01 1.90E+01 1.94E+01 1.98E+01 2.02E+01 2.06E+01 2.10E+01 2.14E+01 2.18E+01 2.25E+01 2.29E+01 2.32E+01 2.35E+01 2.40E+01 2.44E+01 2.49E+01 2.55E+01 2.61E+01 2.69E+01 2.75E+01 2.83E+01 2.91E+01 2.97E+01 3.03E+01 3.04E+01 3.07E+01 3.15E+01 Body Burden (ng/kg) 8.77E+00 8.98E+00 9.19E+00 9.41E+00 9.62E+00 9.85E+00 1.01E+01 1.03E+01 1.06E+01 1.08E+01 1.11E+01 1.15E+01 1.17E+01 1.18E+01 1.20E+01 1.23E+01 1.26E+01 1.29E+01 1.33E+01 1.36E+01 1.41E+01 1.45E+01 1.51E+01 1.55E+01 1.59E+01 1.63E+01 1.64E+01 1.65E+01 1.71E+01 Blood (ng/kg) 1.89E-01 1.92E-01 1.96E-01 2.00E-01 2.04E-01 2.08E-01 2.13E-01 2.17E-01 2.21E-01 2.25E-01 2.30E-01 2.37E-01 2.41E-01 2.44E-01 2.48E-01 2.52E-01 2.57E-01 2.63E-01 2.69E-01 2.74E-01 2.83E-01 2.90E-01 2.98E-01 3.06E-01 3.13E-01 3.19E-01 3.20E-01 3.23E-01 3.31E-01 Nongestational Lifetime Intake (ng/kg/ day) 6.38E-03 6.57E-03 6.77E-03 6.98E-03 7.18E-03 7.40E-03 7.51E-03 7.62E-03 7.85E-03 8.09E-03 8.33E-03 8.58E-03 8.71E-03 8.84E-03 9.10E-03 9.37E-03 9.66E-03 9.94E-03 1.02E-02 1.06E-02 1.09E-02 1.12E-02 1.15E-02 1.19E-02 1.22E-02 1.26E-02 1.30E-02 1.34E-02 1.38E-02 Fat (ng/kg) 3.22E+01 3.28E+01 3.35E+01 3.42E+01 3.50E+01 3.57E+01 3.61E+01 3.63E+01 3.67E+01 3.70E+01 3.75E+01 3.89E+01 3.93E+01 3.97E+01 4.04E+01 4.13E+01 4.21E+01 4.31E+01 4.39E+01 4.47E+01 4.56E+01 4.66E+01 4.75E+01 4.82E+01 4.91E+01 5.00E+01 5.12E+01 5.24E+01 5.36E+01 Body Burden (ng/kg) 1.76E+01 1.80E+01 1.84E+01 1.89E+01 1.94E+01 1.99E+01 2.02E+01 2.03E+01 2.06E+01 2.07E+01 2.10E+01 2.21E+01 2.24E+01 2.26E+01 2.31E+01 2.38E+01 2.43E+01 2.50E+01 2.56E+01 2.62E+01 2.68E+01 2.75E+01 2.82E+01 2.87E+01 2.94E+01 3.00E+01 3.09E+01 3.19E+01 3.28E+01 Blood (ng/kg) 3.39E-01 3.46E-01 3.53E-01 3.60E-01 3.68E-01 3.76E-01 3.80E-01 3.82E-01 3.87E-01 3.89E-01 3.94E-01 4.09E-01 4.14E-01 4.18E-01 4.25E-01 4.35E-01 4.44E-01 4.53E-01 4.62E-01 4.71E-01 4.80E-01 4.90E-01 5.00E-01 5.07E-01 5.17E-01 5.26E-01 5.39E-01 5.52E-01 5.65E-01 D R A FT: D O N O T C IT E O R Q U O T E C -189 This document is a draftfor review purposes only and does not constitute Agency policy. Nongestational Lifetime Intake (ng/kg/ day) 1.42E-02 1.46E-02 1.50E-02 1.55E-02 1.60E-02 1.64E-02 1.69E-02 1.74E-02 1.80E-02 1.85E-02 1.91E-02 1.96E-02 2.02E-02 2.08E-02 2.14E-02 2.21E-02 2.28E-02 2.34E-02 2.41E-02 2.49E-02 2.56E-02 2.64E-02 2.72E-02 2.80E-02 2.88E-02 2.97E-02 3.06E-02 3.15E-02 3.24E-02 Fat (ng/kg) 5.48E+01 5.57E+01 5.68E+01 5.78E+01 5.88E+01 5.97E+01 6.10E+01 6.22E+01 6.34E+01 6.47E+01 6.60E+01 6.73E+01 6.86E+01 7.00E+01 7.13E+01 7.28E+01 7.42E+01 7.57E+01 7.71E+01 7.87E+01 8.02E+01 8.18E+01 8.33E+01 8.50E+01 8.67E+01 8.83E+01 9.03E+01 9.21E+01 9.40E+01 Body Burden (ng/kg) 3.37E+01 3.44E+01 3.52E+01 3.60E+01 3.67E+01 3.75E+01 3.85E+01 3.95E+01 4.04E+01 4.14E+01 4.25E+01 4.35E+01 4.46E+01 4.57E+01 4.69E+01 4.81E+01 4.93E+01 5.05E+01 5.18E+01 5.31E+01 5.44E+01 5.58E+01 5.71E+01 5.86E+01 6.01E+01 6.16E+01 6.34E+01 6.50E+01 6.67E+01 Blood (ng/kg) 5.77E-01 5.87E-01 5.97E-01 6.08E-01 6.19E-01 6.29E-01 6.42E-01 6.55E-01 6.68E-01 6.81E-01 6.94E-01 7.08E-01 7.22E-01 7.36E-01 7.51E-01 7.66E-01 7.81E-01 7.97E-01 8.12E-01 8.28E-01 8.44E-01 8.61E-01 8.77E-01 8.95E-01 9.12E-01 9.30E-01 9.50E-01 9.69E-01 9.89E-01 Nongestational Lifetime Intake (ng/kg/ day) 3.34E-02 3.44E-02 3.54E-02 3.65E-02 3.76E-02 3.87E-02 3.99E-02 4.11E-02 4.23E-02 4.36E-02 4.49E-02 4.63E-02 4.76E-02 4.91E-02 5.05E-02 5.21E-02 5.36E-02 5.52E-02 5.69E-02 5.86E-02 6.03E-02 6.22E-02 6.40E-02 6.59E-02 6.79E-02 7.00E-02 7.21E-02 7.42E-02 7.64E-02 Fat (ng/kg) 9.57E+01 9.74E+01 9.92E+01 1.01E+02 1.03E+02 1.05E+02 1.07E+02 1.09E+02 1.11E+02 1.14E+02 1.16E+02 1.18E+02 1.21E+02 1.23E+02 1.25E+02 1.28E+02 1.30E+02 1.33E+02 1.35E+02 1.38E+02 1.41E+02 1.43E+02 1.46E+02 1.49E+02 1.52E+02 1.55E+02 1.58E+02 1.61E+02 1.64E+02 Body Burden (ng/kg) 6.83E+01 6.99E+01 7.15E+01 7.32E+01 7.51E+01 7.69E+01 7.89E+01 8.09E+01 8.30E+01 8.53E+01 8.76E+01 8.99E+01 9.22E+01 9.46E+01 9.70E+01 9.95E+01 1.02E+02 1.05E+02 1.07E+02 1.10E+02 1.13E+02 1.16E+02 1.19E+02 1.22E+02 1.25E+02 1.28E+02 1.31E+02 1.35E+02 1.38E+02 Blood (ng/kg) 1.01E+00 1.03E+00 1.04E+00 1.06E+00 1.08E+00 1.10E+00 1.13E+00 1.15E+00 1.17E+00 1.20E+00 1.22E+00 1.24E+00 1.27E+00 1.29E+00 1.32E+00 1.34E+00 1.37E+00 1.40E+00 1.43E+00 1.45E+00 1.48E+00 1.51E+00 1.54E+00 1.57E+00 1.60E+00 1.63E+00 1.66E+00 1.69E+00 1.73E+00 Nongestational Lifetime Intake (ng/kg/ day) 7.87E-02 8.11E-02 8.35E-02 8.60E-02 8.86E-02 9.13E-02 9.40E-02 9.68E-02 9.97E-02 1.03E-01 1.06E-01 1.09E-01 1.12E-01 1.16E-01 1.19E-01 1.23E-01 1.26E-01 1.30E-01 1.34E-01 1.38E-01 1.42E-01 1.46E-01 1.51E-01 1.55E-01 1.60E-01 1.65E-01 1.70E-01 1.75E-01 1.80E-01 Fat (ng/kg) 1.67E+02 1.71E+02 1.74E+02 1.78E+02 1.81E+02 1.85E+02 1.88E+02 1.92E+02 1.96E+02 1.99E+02 2.03E+02 2.07E+02 2.11E+02 2.15E+02 2.20E+02 2.24E+02 2.28E+02 2.33E+02 2.38E+02 2.42E+02 2.47E+02 2.52E+02 2.57E+02 2.62E+02 2.67E+02 2.72E+02 2.78E+02 2.83E+02 2.89E+02 Body Burden (ng/kg) 1.42E+02 1.46E+02 1.50E+02 1.54E+02 1.58E+02 1.62E+02 1.66E+02 1.70E+02 1.75E+02 1.79E+02 1.84E+02 1.89E+02 1.94E+02 1.99E+02 2.04E+02 2.10E+02 2.15E+02 2.21E+02 2.27E+02 2.33E+02 2.39E+02 2.46E+02 2.52E+02 2.59E+02 2.66E+02 2.73E+02 2.80E+02 2.88E+02 2.95E+02 Blood (ng/kg) 1.76E+00 1.80E+00 1.83E+00 1.87E+00 1.90E+00 1.94E+00 1.98E+00 2.02E+00 2.06E+00 2.10E+00 2.14E+00 2.18E+00 2.22E+00 2.27E+00 2.31E+00 2.36E+00 2.40E+00 2.45E+00 2.50E+00 2.55E+00 2.60E+00 2.65E+00 2.70E+00 2.75E+00 2.81E+00 2.86E+00 2.92E+00 2.98E+00 3.04E+00 D R A FT: D O N O T C IT E O R Q U O T E C -190 This document is a draftfor review purposes only and does not constitute Agency policy. Nongestational Lifetime Intake (ng/kg/ day) 1.86E-01 1.91E-01 1.97E-01 2.03E-01 2.09E-01 2.15E-01 2.22E-01 2.28E-01 2.35E-01 2.42E-01 2.49E-01 2.57E-01 2.65E-01 2.72E-01 2.81E-01 2.89E-01 2.98E-01 3.07E-01 3.16E-01 3.25E-01 3.35E-01 3.45E-01 3.56E-01 3.66E-01 3.77E-01 3.89E-01 4.00E-01 4.12E-01 4.25E-01 Fat (ng/kg) 2.94E+02 3.00E+02 3.06E+02 3.12E+02 3.18E+02 3.25E+02 3.31E+02 3.38E+02 3.44E+02 3.51E+02 3.58E+02 3.65E+02 3.73E+02 3.80E+02 3.88E+02 3.95E+02 4.03E+02 4.11E+02 4.19E+02 4.28E+02 4.36E+02 4.45E+02 4.54E+02 4.63E+02 4.72E+02 4.82E+02 4.91E+02 5.01E+02 5.11E+02 Body Burden (ng/kg) 3.03E+02 3.12E+02 3.20E+02 3.28E+02 3.37E+02 3.46E+02 3.56E+02 3.65E+02 3.75E+02 3.86E+02 3.96E+02 4.07E+02 4.18E+02 4.29E+02 4.41E+02 4.53E+02 4.65E+02 4.77E+02 4.90E+02 5.04E+02 5.18E+02 5.32E+02 5.47E+02 5.62E+02 5.77E+02 5.93E+02 6.09E+02 6.26E+02 6.43E+02 Blood (ng/kg) 3.10E+00 3.16E+00 3.22E+00 3.28E+00 3.35E+00 3.42E+00 3.48E+00 3.55E+00 3.62E+00 3.70E+00 3.77E+00 3.85E+00 3.92E+00 4.00E+00 4.08E+00 4.16E+00 4.24E+00 4.33E+00 4.41E+00 4.50E+00 4.59E+00 4.68E+00 4.78E+00 4.87E+00 4.97E+00 5.07E+00 5.17E+00 5.28E+00 5.38E+00 Nongestational Lifetime Intake (ng/kg/ day) 4.37E-01 4.50E-01 4.64E-01 4.92E-01 5.07E-01 5.22E-01 5.54E-01 5.71E-01 5.88E-01 6.05E-01 6.23E-01 6.61E-01 6.81E-01 7.02E-01 7.23E-01 7.44E-01 7.67E-01 7.90E-01 8.13E-01 8.38E-01 8.63E-01 8.89E-01 9.16E-01 9.43E-01 9.71E-01 1.00E+00 1.06E+00 1.09E+00 1.13E+00 Fat (ng/kg) 5.22E+02 5.32E+02 5.43E+02 5.65E+02 5.76E+02 5.88E+02 6.12E+02 6.25E+02 6.37E+02 6.50E+02 6.64E+02 6.91E+02 7.05E+02 7.20E+02 7.34E+02 7.49E+02 7.65E+02 7.80E+02 7.97E+02 8.13E+02 8.30E+02 8.47E+02 8.65E+02 8.83E+02 9.01E+02 9.20E+02 9.58E+02 9.78E+02 9.99E+02 Body Burden (ng/kg) 6.61E+02 6.79E+02 6.98E+02 7.37E+02 7.57E+02 7.78E+02 8.22E+02 8.44E+02 8.68E+02 8.92E+02 9.17E+02 9.68E+02 9.95E+02 1.02E+03 1.05E+03 1.08E+03 1.11E+03 1.14E+03 1.17E+03 1.21E+03 1.24E+03 1.28E+03 1.31E+03 1.35E+03 1.39E+03 1.43E+03 1.51E+03 1.55E+03 1.59E+03 Blood (ng/kg) 5.49E+00 5.60E+00 5.71E+00 5.95E+00 6.07E+00 6.19E+00 6.44E+00 6.58E+00 6.71E+00 6.84E+00 6.98E+00 7.27E+00 7.42E+00 7.57E+00 7.73E+00 7.89E+00 8.05E+00 8.21E+00 8.38E+00 8.56E+00 8.73E+00 8.91E+00 9.10E+00 9.29E+00 9.48E+00 9.68E+00 1.01E+01 1.03E+01 1.05E+01 Nongestational Lifetime Intake (ng/kg/ day) 1.16E+00 1.19E+00 1.23E+00 1.27E+00 1.31E+00 1.34E+00 1.38E+00 1.43E+00 1.47E+00 1.51E+00 1.56E+00 1.61E+00 1.65E+00 1.70E+00 1.75E+00 1.81E+00 1.86E+00 1.92E+00 1.97E+00 2.03E+00 2.09E+00 2.16E+00 2.22E+00 2.29E+00 2.36E+00 2.43E+00 2.50E+00 2.58E+00 2.65E+00 Fat (ng/kg) 1.02E+03 1.04E+03 1.06E+03 1.09E+03 1.11E+03 1.13E+03 1.16E+03 1.18E+03 1.21E+03 1.23E+03 1.26E+03 1.28E+03 1.31E+03 1.34E+03 1.37E+03 1.40E+03 1.43E+03 1.46E+03 1.49E+03 1.52E+03 1.56E+03 1.59E+03 1.62E+03 1.66E+03 1.70E+03 1.73E+03 1.77E+03 1.81E+03 1.85E+03 Body Burden (ng/kg) 1.64E+03 1.68E+03 1.73E+03 1.78E+03 1.83E+03 1.88E+03 1.94E+03 1.99E+03 2.05E+03 2.11E+03 2.17E+03 2.23E+03 2.29E+03 2.36E+03 2.42E+03 2.49E+03 2.56E+03 2.64E+03 2.71E+03 2.79E+03 2.87E+03 2.95E+03 3.03E+03 3.12E+03 3.21E+03 3.30E+03 3.40E+03 3.49E+03 3.59E+03 Blood (ng/kg) 1.07E+01 1.10E+01 1.12E+01 1.14E+01 1.17E+01 1.19E+01 1.22E+01 1.24E+01 1.27E+01 1.30E+01 1.32E+01 1.35E+01 1.38E+01 1.41E+01 1.44E+01 1.47E+01 1.50E+01 1.54E+01 1.57E+01 1.60E+01 1.64E+01 1.67E+01 1.71E+01 1.75E+01 1.79E+01 1.82E+01 1.86E+01 1.91E+01 1.95E+01 D R A FT: D O N O T C IT E O R Q U O T E C -191 This document is a draftfor review purposes only and does not constitute Agency policy. Nongestational Lifetime Intake (ng/kg/ day) 2.73E+00 2.82E+00 2.90E+00 2.99E+00 3.08E+00 3.17E+00 3.26E+00 3.36E+00 3.46E+00 3.57E+00 3.67E+00 3.78E+00 3.90E+00 4.01E+00 4.13E+00 4.26E+00 4.39E+00 4.52E+00 4.65E+00 4.79E+00 4.94E+00 5.08E+00 5.24E+00 5.39E+00 5.56E+00 5.72E+00 5.89E+00 6.07E+00 6.25E+00 Fat (ng/kg) 1.89E+03 1.93E+03 1.98E+03 2.02E+03 2.07E+03 2.11E+03 2.16E+03 2.21E+03 2.26E+03 2.31E+03 2.36E+03 2.42E+03 2.47E+03 2.53E+03 2.58E+03 2.64E+03 2.70E+03 2.77E+03 2.83E+03 2.90E+03 2.96E+03 3.03E+03 3.10E+03 3.18E+03 3.25E+03 3.33E+03 3.41E+03 3.49E+03 3.57E+03 Body Burden (ng/kg) 3.70E+03 3.80E+03 3.91E+03 4.03E+03 4.14E+03 4.26E+03 4.38E+03 4.51E+03 4.64E+03 4.77E+03 4.91E+03 5.05E+03 5.20E+03 5.35E+03 5.50E+03 5.66E+03 5.83E+03 6.00E+03 6.17E+03 6.35E+03 6.53E+03 6.72E+03 6.92E+03 7.12E+03 7.33E+03 7.54E+03 7.76E+03 7.98E+03 8.22E+03 Blood (ng/kg) 1.99E+01 2.04E+01 2.08E+01 2.13E+01 2.17E+01 2.22E+01 2.27E+01 2.32E+01 2.38E+01 2.43E+01 2.49E+01 2.54E+01 2.60E+01 2.66E+01 2.72E+01 2.78E+01 2.85E+01 2.91E+01 2.98E+01 3.05E+01 3.12E+01 3.19E+01 3.27E+01 3.34E+01 3.42E+01 3.50E+01 3.58E+01 3.67E+01 3.76E+01 Nongestational Lifetime Intake (ng/kg/ day) 6.44E+00 6.63E+00 6.83E+00 7.04E+00 7.25E+00 7.47E+00 7.69E+00 7.92E+00 8.16E+00 8.40E+00 8.66E+00 8.92E+00 9.18E+00 9.46E+00 9.74E+00 1.00E+01 1.00E+01 1.34E+01 1.67E+01 2.00E+01 2.33E+01 2.67E+01 3.00E+01 3.33E+01 3.67E+01 4.00E+01 4.33E+01 4.67E+01 5.00E+01 Fat (ng/kg) 3.65E+03 3.74E+03 3.83E+03 3.92E+03 4.02E+03 4.11E+03 4.21E+03 4.32E+03 4.42E+03 4.53E+03 4.64E+03 4.75E+03 4.87E+03 4.99E+03 5.11E+03 5.22E+03 5.24E+03 6.64E+03 8.04E+03 9.45E+03 1.08E+04 1.22E+04 1.36E+04 1.49E+04 1.63E+04 1.77E+04 1.90E+04 2.04E+04 2.17E+04 Body Burden (ng/kg) 8.45E+03 8.70E+03 8.95E+03 9.21E+03 9.48E+03 9.76E+03 1.00E+04 1.03E+04 1.06E+04 1.10E+04 1.13E+04 1.16E+04 1.19E+04 1.23E+04 1.26E+04 1.30E+04 1.30E+04 1.72E+04 2.14E+04 2.56E+04 2.97E+04 3.39E+04 3.81E+04 4.22E+04 4.63E+04 5.05E+04 5.46E+04 5.87E+04 6.28E+04 Blood (ng/kg) 3.85E+01 3.94E+01 4.03E+01 4.13E+01 4.23E+01 4.33E+01 4.43E+01 4.54E+01 4.65E+01 4.77E+01 4.88E+01 5.00E+01 5.13E+01 5.25E+01 5.38E+01 5.50E+01 5.51E+01 6.99E+01 8.47E+01 9.94E+01 1.14E+02 1.28E+02 1.43E+02 1.57E+02 1.72E+02 1.86E+02 2.00E+02 2.15E+02 2.29E+02 Nongestational Lifetime Intake (ng/kg/ day) Fat (ng/kg) Body Burden (ng/kg) Blood (ng/kg) 5.33E+01 2.31E+04 6.69E+04 2.43E+02 5.67E+01 2.45E+04 7.10E+04 2.57E+02 6.00E+01 2.58E+04 7.51E+04 2.72E+02 6.33E+01 2.72E+04 7.92E+04 2.86E+02 6.67E+01 2.85E+04 8.32E+04 3.00E+02 7.00E+01 2.99E+04 8.73E+04 3.14E+02 7.33E+01 3.12E+04 9.13E+04 3.29E+02 7.67E+01 3.26E+04 9.54E+04 3.43E+02 8.00E+01 3.39E+04 9.94E+04 3.57E+02 8.33E+01 3.53E+04 1.03E+05 3.71E+02 8.67E+01 3.66E+04 1.07E+05 3.86E+02 9.00E+01 3.80E+04 1.12E+05 4.00E+02 9.33E+01 3.94E+04 1.16E+05 4.14E+02 9.67E+01 4.07E+04 1.20E+05 4.28E+02 1.00E+02 4.21E+04 1.24E+05 4.43E+02 1 D R A FT: D O N O T C IT E O R Q U O T E C -192 This document is a draftfor review purposes only and does not constitute Agency policy. C .4 .2 . N o n g e sta tio n a l 5 -Y e a r A verage Non-gestational 5-year Average Intake (ng/kg/ day) 1.00E-09 Fat (ng/kg) 5.18E-05 Body Burden (ng/kg) 1.87E-05 1.33E-09 6.90E-05 2.50E-05 1.67E-09 8.62E-05 3.12E-05 2.00E-09 2.33E-09 2.67E-09 3.00E-09 3.33E-09 1.03E-04 1.21E-04 1.38E-04 1.55E-04 1.72E-04 3.74E-05 4.36E-05 4.99E-05 5.61E-05 6.23E-05 3.67E-09 1.90E-04 6.86E-05 4.00E-09 2.07E-04 7.48E-05 4.33E-09 2.24E-04 8.10E-05 4.67E-09 2.41E-04 8.72E-05 5.00E-09 2.58E-04 9.35E-05 5.33E-09 5.67E-09 6.00E-09 6.33E-09 6.67E-09 2.76E-04 2.93E-04 3.10E-04 3.27E-04 3.44E-04 9.97E-05 1.06E-04 1.12E-04 1.18E-04 1.25E-04 7.00E-09 3.61E-04 1.31E-04 7.33E-09 3.79E-04 1.37E-04 7.67E-09 3.96E-04 1.43E-04 8.00E-09 4.13E-04 1.49E-04 8.33E-09 8.67E-09 9.00E-09 9.33E-09 9.67E-09 4.30E-04 4.47E-04 4.65E-04 4.82E-04 4.99E-04 1.56E-04 1.62E-04 1.68E-04 1.74E-04 1.80E-04 1.00E-08 5.16E-04 1.87E-04 1.33E-08 6.87E-04 2.48E-04 1.67E-08 8.57E-04 3.10E-04 2.00E-08 1.03E-03 3.72E-04 2.33E-08 1.20E-03 4.34E-04 2.67E-08 3.00E-08 3.33E-08 1.37E-03 1.54E-03 1.71E-03 4.96E-04 5.57E-04 6.19E-04 Blood (ng/kg) 5.45E-07 7.26E-07 9.07E-07 1.09E-06 1.27E-06 1.45E-06 1.63E-06 1.81E-06 1.99E-06 2.17E-06 2.36E-06 2.54E-06 2.72E-06 2.90E-06 3.08E-06 3.26E-06 3.44E-06 3.62E-06 3.80E-06 3.98E-06 4.16E-06 4.34E-06 4.52E-06 4.70E-06 4.89E-06 5.07E-06 5.25E-06 5.43E-06 7.22E-06 9.01E-06 1.08E-05 1.26E-05 1.44E-05 1.62E-05 1.80E-05 Non-gestational 5-year Average Intake (ng/kg/ day) 3.67E-08 4.00E-08 4.33E-08 4.67E-08 5.00E-08 5.33E-08 5.67E-08 6.00E-08 6.33E-08 6.67E-08 7.00E-08 7.33E-08 7.67E-08 8.00E-08 8.33E-08 8.67E-08 9.00E-08 9.33E-08 9.67E-08 1.00E-07 1.33E-07 1.67E-07 2.00E-07 2.33E-07 2.67E-07 3.00E-07 3.33E-07 3.67E-07 4.00E-07 4.33E-07 4.67E-07 5.00E-07 5.33E-07 5.66E-07 5.99E-07 6.33E-07 6.66E-07 6.99E-07 Fat (ng/kg) 1.88E-03 2.05E-03 2.22E-03 2.39E-03 2.56E-03 2.73E-03 2.90E-03 3.07E-03 3.24E-03 3.41E-03 3.57E-03 3.74E-03 3.91E-03 4.08E-03 4.25E-03 4.42E-03 4.59E-03 4.76E-03 4.93E-03 5.09E-03 6.74E-03 8.39E-03 1.00E-02 1.17E-02 1.33E-02 1.50E-02 1.66E-02 1.83E-02 1.99E-02 2.16E-02 2.32E-02 2.49E-02 2.64E-02 2.80E-02 2.96E-02 3.11E-02 3.27E-02 3.43E-02 Body Burden (ng/kg) 6.81E-04 7.43E-04 8.04E-04 8.66E-04 9.28E-04 9.89E-04 1.05E-03 1.11E-03 1.17E-03 1.23E-03 1.30E-03 1.36E-03 1.42E-03 1.48E-03 1.54E-03 1.60E-03 1.66E-03 1.72E-03 1.79E-03 1.85E-03 2.45E-03 3.05E-03 3.65E-03 4.25E-03 4.85E-03 5.45E-03 6.05E-03 6.65E-03 7.25E-03 7.85E-03 8.45E-03 9.05E-03 9.63E-03 1.02E-02 1.08E-02 1.14E-02 1.19E-02 1.25E-02 Blood (ng/kg) 1.98E-05 2.16E-05 2.34E-05 2.51E-05 2.69E-05 2.87E-05 3.05E-05 3.23E-05 3.40E-05 3.58E-05 3.76E-05 3.94E-05 4.11E-05 4.29E-05 4.47E-05 4.65E-05 4.82E-05 5.00E-05 5.18E-05 5.36E-05 7.09E-05 8.82E-05 1.06E-04 1.23E-04 1.40E-04 1.57E-04 1.75E-04 1.92E-04 2.09E-04 2.27E-04 2.44E-04 2.61E-04 2.78E-04 2.94E-04 3.11E-04 3.28E-04 3.44E-04 3.61E-04 Non-gestational 5-year Average Intake (ng/kg/ day) 7.32E-07 7.65E-07 7.98E-07 8.32E-07 8.65E-07 8.98E-07 9.31E-07 9.64E-07 9.97E-07 1.01E-06 1.03E-06 1.04E-06 1.06E-06 1.07E-06 1.09E-06 1.11E-06 1.12E-06 1.14E-06 1.16E-06 1.17E-06 1.19E-06 1.21E-06 1.23E-06 1.24E-06 1.26E-06 1.28E-06 1.30E-06 1.32E-06 1.34E-06 1.36E-06 1.38E-06 1.40E-06 1.42E-06 1.44E-06 1.46E-06 1.49E-06 1.53E-06 1.58E-06 Fat (ng/kg) 3.59E-02 3.74E-02 3.90E-02 4.06E-02 4.22E-02 4.37E-02 4.53E-02 4.69E-02 4.85E-02 4.92E-02 4.99E-02 5.06E-02 5.13E-02 5.20E-02 5.28E-02 5.35E-02 5.43E-02 5.51E-02 5.59E-02 5.67E-02 5.75E-02 5.83E-02 5.92E-02 6.00E-02 6.09E-02 6.17E-02 6.26E-02 6.35E-02 6.44E-02 6.53E-02 6.63E-02 6.72E-02 6.82E-02 6.91E-02 7.02E-02 7.12E-02 7.32E-02 7.53E-02 Body Burden (ng/kg) 1.31E-02 1.37E-02 1.42E-02 1.48E-02 1.54E-02 1.60E-02 1.66E-02 1.71E-02 1.77E-02 1.80E-02 1.82E-02 1.85E-02 1.88E-02 1.90E-02 1.93E-02 1.96E-02 1.99E-02 2.01E-02 2.04E-02 2.07E-02 2.10E-02 2.13E-02 2.16E-02 2.20E-02 2.23E-02 2.26E-02 2.29E-02 2.32E-02 2.36E-02 2.39E-02 2.43E-02 2.46E-02 2.50E-02 2.53E-02 2.57E-02 2.61E-02 2.68E-02 2.76E-02 Blood (ng/kg) 3.77E-04 3.94E-04 4.10E-04 4.27E-04 4.43E-04 4.60E-04 4.77E-04 4.93E-04 5.10E-04 5.17E-04 5.24E-04 5.32E-04 5.40E-04 5.47E-04 5.55E-04 5.63E-04 5.71E-04 5.79E-04 5.88E-04 5.96E-04 6.05E-04 6.13E-04 6.22E-04 6.31E-04 6.40E-04 6.49E-04 6.58E-04 6.68E-04 6.77E-04 6.87E-04 6.97E-04 7.07E-04 7.17E-04 7.27E-04 7.38E-04 7.48E-04 7.70E-04 7.92E-04 D R A FT: D O N O T C IT E O R Q U O T E C -193 This document is a draftfor review purposes only and does not constitute Agency policy. Non-gestational 5-year Average Intake (ng/kg/ day) 1.62E-06 1.67E-06 1.72E-06 1.77E-06 1.83E-06 1.88E-06 1.94E-06 2.00E-06 2.06E-06 2.12E-06 2.18E-06 2.25E-06 2.32E-06 2.39E-06 2.46E-06 2.53E-06 2.61E-06 2.68E-06 2.76E-06 2.85E-06 2.93E-06 3.02E-06 3.11E-06 3.21E-06 3.30E-06 3.40E-06 3.50E-06 3.61E-06 3.72E-06 3.83E-06 3.94E-06 4.06E-06 4.18E-06 4.31E-06 4.44E-06 4.57E-06 4.71E-06 4.85E-06 Fat (ng/kg) 7.74E-02 7.96E-02 8.19E-02 8.42E-02 8.66E-02 8.91E-02 9.16E-02 9.42E-02 9.69E-02 9.96E-02 1.02E-01 1.05E-01 1.08E-01 1.11E-01 1.15E-01 1.18E-01 1.21E-01 1.24E-01 1.28E-01 1.32E-01 1.35E-01 1.39E-01 1.43E-01 1.47E-01 1.51E-01 1.55E-01 1.59E-01 1.64E-01 1.68E-01 1.73E-01 1.78E-01 1.83E-01 1.88E-01 1.93E-01 1.98E-01 2.04E-01 2.09E-01 2.15E-01 Body Burden (ng/kg) 2.84E-02 2.92E-02 3.00E-02 3.09E-02 3.18E-02 3.27E-02 3.36E-02 3.46E-02 3.56E-02 3.66E-02 3.77E-02 3.87E-02 3.98E-02 4.10E-02 4.21E-02 4.33E-02 4.46E-02 4.58E-02 4.71E-02 4.85E-02 4.98E-02 5.13E-02 5.27E-02 5.42E-02 5.57E-02 5.73E-02 5.89E-02 6.05E-02 6.22E-02 6.40E-02 6.58E-02 6.76E-02 6.95E-02 7.15E-02 7.34E-02 7.55E-02 7.76E-02 7.98E-02 Blood (ng/kg) 8.14E-04 8.37E-04 8.61E-04 8.86E-04 9.11E-04 9.37E-04 9.63E-04 9.91E-04 1.02E-03 1.05E-03 1.08E-03 1.11E-03 1.14E-03 1.17E-03 1.20E-03 1.24E-03 1.27E-03 1.31E-03 1.35E-03 1.38E-03 1.42E-03 1.46E-03 1.50E-03 1.54E-03 1.59E-03 1.63E-03 1.68E-03 1.72E-03 1.77E-03 1.82E-03 1.87E-03 1.92E-03 1.97E-03 2.03E-03 2.08E-03 2.14E-03 2.20E-03 2.26E-03 Non-gestational 5-year Average Intake (ng/kg/ day) 4.99E-06 5.14E-06 5.30E-06 5.46E-06 5.62E-06 5.79E-06 5.96E-06 6.14E-06 6.33E-06 6.52E-06 6.71E-06 6.91E-06 7.12E-06 7.33E-06 7.55E-06 7.78E-06 8.01E-06 8.25E-06 8.50E-06 8.76E-06 9.02E-06 9.29E-06 9.57E-06 9.86E-06 1.02E-05 1.05E-05 1.08E-05 1.11E-05 1.14E-05 1.18E-05 1.21E-05 1.25E-05 1.29E-05 1.32E-05 1.36E-05 1.41E-05 1.45E-05 1.49E-05 Fat (ng/kg) 2.21E-01 2.27E-01 2.33E-01 2.39E-01 2.46E-01 2.53E-01 2.59E-01 2.66E-01 2.74E-01 2.81E-01 2.88E-01 2.96E-01 3.04E-01 3.12E-01 3.20E-01 3.28E-01 3.37E-01 3.46E-01 3.55E-01 3.64E-01 3.74E-01 3.84E-01 3.94E-01 4.05E-01 4.15E-01 4.26E-01 4.37E-01 4.48E-01 4.60E-01 4.72E-01 4.84E-01 4.96E-01 5.09E-01 5.22E-01 5.35E-01 5.49E-01 5.63E-01 5.77E-01 Body Burden (ng/kg) 8.20E-02 8.42E-02 8.66E-02 8.90E-02 9.14E-02 9.39E-02 9.65E-02 9.92E-02 1.02E-01 1.05E-01 1.07E-01 1.10E-01 1.13E-01 1.16E-01 1.19E-01 1.23E-01 1.26E-01 1.29E-01 1.33E-01 1.36E-01 1.40E-01 1.44E-01 1.48E-01 1.52E-01 1.56E-01 1.60E-01 1.64E-01 1.68E-01 1.73E-01 1.78E-01 1.82E-01 1.87E-01 1.92E-01 1.97E-01 2.02E-01 2.08E-01 2.13E-01 2.18E-01 Blood (ng/kg) 2.32E-03 2.39E-03 2.45E-03 2.52E-03 2.59E-03 2.66E-03 2.73E-03 2.80E-03 2.88E-03 2.95E-03 3.03E-03 3.11E-03 3.19E-03 3.28E-03 3.36E-03 3.45E-03 3.54E-03 3.64E-03 3.73E-03 3.83E-03 3.93E-03 4.04E-03 4.15E-03 4.25E-03 4.36E-03 4.48E-03 4.59E-03 4.71E-03 4.83E-03 4.96E-03 5.08E-03 5.21E-03 5.35E-03 5.49E-03 5.63E-03 5.77E-03 5.92E-03 6.07E-03 Non-gestational 5-year Average Intake (ng/kg/ day) 1.54E-05 1.58E-05 1.63E-05 1.68E-05 1.73E-05 1.78E-05 1.83E-05 1.89E-05 1.95E-05 2.00E-05 2.06E-05 2.13E-05 2.19E-05 2.25E-05 2.32E-05 2.39E-05 2.46E-05 2.54E-05 2.61E-05 2.69E-05 2.77E-05 2.86E-05 2.94E-05 3.03E-05 3.12E-05 3.21E-05 3.31E-05 3.41E-05 3.51E-05 3.62E-05 3.73E-05 3.84E-05 3.95E-05 4.07E-05 4.19E-05 4.32E-05 4.45E-05 4.58E-05 Fat (ng/kg) 5.92E-01 6.07E-01 6.23E-01 6.38E-01 6.54E-01 6.71E-01 6.88E-01 7.05E-01 7.23E-01 7.41E-01 7.60E-01 7.79E-01 7.98E-01 8.18E-01 8.38E-01 8.59E-01 8.80E-01 9.02E-01 9.24E-01 9.47E-01 9.70E-01 9.94E-01 1.02E+00 1.04E+00 1.07E+00 1.09E+00 1.12E+00 1.15E+00 1.18E+00 1.21E+00 1.24E+00 1.26E+00 1.29E+00 1.32E+00 1.35E+00 1.38E+00 1.41E+00 1.45E+00 Body Burden (ng/kg) 2.24E-01 2.30E-01 2.36E-01 2.42E-01 2.49E-01 2.55E-01 2.62E-01 2.69E-01 2.75E-01 2.83E-01 2.90E-01 2.97E-01 3.05E-01 3.13E-01 3.21E-01 3.29E-01 3.38E-01 3.46E-01 3.55E-01 3.64E-01 3.73E-01 3.83E-01 3.92E-01 4.02E-01 4.12E-01 4.23E-01 4.35E-01 4.46E-01 4.57E-01 4.68E-01 4.80E-01 4.92E-01 5.04E-01 5.14E-01 5.26E-01 5.39E-01 5.52E-01 5.66E-01 Blood (ng/kg) 6.23E-03 6.38E-03 6.55E-03 6.71E-03 6.88E-03 7.05E-03 7.23E-03 7.41E-03 7.60E-03 7.79E-03 7.99E-03 8.18E-03 8.39E-03 8.60E-03 8.81E-03 9.03E-03 9.25E-03 9.48E-03 9.71E-03 9.95E-03 1.02E-02 1.04E-02 1.07E-02 1.10E-02 1.12E-02 1.15E-02 1.18E-02 1.21E-02 1.23E-02 1.27E-02 1.30E-02 1.33E-02 1.35E-02 1.39E-02 1.42E-02 1.45E-02 1.49E-02 1.52E-02 D R A FT: D O N O T C IT E O R Q U O T E C -194 This document is a draftfor review purposes only and does not constitute Agency policy. Non-gestational 5-year Average Intake (ng/kg/ day) 4.72E-05 4.86E-05 5.01E-05 5.16E-05 5.31E-05 5.47E-05 5.64E-05 5.81E-05 5.98E-05 6.16E-05 6.34E-05 6.54E-05 6.73E-05 6.93E-05 7.14E-05 7.36E-05 7.58E-05 7.80E-05 8.04E-05 8.28E-05 8.53E-05 8.78E-05 9.05E-05 9.32E-05 9.60E-05 9.89E-05 1.02E-04 1.05E-04 1.08E-04 1.11E-04 1.15E-04 1.18E-04 1.22E-04 1.25E-04 1.29E-04 1.33E-04 1.37E-04 1.41E-04 Fat (ng/kg) 1.48E+00 1.52E+00 1.55E+00 1.59E+00 1.62E+00 1.66E+00 1.70E+00 1.74E+00 1.78E+00 1.82E+00 1.86E+00 1.90E+00 1.95E+00 1.99E+00 2.04E+00 2.06E+00 2.11E+00 2.15E+00 2.20E+00 2.25E+00 2.30E+00 2.35E+00 2.40E+00 2.46E+00 2.51E+00 2.57E+00 2.62E+00 2.68E+00 2.74E+00 2.80E+00 2.86E+00 2.92E+00 2.98E+00 3.05E+00 3.11E+00 3.18E+00 3.25E+00 3.32E+00 Body Burden (ng/kg) 5.80E-01 5.94E-01 6.08E-01 6.23E-01 6.38E-01 6.53E-01 6.69E-01 6.85E-01 7.02E-01 7.19E-01 7.36E-01 7.53E-01 7.71E-01 7.90E-01 8.08E-01 8.18E-01 8.37E-01 8.57E-01 8.77E-01 8.98E-01 9.19E-01 9.40E-01 9.62E-01 9.84E-01 1.01E+00 1.03E+00 1.05E+00 1.08E+00 1.10E+00 1.13E+00 1.15E+00 1.18E+00 1.21E+00 1.24E+00 1.26E+00 1.29E+00 1.32E+00 1.35E+00 Blood (ng/kg) 1.56E-02 1.59E-02 1.63E-02 1.67E-02 1.71E-02 1.75E-02 1.79E-02 1.83E-02 1.87E-02 1.91E-02 1.96E-02 2.00E-02 2.05E-02 2.09E-02 2.14E-02 2.16E-02 2.21E-02 2.26E-02 2.31E-02 2.36E-02 2.42E-02 2.47E-02 2.52E-02 2.58E-02 2.64E-02 2.69E-02 2.75E-02 2.81E-02 2.88E-02 2.94E-02 3.00E-02 3.07E-02 3.13E-02 3.20E-02 3.27E-02 3.34E-02 3.41E-02 3.48E-02 Non-gestational 5-year Average Intake (ng/kg/ day) 1.45E-04 1.50E-04 1.54E-04 1.59E-04 1.63E-04 1.68E-04 1.73E-04 1.79E-04 1.84E-04 1.89E-04 1.95E-04 2.01E-04 2.07E-04 2.13E-04 2.20E-04 2.26E-04 2.33E-04 2.40E-04 2.47E-04 2.55E-04 2.62E-04 2.70E-04 2.78E-04 2.86E-04 2.95E-04 3.04E-04 3.13E-04 3.22E-04 3.32E-04 3.42E-04 3.52E-04 3.63E-04 3.74E-04 3.85E-04 3.97E-04 4.08E-04 4.21E-04 4.33E-04 Fat (ng/kg) 3.45E+00 3.46E+00 3.53E+00 3.67E+00 3.75E+00 3.77E+00 3.86E+00 3.95E+00 4.04E+00 4.13E+00 4.22E+00 4.31E+00 4.44E+00 4.49E+00 4.59E+00 4.72E+00 4.81E+00 4.91E+00 5.00E+00 5.10E+00 5.21E+00 5.33E+00 5.44E+00 5.55E+00 5.66E+00 5.78E+00 5.90E+00 6.02E+00 6.14E+00 6.26E+00 6.39E+00 6.52E+00 6.65E+00 6.78E+00 6.92E+00 7.06E+00 7.20E+00 7.34E+00 Body Burden (ng/kg) 1.41E+00 1.41E+00 1.45E+00 1.51E+00 1.54E+00 1.55E+00 1.59E+00 1.63E+00 1.67E+00 1.71E+00 1.75E+00 1.79E+00 1.84E+00 1.87E+00 1.92E+00 1.97E+00 2.02E+00 2.06E+00 2.10E+00 2.15E+00 2.19E+00 2.25E+00 2.30E+00 2.35E+00 2.40E+00 2.46E+00 2.51E+00 2.57E+00 2.63E+00 2.68E+00 2.74E+00 2.80E+00 2.87E+00 2.93E+00 3.00E+00 3.06E+00 3.13E+00 3.20E+00 Blood (ng/kg) 3.62E-02 3.63E-02 3.71E-02 3.86E-02 3.94E-02 3.96E-02 4.06E-02 4.15E-02 4.24E-02 4.33E-02 4.43E-02 4.52E-02 4.66E-02 4.72E-02 4.82E-02 4.95E-02 5.05E-02 5.16E-02 5.24E-02 5.35E-02 5.47E-02 5.60E-02 5.71E-02 5.83E-02 5.94E-02 6.07E-02 6.19E-02 6.32E-02 6.44E-02 6.57E-02 6.71E-02 6.84E-02 6.98E-02 7.12E-02 7.26E-02 7.41E-02 7.56E-02 7.71E-02 Non-gestational 5-year Average Intake (ng/kg/ day) 4.46E-04 4.60E-04 4.74E-04 4.88E-04 5.02E-04 5.17E-04 5.33E-04 5.49E-04 5.65E-04 5.82E-04 6.00E-04 6.18E-04 6.36E-04 6.55E-04 6.75E-04 6.95E-04 7.16E-04 7.38E-04 7.60E-04 7.83E-04 8.06E-04 8.30E-04 8.55E-04 8.81E-04 9.07E-04 9.21E-04 9.35E-04 9.49E-04 9.63E-04 9.69E-04 9.77E-04 9.84E-04 9.91E-04 1.37E-03 1.39E-03 1.41E-03 1.43E-03 1.46E-03 Fat (ng/kg) 7.49E+00 7.64E+00 7.79E+00 7.95E+00 8.10E+00 8.26E+00 8.43E+00 8.59E+00 8.76E+00 8.93E+00 9.11E+00 9.29E+00 9.47E+00 9.65E+00 9.84E+00 1.00E+01 1.02E+01 1.04E+01 1.06E+01 1.08E+01 1.10E+01 1.13E+01 1.15E+01 1.17E+01 1.19E+01 1.20E+01 1.22E+01 1.30E+01 1.38E+01 1.43E+01 1.48E+01 1.52E+01 1.55E+01 1.56E+01 1.57E+01 1.59E+01 1.60E+01 1.62E+01 Body Burden (ng/kg) 3.27E+00 3.34E+00 3.42E+00 3.49E+00 3.57E+00 3.65E+00 3.73E+00 3.81E+00 3.89E+00 3.98E+00 4.07E+00 4.16E+00 4.25E+00 4.34E+00 4.44E+00 4.54E+00 4.64E+00 4.74E+00 4.84E+00 4.95E+00 5.06E+00 5.17E+00 5.28E+00 5.40E+00 5.52E+00 5.58E+00 5.64E+00 6.23E+00 6.92E+00 7.14E+00 7.34E+00 7.50E+00 7.64E+00 7.50E+00 7.58E+00 7.66E+00 7.75E+00 7.83E+00 Blood (ng/kg) 7.86E-02 8.02E-02 8.18E-02 8.34E-02 8.50E-02 8.67E-02 8.84E-02 9.02E-02 9.19E-02 9.37E-02 9.56E-02 9.74E-02 9.94E-02 1.01E-01 1.03E-01 1.05E-01 1.07E-01 1.09E-01 1.12E-01 1.14E-01 1.16E-01 1.18E-01 1.20E-01 1.23E-01 1.25E-01 1.26E-01 1.27E-01 1.37E-01 1.45E-01 1.50E-01 1.55E-01 1.59E-01 1.63E-01 1.63E-01 1.65E-01 1.66E-01 1.68E-01 1.69E-01 D R A FT: D O N O T C IT E O R Q U O T E C -195 This document is a draftfor review purposes only and does not constitute Agency policy. Non-gestational 5-year Average Intake (ng/kg/ day) 1.48E-03 1.50E-03 1.52E-03 1.54E-03 1.57E-03 1.59E-03 1.61E-03 1.64E-03 1.66E-03 1.69E-03 1.71E-03 1.74E-03 1.76E-03 1.79E-03 1.82E-03 1.84E-03 1.87E-03 1.90E-03 1.93E-03 1.96E-03 2.27E-03 2.34E-03 2.41E-03 2.48E-03 2.55E-03 2.63E-03 2.71E-03 2.79E-03 2.87E-03 2.96E-03 3.05E-03 3.14E-03 3.23E-03 3.33E-03 3.43E-03 3.53E-03 3.64E-03 3.75E-03 Fat (ng/kg) 1.63E+01 1.65E+01 1.66E+01 1.68E+01 1.69E+01 1.71E+01 1.73E+01 1.74E+01 1.76E+01 1.78E+01 1.79E+01 1.81E+01 1.83E+01 1.84E+01 1.86E+01 1.88E+01 1.91E+01 1.98E+01 2.05E+01 2.05E+01 2.14E+01 2.18E+01 2.22E+01 2.26E+01 2.31E+01 2.35E+01 2.39E+01 2.44E+01 2.49E+01 2.53E+01 2.58E+01 2.63E+01 2.68E+01 2.73E+01 2.78E+01 2.83E+01 2.88E+01 2.96E+01 Body Burden (ng/kg) 7.92E+00 8.00E+00 8.09E+00 8.18E+00 8.27E+00 8.36E+00 8.46E+00 8.55E+00 8.64E+00 8.74E+00 8.83E+00 8.93E+00 9.03E+00 9.13E+00 9.23E+00 9.33E+00 9.53E+00 1.01E+01 1.08E+01 1.05E+01 1.09E+01 1.11E+01 1.14E+01 1.16E+01 1.19E+01 1.22E+01 1.24E+01 1.27E+01 1.30E+01 1.33E+01 1.36E+01 1.39E+01 1.42E+01 1.45E+01 1.49E+01 1.52E+01 1.55E+01 1.61E+01 Blood (ng/kg) 1.71E-01 1.73E-01 1.74E-01 1.76E-01 1.78E-01 1.79E-01 1.81E-01 1.83E-01 1.84E-01 1.86E-01 1.88E-01 1.90E-01 1.92E-01 1.93E-01 1.95E-01 1.97E-01 2.00E-01 2.08E-01 2.14E-01 2.14E-01 2.25E-01 2.29E-01 2.33E-01 2.37E-01 2.42E-01 2.46E-01 2.51E-01 2.56E-01 2.61E-01 2.66E-01 2.71E-01 2.76E-01 2.81E-01 2.86E-01 2.91E-01 2.97E-01 3.02E-01 3.10E-01 Non-gestational 5-year Average Intake (ng/kg/ day) 3.81E-03 3.86E-03 4.22E-03 4.35E-03 4.48E-03 4.61E-03 4.75E-03 4.89E-03 5.04E-03 5.19E-03 5.35E-03 5.51E-03 5.67E-03 5.84E-03 5.93E-03 6.02E-03 6.20E-03 6.38E-03 6.57E-03 6.77E-03 6.98E-03 7.18E-03 7.40E-03 7.51E-03 7.62E-03 7.85E-03 8.09E-03 8.33E-03 8.58E-03 8.71E-03 8.84E-03 9.10E-03 9.37E-03 9.66E-03 9.94E-03 1.02E-02 1.06E-02 1.09E-02 Fat (ng/kg) 2.99E+01 3.00E+01 3.04E+01 3.10E+01 3.16E+01 3.21E+01 3.28E+01 3.34E+01 3.44E+01 3.57E+01 3.72E+01 3.81E+01 3.88E+01 3.95E+01 3.98E+01 4.00E+01 4.10E+01 4.18E+01 4.26E+01 4.34E+01 4.42E+01 4.50E+01 4.59E+01 4.63E+01 4.66E+01 4.71E+01 4.72E+01 4.74E+01 4.93E+01 4.98E+01 5.03E+01 5.13E+01 5.23E+01 5.33E+01 5.44E+01 5.54E+01 5.64E+01 5.75E+01 Body Burden (ng/kg) 1.63E+01 1.63E+01 1.66E+01 1.69E+01 1.73E+01 1.77E+01 1.81E+01 1.86E+01 1.94E+01 2.06E+01 2.12E+01 2.17E+01 2.23E+01 2.28E+01 2.30E+01 2.33E+01 2.38E+01 2.44E+01 2.49E+01 2.55E+01 2.61E+01 2.67E+01 2.73E+01 2.77E+01 2.78E+01 2.81E+01 2.79E+01 2.83E+01 2.99E+01 3.03E+01 3.06E+01 3.15E+01 3.22E+01 3.29E+01 3.38E+01 3.46E+01 3.54E+01 3.62E+01 Blood (ng/kg) 3.14E-01 3.14E-01 3.19E-01 3.25E-01 3.31E-01 3.37E-01 3.44E-01 3.50E-01 3.60E-01 3.74E-01 3.90E-01 3.99E-01 4.07E-01 4.14E-01 4.18E-01 4.20E-01 4.30E-01 4.38E-01 4.46E-01 4.55E-01 4.63E-01 4.72E-01 4.81E-01 4.85E-01 4.88E-01 4.94E-01 4.95E-01 4.97E-01 5.17E-01 5.22E-01 5.27E-01 5.38E-01 5.49E-01 5.59E-01 5.70E-01 5.81E-01 5.92E-01 6.03E-01 Non-gestational 5-year Average Intake (ng/kg/ day) 1.12E-02 1.15E-02 1.19E-02 1.22E-02 1.26E-02 1.30E-02 1.34E-02 1.38E-02 1.42E-02 1.46E-02 1.50E-02 1.55E-02 1.60E-02 1.64E-02 1.69E-02 1.74E-02 1.80E-02 1.85E-02 1.91E-02 1.96E-02 2.02E-02 2.08E-02 2.14E-02 2.21E-02 2.28E-02 2.34E-02 2.41E-02 2.49E-02 2.56E-02 2.64E-02 2.72E-02 2.80E-02 2.88E-02 2.97E-02 3.06E-02 3.15E-02 3.24E-02 3.34E-02 Fat (ng/kg) 5.86E+01 5.96E+01 6.05E+01 6.14E+01 6.24E+01 6.38E+01 6.56E+01 6.74E+01 6.87E+01 6.94E+01 7.06E+01 7.19E+01 7.30E+01 7.38E+01 7.55E+01 7.69E+01 7.84E+01 7.99E+01 8.13E+01 8.29E+01 8.44E+01 8.60E+01 8.76E+01 8.92E+01 9.09E+01 9.26E+01 9.44E+01 9.64E+01 9.79E+01 9.98E+01 1.02E+02 1.04E+02 1.06E+02 1.07E+02 1.10E+02 1.12E+02 1.14E+02 1.16E+02 Body Burden (ng/kg) 3.71E+01 3.79E+01 3.85E+01 3.92E+01 4.01E+01 4.15E+01 4.31E+01 4.42E+01 4.53E+01 4.59E+01 4.69E+01 4.78E+01 4.87E+01 4.96E+01 5.11E+01 5.23E+01 5.36E+01 5.49E+01 5.62E+01 5.75E+01 5.89E+01 6.03E+01 6.18E+01 6.32E+01 6.48E+01 6.63E+01 6.80E+01 6.98E+01 7.13E+01 7.30E+01 7.48E+01 7.66E+01 7.85E+01 8.04E+01 8.28E+01 8.51E+01 8.69E+01 8.88E+01 Blood (ng/kg) 6.14E-01 6.25E-01 6.35E-01 6.43E-01 6.54E-01 6.69E-01 6.87E-01 7.07E-01 7.20E-01 7.28E-01 7.40E-01 7.54E-01 7.66E-01 7.74E-01 7.92E-01 8.07E-01 8.22E-01 8.37E-01 8.53E-01 8.69E-01 8.85E-01 9.01E-01 9.18E-01 9.35E-01 9.53E-01 9.71E-01 9.89E-01 1.01E+00 1.03E+00 1.05E+00 1.07E+00 1.09E+00 1.11E+00 1.13E+00 1.15E+00 1.17E+00 1.20E+00 1.22E+00 D R A FT: D O N O T C IT E O R Q U O T E C -196 This document is a draftfor review purposes only and does not constitute Agency policy. Non-gestational 5-year Average Intake (ng/kg/ day) 3.44E-02 3.54E-02 3.65E-02 3.76E-02 3.87E-02 3.99E-02 4.11E-02 4.23E-02 4.36E-02 4.49E-02 4.63E-02 4.76E-02 4.91E-02 5.05E-02 5.21E-02 5.36E-02 5.52E-02 5.69E-02 5.86E-02 6.03E-02 6.22E-02 6.40E-02 6.59E-02 6.79E-02 7.00E-02 7.21E-02 7.42E-02 7.64E-02 7.87E-02 8.11E-02 8.35E-02 8.60E-02 8.86E-02 9.13E-02 9.40E-02 9.68E-02 9.97E-02 1.03E-01 Fat (ng/kg) 1.18E+02 1.20E+02 1.22E+02 1.24E+02 1.27E+02 1.29E+02 1.32E+02 1.34E+02 1.37E+02 1.40E+02 1.43E+02 1.45E+02 1.48E+02 1.51E+02 1.53E+02 1.56E+02 1.59E+02 1.62E+02 1.65E+02 1.69E+02 1.72E+02 1.74E+02 1.78E+02 1.81E+02 1.84E+02 1.88E+02 1.91E+02 1.95E+02 1.99E+02 2.03E+02 2.07E+02 2.11E+02 2.15E+02 2.19E+02 2.23E+02 2.27E+02 2.32E+02 2.36E+02 Body Burden (ng/kg) 9.08E+01 9.28E+01 9.47E+01 9.73E+01 9.96E+01 1.02E+02 1.04E+02 1.07E+02 1.10E+02 1.13E+02 1.16E+02 1.19E+02 1.22E+02 1.25E+02 1.28E+02 1.31E+02 1.34E+02 1.38E+02 1.41E+02 1.45E+02 1.48E+02 1.52E+02 1.55E+02 1.59E+02 1.63E+02 1.67E+02 1.71E+02 1.76E+02 1.81E+02 1.86E+02 1.90E+02 1.95E+02 2.00E+02 2.05E+02 2.10E+02 2.16E+02 2.22E+02 2.27E+02 Blood (ng/kg) 1.24E+00 1.26E+00 1.28E+00 1.30E+00 1.33E+00 1.35E+00 1.38E+00 1.40E+00 1.43E+00 1.47E+00 1.49E+00 1.52E+00 1.55E+00 1.58E+00 1.61E+00 1.64E+00 1.67E+00 1.70E+00 1.73E+00 1.77E+00 1.80E+00 1.83E+00 1.86E+00 1.90E+00 1.93E+00 1.97E+00 2.01E+00 2.05E+00 2.09E+00 2.13E+00 2.17E+00 2.21E+00 2.25E+00 2.30E+00 2.34E+00 2.38E+00 2.43E+00 2.48E+00 Non-gestational 5-year Average Intake (ng/kg/ day) 1.06E-01 1.09E-01 1.12E-01 1.16E-01 1.19E-01 1.23E-01 1.26E-01 1.30E-01 1.34E-01 1.38E-01 1.42E-01 1.46E-01 1.51E-01 1.55E-01 1.60E-01 1.65E-01 1.70E-01 1.75E-01 1.80E-01 1.86E-01 1.91E-01 1.97E-01 2.03E-01 2.09E-01 2.15E-01 2.22E-01 2.28E-01 2.35E-01 2.42E-01 2.49E-01 2.57E-01 2.65E-01 2.72E-01 2.81E-01 2.89E-01 2.98E-01 3.07E-01 3.16E-01 Fat (ng/kg) 2.41E+02 2.45E+02 2.50E+02 2.55E+02 2.60E+02 2.65E+02 2.70E+02 2.75E+02 2.80E+02 2.86E+02 2.92E+02 2.97E+02 3.03E+02 3.08E+02 3.14E+02 3.20E+02 3.27E+02 3.33E+02 3.39E+02 3.46E+02 3.53E+02 3.60E+02 3.66E+02 3.73E+02 3.81E+02 3.88E+02 3.96E+02 4.03E+02 4.11E+02 4.20E+02 4.28E+02 4.36E+02 4.45E+02 4.53E+02 4.62E+02 4.71E+02 4.80E+02 4.90E+02 Body Burden (ng/kg) 2.33E+02 2.39E+02 2.44E+02 2.51E+02 2.57E+02 2.64E+02 2.71E+02 2.78E+02 2.86E+02 2.93E+02 3.01E+02 3.09E+02 3.16E+02 3.24E+02 3.33E+02 3.42E+02 3.51E+02 3.60E+02 3.69E+02 3.79E+02 3.89E+02 3.99E+02 4.09E+02 4.20E+02 4.31E+02 4.43E+02 4.55E+02 4.67E+02 4.79E+02 4.92E+02 5.05E+02 5.19E+02 5.32E+02 5.46E+02 5.61E+02 5.75E+02 5.91E+02 6.07E+02 Blood (ng/kg) 2.52E+00 2.57E+00 2.62E+00 2.67E+00 2.72E+00 2.77E+00 2.83E+00 2.88E+00 2.94E+00 3.00E+00 3.06E+00 3.11E+00 3.17E+00 3.23E+00 3.29E+00 3.36E+00 3.42E+00 3.49E+00 3.56E+00 3.63E+00 3.70E+00 3.77E+00 3.84E+00 3.91E+00 3.99E+00 4.07E+00 4.15E+00 4.23E+00 4.31E+00 4.40E+00 4.48E+00 4.57E+00 4.66E+00 4.75E+00 4.84E+00 4.93E+00 5.03E+00 5.13E+00 Non-gestational 5-year Average Intake (ng/kg/ day) 3.25E-01 3.35E-01 3.45E-01 3.56E-01 3.66E-01 3.77E-01 3.89E-01 4.00E-01 4.12E-01 4.25E-01 4.37E-01 4.50E-01 4.64E-01 4.92E-01 5.07E-01 5.22E-01 5.54E-01 5.71E-01 5.88E-01 6.05E-01 6.23E-01 6.61E-01 6.81E-01 7.02E-01 7.23E-01 7.44E-01 7.67E-01 7.90E-01 8.13E-01 8.38E-01 8.63E-01 8.89E-01 9.16E-01 9.43E-01 9.71E-01 1.00E+00 1.06E+00 1.09E+00 Fat (ng/kg) 4.99E+02 5.09E+02 5.19E+02 5.30E+02 5.40E+02 5.51E+02 5.62E+02 5.73E+02 5.84E+02 5.96E+02 6.08E+02 6.20E+02 6.32E+02 6.58E+02 6.71E+02 6.85E+02 7.12E+02 7.27E+02 7.41E+02 7.56E+02 7.71E+02 8.03E+02 8.19E+02 8.36E+02 8.53E+02 8.70E+02 8.88E+02 9.06E+02 9.25E+02 9.44E+02 9.63E+02 9.83E+02 1.00E+03 1.02E+03 1.05E+03 1.07E+03 1.11E+03 1.14E+03 Body Burden (ng/kg) 6.23E+02 6.40E+02 6.57E+02 6.75E+02 6.93E+02 7.12E+02 7.31E+02 7.51E+02 7.71E+02 7.92E+02 8.13E+02 8.35E+02 8.58E+02 9.05E+02 9.29E+02 9.55E+02 1.01E+03 1.04E+03 1.06E+03 1.09E+03 1.12E+03 1.18E+03 1.22E+03 1.25E+03 1.28E+03 1.32E+03 1.36E+03 1.39E+03 1.43E+03 1.47E+03 1.51E+03 1.55E+03 1.60E+03 1.64E+03 1.69E+03 1.73E+03 1.83E+03 1.88E+03 Blood (ng/kg) 5.23E+00 5.34E+00 5.44E+00 5.55E+00 5.66E+00 5.77E+00 5.89E+00 6.00E+00 6.12E+00 6.25E+00 6.37E+00 6.50E+00 6.63E+00 6.89E+00 7.03E+00 7.17E+00 7.46E+00 7.61E+00 7.77E+00 7.92E+00 8.08E+00 8.41E+00 8.58E+00 8.76E+00 8.94E+00 9.12E+00 9.31E+00 9.50E+00 9.69E+00 9.89E+00 1.01E+01 1.03E+01 1.05E+01 1.07E+01 1.10E+01 1.12E+01 1.16E+01 1.19E+01 D R A FT: D O N O T C IT E O R Q U O T E C -197 This document is a draftfor review purposes only and does not constitute Agency policy. Non-gestational 5-year Average Intake (ng/kg/ day) 1.13E+00 1.16E+00 1.19E+00 1.23E+00 1.27E+00 1.31E+00 1.34E+00 1.38E+00 1.43E+00 1.47E+00 1.51E+00 1.56E+00 1.61E+00 1.65E+00 1.70E+00 1.75E+00 1.81E+00 1.86E+00 1.92E+00 1.97E+00 2.03E+00 2.09E+00 2.16E+00 2.22E+00 2.29E+00 2.36E+00 2.43E+00 2.50E+00 2.58E+00 2.65E+00 2.73E+00 2.82E+00 2.90E+00 2.99E+00 3.08E+00 3.17E+00 3.26E+00 3.36E+00 Fat (ng/kg) 1.16E+03 1.18E+03 1.21E+03 1.23E+03 1.26E+03 1.29E+03 1.31E+03 1.34E+03 1.37E+03 1.40E+03 1.43E+03 1.46E+03 1.49E+03 1.52E+03 1.55E+03 1.59E+03 1.62E+03 1.66E+03 1.69E+03 1.73E+03 1.77E+03 1.80E+03 1.84E+03 1.88E+03 1.92E+03 1.97E+03 2.01E+03 2.05E+03 2.10E+03 2.15E+03 2.19E+03 2.24E+03 2.29E+03 2.34E+03 2.40E+03 2.45E+03 2.51E+03 2.56E+03 Body Burden (ng/kg) 1.94E+03 1.99E+03 2.04E+03 2.10E+03 2.16E+03 2.22E+03 2.28E+03 2.35E+03 2.41E+03 2.48E+03 2.55E+03 2.62E+03 2.69E+03 2.77E+03 2.85E+03 2.93E+03 3.01E+03 3.10E+03 3.18E+03 3.27E+03 3.37E+03 3.46E+03 3.56E+03 3.66E+03 3.76E+03 3.87E+03 3.98E+03 4.09E+03 4.21E+03 4.33E+03 4.45E+03 4.58E+03 4.71E+03 4.85E+03 4.98E+03 5.13E+03 5.27E+03 5.42E+03 Blood (ng/kg) 1.21E+01 1.24E+01 1.27E+01 1.29E+01 1.32E+01 1.35E+01 1.38E+01 1.40E+01 1.43E+01 1.46E+01 1.50E+01 1.53E+01 1.56E+01 1.59E+01 1.63E+01 1.66E+01 1.70E+01 1.74E+01 1.77E+01 1.81E+01 1.85E+01 1.89E+01 1.93E+01 1.97E+01 2.02E+01 2.06E+01 2.11E+01 2.15E+01 2.20E+01 2.25E+01 2.30E+01 2.35E+01 2.40E+01 2.46E+01 2.51E+01 2.57E+01 2.63E+01 2.69E+01 Non-gestational 5-year Average Intake (ng/kg/ day) 3.46E+00 3.57E+00 3.67E+00 3.78E+00 3.90E+00 4.01E+00 4.13E+00 4.26E+00 4.39E+00 4.52E+00 4.65E+00 4.79E+00 4.94E+00 5.08E+00 5.24E+00 5.39E+00 5.56E+00 5.72E+00 5.89E+00 6.07E+00 6.25E+00 6.44E+00 6.63E+00 6.83E+00 7.04E+00 7.25E+00 7.47E+00 7.69E+00 7.92E+00 8.16E+00 8.40E+00 8.66E+00 8.92E+00 9.18E+00 9.46E+00 9.74E+00 1.00E+01 1.00E+01 Fat (ng/kg) 2.62E+03 2.68E+03 2.74E+03 2.80E+03 2.87E+03 2.93E+03 3.00E+03 3.07E+03 3.14E+03 3.22E+03 3.29E+03 3.37E+03 3.45E+03 3.53E+03 3.61E+03 3.69E+03 3.78E+03 3.87E+03 3.96E+03 4.06E+03 4.15E+03 4.25E+03 4.36E+03 4.46E+03 4.57E+03 4.68E+03 4.79E+03 4.91E+03 5.03E+03 5.15E+03 5.28E+03 5.41E+03 5.54E+03 5.68E+03 5.82E+03 5.97E+03 6.10E+03 6.12E+03 Body Burden (ng/kg) 5.58E+03 5.74E+03 5.90E+03 6.07E+03 6.25E+03 6.42E+03 6.61E+03 6.80E+03 6.99E+03 7.20E+03 7.40E+03 7.62E+03 7.83E+03 8.06E+03 8.29E+03 8.53E+03 8.78E+03 9.03E+03 9.29E+03 9.56E+03 9.84E+03 1.01E+04 1.04E+04 1.07E+04 1.10E+04 1.13E+04 1.17E+04 1.20E+04 1.24E+04 1.27E+04 1.31E+04 1.35E+04 1.39E+04 1.43E+04 1.47E+04 1.51E+04 1.55E+04 1.56E+04 Blood (ng/kg) 2.75E+01 2.81E+01 2.87E+01 2.94E+01 3.01E+01 3.07E+01 3.15E+01 3.22E+01 3.29E+01 3.37E+01 3.45E+01 3.53E+01 3.61E+01 3.69E+01 3.78E+01 3.87E+01 3.96E+01 4.06E+01 4.15E+01 4.25E+01 4.35E+01 4.46E+01 4.56E+01 4.67E+01 4.79E+01 4.90E+01 5.02E+01 5.15E+01 5.27E+01 5.40E+01 5.53E+01 5.67E+01 5.81E+01 5.95E+01 6.10E+01 6.25E+01 6.39E+01 6.41E+01 Non-gestational 5-year Average Intake (ng/kg/ day) 1.34E+01 1.67E+01 2.00E+01 2.33E+01 2.67E+01 3.00E+01 3.33E+01 3.67E+01 4.00E+01 4.33E+01 4.67E+01 5.00E+01 5.33E+01 5.67E+01 6.00E+01 6.33E+01 6.67E+01 7.00E+01 7.33E+01 7.67E+01 8.00E+01 8.33E+01 8.67E+01 9.00E+01 9.33E+01 9.67E+01 1.00E+02 1.10E+02 1.20E+02 Fat (ng/kg) 7.77E+03 9.43E+03 1.11E+04 1.27E+04 1.43E+04 1.60E+04 1.76E+04 1.92E+04 2.08E+04 2.24E+04 2.40E+04 2.57E+04 2.73E+04 2.89E+04 3.05E+04 3.21E+04 3.37E+04 3.53E+04 3.69E+04 3.85E+04 4.01E+04 4.17E+04 4.33E+04 4.49E+04 4.65E+04 4.81E+04 4.97E+04 5.45E+04 5.94E+04 Body Burden (ng/kg) 2.05E+04 2.55E+04 3.05E+04 3.54E+04 4.03E+04 4.53E+04 5.02E+04 5.51E+04 6.00E+04 6.49E+04 6.97E+04 7.46E+04 7.94E+04 8.43E+04 8.91E+04 9.39E+04 9.87E+04 1.04E+05 1.08E+05 1.13E+05 1.18E+05 1.23E+05 1.27E+05 1.32E+05 1.37E+05 1.41E+05 1.46E+05 1.60E+05 1.74E+05 1 Blood (ng/kg) 8.15E+01 9.88E+01 1.16E+02 1.33E+02 1.50E+02 1.67E+02 1.84E+02 2.01E+02 2.18E+02 2.35E+02 2.52E+02 2.69E+02 2.86E+02 3.03E+02 3.19E+02 3.36E+02 3.53E+02 3.70E+02 3.87E+02 4.04E+02 4.20E+02 4.37E+02 4.54E+02 4.71E+02 4.88E+02 5.04E+02 5.21E+02 5.72E+02 6.22E+02 B o d y B u r d e n I1 B lo o d I ____(n g /k g )___ 8 .6 6 E -0 3 | 2 .3 0 E -0 4 || 9 .0 1 E -0 3 | 2 .3 9 E -0 4 II 9 .3 7 E -0 3 | 2 .4 9 E -0 4 || 9 .7 2 E -0 3 | 2 .5 8 E -0 4 II 1 .0 1 E -0 2 | 2 .6 7 E -0 4 II 1 .0 4 E -0 2 | 2 .7 7 E -0 4 || 1 .0 8 E -0 2 | 2 .8 6 E -0 4 || 1 .0 9 E -0 2 | 2 .9 0 E -0 4 || 1 .1 1 E -0 2 | 2 .9 5 E -0 4 II 1 .1 3 E -0 2 | 2 .9 9 E -0 4 II 1 .1 4 E -0 2 | 3 .0 3 E -0 4 || 1 .1 6 E -0 2 | 3 .0 8 E -0 4 II 1 .1 8 E -0 2 | 3 .1 2 E -0 4 || 1 .2 0 E -0 2 | 3 .1 7 E -0 4 || 1 .2 1 E -0 2 | 3 .2 2 E -0 4 || 1 .2 3 E -0 2 | 3 .2 6 E -0 4 || 1 .2 5 E -0 2 | 3 .3 1 E -0 4 || 1 .2 7 E -0 2 | 3 .3 6 E -0 4 II 1 .2 9 E -0 2 | 3 .4 1 E -0 4 || 1 .3 1 E -0 2 | 3 .4 6 E -0 4 II 1 .3 2 E -0 2 | 3 .5 1 E -0 4 || 1 .3 4 E -0 2 | 3 .5 6 E -0 4 || 1 .3 6 E -0 2 | 3 .6 1 E -0 4 || 1 .3 8 E -0 2 | 3 .6 7 E -0 4 || 1 .4 0 E -0 2 | 3 .7 2 E -0 4 || 1 .4 2 E -0 2 | 3 .7 7 E -0 4 || 1 .4 5 E -0 2 | 3 .8 3 E -0 4 || 1 .4 7 E -0 2 | 3 .8 9 E -0 4 || 1 .4 9 E -0 2 | 3 .9 4 E -0 4 || 1 .5 1 E -0 2 | 4 .0 0 E -0 4 || 1 .5 3 E -0 2 | 4 .0 6 E -0 4 II 1 .5 6 E -0 2 | 4 .1 2 E -0 4 || 1 .5 8 E -0 2 | 4 .1 8 E -0 4 || 1 .6 0 E -0 2 | 4 .2 5 E -0 4 II 1 .6 5 E -0 2 | 4 .3 7 E -0 4 II 1 .7 0 E -0 2 | 4 .5 0 E -0 4 || 1 .7 5 E -0 2 | 4 .6 3 E -0 4 || 1 .8 0 E -0 2 | 4 .7 6 E -0 4 || 'bi -bax 3 R fl JO flCS2 $w bf CoT CT o CT O COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT COT w bi 00 wO '-- 1 C T wkcOo w>n w in wco kO w Ci>T wkO w Ok l> 00 w 00 w 00 00 wcOok w ro~k- w O wkO wO w<n wOk w o '-- 1 '-- 1 '-- 1 C T wOk CT w co w 00 CO wco w 00 w T j- i n wOk w n kO w o\ ko wn wo r - 00 wkO w CT 00 ww 00 r c o w<n wO '-- 1 C T wO w rc n C T C T C T C T C T C T C T C T C T C T C T C T C T co co co co co co co co co co co co r c r c r c r c r c r c r c r c I r~- r~- r~- r~- r~- r~- r~- kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO "O o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o ' h i w00 w CT w in O k co k O www oo O k co k O wR ok w o wco w oo w kO o wR o w O ww CT '-- 1 w w kO '-- 1 wR '-- 1 w Ok '-- 1 w CT wco w CT CT w kO CT www 00 CT fco> CrcT wwww rc Urc) 0rc0 o w CT w ww VO w rcn w0n0 w CkOT wRkO bX R 00 00 00 cK cK cK ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 IB o d y B u rd e n I1 B lo o d | ___ (ng/kg)___ | 4.78E -04 | 1.27E-05 || | 5.14E -04 | 1.37E-05 || | 5.51E -04 | 1.47E-05 || | 5.88E -04 | 1.57E-05 || | 6.24E -04 | 1.67E-05 || | 6.61E -04 | 1.76E-05 || | 6.97E -04 | 1.86E-05 II | 7.34E -04 | 1.96E-05 II | 7.70E -04 | 2.05E -05 II | 8.07E -04 | 2.15E -05 || | 8.43E -04 | 2.25E -05 || | 2.35E -05 || | 9.17E -04 | 2.44E -05 II | 9.53E -04 | 2.54E -05 || | 9.90E -04 | 2.64E -05 II | 1.03E-03 | 2.74E -05 || | 1.06E-03 | 2.83E -05 || | 1.10E-03 | 2.93E -05 || | 1.46E-03 | 3.90E -05 || | 1.83E-03 | 4.86E -05 II | 2.19E -03 | 5.83E -05 || | 2.55E -03 | 6.80E -05 || | 2.92E -03 | 7.76E -05 II | 3.28E -03 | 8.73E -05 || | 3.64E -03 | 9.69E -05 II | 4.01E -03 | 1.07E -04 II | 4.37E -03 | 1.16E -04 II | 4.74E -03 | 1.26E -04 || | 5.10E -03 | 1.36E -04 II | 5.46E -03 | 1.45E -04 || | 5.82E -03 | 1.55E -04 || | 6.17E -03 | 1.64E -04 || | 6.53E -03 | 1.73E -04 || | 6.88E -03 | 1.83E -04 || | 7.24E -03 | 1.92E -04 II | 7.59E -03 | 2.02E -04 II | 7.95E -03 | 2.11E -04 || | 8.30E -03 | 2.20E -04 || 'bJ -3bax o1 C5 = Ho00 00 1w CC bi O CC O CC O CC O CC O CC O CC O CC O cc O cc O CC O roc roc CC O CC O cc O CC O rc O CO O CO O CO O CO O CO O CO O CO O CT O CT O CT O CT O CT O CT O CT O CT O CT O CT O CT O CT O CT O C5 'b=i w CT worc wO w 0\ w0n0 wI"- kO wIl*>-- w k0O0 w<n Ok w o w wrc CT w CrcT w w n wo kO w Ok kO w Or k- wOl>k w CT kO w in w kO wI*-- w Ok CO CT w CT w O wO'-- wO CT w Ok CT w00 wr- CO w knO w>n kO w ww C0O0 CT Ok w O w Ok O CT CT CT CT CT CT CT CT CT CO >n kO R 00 Ok CT CT I"O 0o0 0o0 0o0 0o0 0o0 0o0 0o0 0o0 0o0 0o0 0o0 0o0 0o0 0o0 0o0 0o0 0o0 ro- ro- ro- ro- ro- ro- ro- ro~- ro- ro- ro- ro- ro- ro- ro- ro- ro- ro- ro- ro~- ro- 1cb "hi w crcc 3bx wR kO wfo > w crcc ri n wR kO <n wfo > w crcc kO kO wR kO kO wfo > w crcc RR wR kO R wfo> wrrcc 00 00 wR kO 00 wfo> wrrcc Ok Ok wR kO Ok wf > w CO wR O CO kO ' 1' 1' wf > w CO O CO CT CT wR kO CT w< > w CO O CO CO CO wR kO CO w< > w CO O CO wR kO w< > w CO wkf5 w w CO O CO kO Ok CO n >n >n n kO wkf5 w kO Ok kO kO w CT CO R w n kO R r~- r- r- r~- r- r- r- r~- kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO m m "O bf o o o o o o o o o o o o o o o o O o o o o o o O o O O O o o o o o o o o o o wwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwww flbi kO C T Ok 00 kO 00 00 00 00 00 r- r- r~- r- r- r- kO kO kO kO kO kO m n m CO Ok r- n CO 00 00 CO Ok Ok Ok Ok 00 00 00 00 O '-- 1 C T CO in kO r- 00 Ok o '-- 1 C T CO in kO r- 00 Ok Ok Ok 00 00 00 00 00 o '-- 1 CT CO n kO R 00 Ok CT CT CT CT CT CT CT CT CT CT CO n kO R 00 Ok = O "uO m bf o m o flcb CO ww bf O '-- i s-'' n mmmmmmm ooooooo wwwwwww 00 CT 00 n m Ok CT CO Ok kO m o CT CT CT CO CO mmmmmmmmmm oooooooooo wwwwwwwwww CT Ok r- kO CO '-- 1 in Ok Ok kO CT CO kO Ok O rCO o n n n ko kO R R R mmmmmm oooooooooo wwwwwwwwww o 00 m CO r- O '-- i 00 C T in Ok O o '-- i 00 00 00 Ok Ok Ok ooooooo wwwwwww r- r- 00 CT in Ok CO kO o CT CT CT CO CO o w fi n n n n n n n n CO CO o bf o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o O O w cb wwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwww 'bi 00 n 00 C T n CO C T C T O Ok Ok 00 00 r- kO kO n CO CT CT Ok CO kO Ok CT kO Ok CT CT CT = 00 r- kO kO n CO O CT CO in in kO r- 00 Ok o --- i C T CO in kO O 00 r- kO in in CO CO O tSl w CT CO n kO R 00 Ok CT CT CT CT CT CT CT CT CT CO n kO R 00 Ok 0 Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok Ok 00 00 00 00 00 00 00 00 00 00 o "O O O O O O O O O O O O O O O O O O O O O O O O O O O O o O o o o o o o o o ro cb 'bx w w CO WO k W CO WO k W CO WO k W CO WO k W CO WO k W CO WO k W CO WO k W CO wO k W CO WO k w CO WO k W CO wO k W CO wo k O CO kO O CO kO O CO kO O CO kO O CO kO O CO kO O CO kO O CO kO o CO kO O CO kO O CO kO o CO kO o bx ,-- 1 .-- 1 C T C T C T CO CO CO t J- n in n kO kO kO R R R 00 00 00 Ok Ok Ok .-- i ,-- 1 .-- 1 C T C T C T CO CO CO u This document is a draftfor review purposes only and does not constitute Agency policy. C-198 DRAFT: DO NOT CITE OR QUOTE B lo o d (ng/kg) | 4 .1 2 E -0 3 || | 4 .2 3 E -0 3 II | 4 .3 5 E -0 3 || | 4 .4 7 E -0 3 II | 4 .5 9 E -0 3 || I 4 .7 1 E -0 3 II | 4 .8 4 E -0 3 II | 4 .9 7 E -0 3 || | 5 .1 0 E -0 3 || | 5 .2 4 E -0 3 || | 5 .3 8 E -0 3 || | 5 .5 3 E -0 3 || | 5 .6 8 E -0 3 || | 5 .8 3 E -0 3 || | 5 .9 8 E -0 3 || | 6 .1 4 E -0 3 || | 6 .3 1 E -0 3 || | 6 .4 8 E -0 3 || | 6 .6 5 E -0 3 || | 6 .8 2 E -0 3 || | 7 .0 1 E -0 3 || | 7 .1 9 E -0 3 || | 7 .3 8 E -0 3 II | 7 .5 8 E -0 3 || | 7 .8 0 E -0 3 || | 8 .0 1 E -0 3 II | 8 .2 2 E -0 3 II | 8 .4 4 E -0 3 || | 8 .6 8 E -0 3 II | 8 .8 9 E -0 3 II I 9 .1 2 E -0 3 II | 9 .3 4 E -0 3 II | 9 .5 9 E -0 3 || | 9 .8 3 E -0 3 || | 1 .0 1 E -0 2 II | 1 .0 3 E -0 2 || | 1 .0 6 E -0 2 || | 1 .0 9 E -0 2 || fl "fi is ^ w o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o fl flmw bx w Ok kO w kO w o\ kO w Cl>O w 0l>0 w CO 00 w 00 00 w CO Ok w 00 Ok w o wo ww n ww r~- wo '-- 1 --- i <N <N CO w o w CO in w CO kO wr - kO w ml> ww Osi CO 00 a \ w 00 wr - o ww ---ni (N w CO CO ww Osi n w kO w o\ kO w ol>\ w o\ 00 w CO o wo'--1 w (N w CO CO fi w <N <N <N <N <N <N <N <N (N (N (N (N <N (N <N CO CO CO CO CO CO CO CO CO CO to Oo 'bi o o o o o o o o o O O O O o O O O o o O o o O o o o O o o o o o o o o o fio ts -a to w w <N Ok wco o wco '--i w in CM w kO CO w 00 w o kO ww Ols>i >n 00 w 00 Ok ww <'--Ni >n <N w O wwww Ok O in kO 00 o www kO Osi o \ '--1 CO w kO kO w 00 w Osi wo O <N w (N w Osi kO ww Osi CO 00 o w (N w n w 00 kO w 00 00 ww --- i n '--i CO w o\ in w CO 00 rti O o 00 00 00 00 00co n n n n n n kO kO kO kO kO kO R R R R R cK cK cK cK a & in o >n o >n n n >n n n n n n n n n oooooooooooo n o n o n o n o n o n o n o n o n o n o n o n o n o n o n o n o n o n o n o n o n o n o n o n o 00 00wco woo wCO w 00 00 '--1 00 00 00kO kO l> i> wCO wOk wOnk w<Ok wkO o wCO wOk '--i wn <N wor CO wOk CO wkO w in w \o ww kO< T \ Rr - wU) w wCO O w(N w <N w CO w w wor wCO w n kO r - wn w Ro w wCN wn w wCN wU) '--i CO in r - " bflx ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 <N <N <N <N <N <N <N (N (N (N (N (N (N (N CO CO CO CO CO CO CO CO CO CO IBody Burden I1 Blood | ____(n g /k g )___ | 5 .5 3 E -0 2 | 1 .4 5 E -0 3 || | 5 .6 9 E -0 2 | 1 .4 9 E -0 3 || | 5 .8 5 E -0 2 | 1 .5 3 E -0 3 || | 6 .0 2 E -0 2 | 1 .5 8 E -0 3 || | 6 .1 9 E -0 2 | 1 .6 2 E -0 3 || | 6 .3 7 E -0 2 | 1 .6 7 E -0 3 || | 6 .5 5 E -0 2 | 1 .7 2 E -0 3 || | 6 .7 4 E -0 2 | 1 .7 6 E -0 3 II | 6 .9 3 E -0 2 | 1 .8 1 E -0 3 II | 7 .1 3 E -0 2 | 1 .8 6 E -0 3 II | 7 .3 3 E -0 2 | 1 .9 2 E -0 3 || | 7 .5 4 E -0 2 | 1 .9 7 E -0 3 II | 7 .7 5 E -0 2 | 2 .0 3 E -0 3 II | 7 .9 7 E -0 2 | 2 .0 8 E -0 3 || | 8 .2 0 E -0 2 | 2 .1 4 E -0 3 || | 8 .4 3 E -0 2 | 2 .2 0 E -0 3 II | 8 .6 7 E -0 2 | 2 .2 6 E -0 3 II | 8 .9 2 E -0 2 | 2 .3 3 E -0 3 || | 9 .1 7 E -0 2 | 2 .3 9 E -0 3 II | 9 .4 3 E -0 2 | 2 .4 6 E -0 3 || | 9 .7 0 E -0 2 | 2 .5 3 E -0 3 || | 9 .9 7 E -0 2 | 2 .6 0 E -0 3 || | 1 .0 3 E -0 1 | 2 .6 7 E -0 3 || | 1 .0 5 E -0 1 | 2 .7 4 E -0 3 || | 1 .0 8 E -0 1 | 2 .8 2 E -0 3 || | 2 .9 0 E -0 3 II | 1 .1 5 E -0 1 | 2 .9 8 E -0 3 || | 1 .1 8 E -0 1 | 3 .0 6 E -0 3 || | 1 .2 1 E -0 1 | 3 .1 4 E -0 3 II | 1 .2 5 E -0 1 | 3 .2 3 E -0 3 || | 1 .2 8 E -0 1 | 3 .3 2 E -0 3 || | 1 .3 2 E -0 1 | 3 .4 1 E -0 3 || | 1 .3 5 E -0 1 | 3 .5 1 E -0 3 II | 1 .3 9 E -0 1 | 3 .6 0 E -0 3 II | 1 .4 3 E -0 1 | 3 .7 0 E -0 3 || | 1 .4 7 E -0 1 | 3 .8 0 E -0 3 || | 1 .5 1 E -0 1 | 3 .9 0 E -0 3 || | 1 .5 5 E -0 1 | 4 .0 1 E -0 3 || 'bX -bax 3 o1 s cb fl fl&w bi O O O O O O O O O O O O O O O O O O O O O O O O o o O o O O O O O o O O O O w bi 00 w <N w kO w CO w w Ok w CO w00 wOsi w r~- wOsi w r~- w CO w00 w CO w Ok w n w w O w w O fl CO in in in kO kO l> r- 00 00 Ok Ok O O --- i (N <N CO w O w in w kO w00 kO w n l> w CO 00 w Ok w o\ Ok w r~o ww kO '-- i (N w CO CO w <N wOinsi w kO w O w 00 <N (N <N (N (N (N <N (N (N (N <N (N (N (N (N CO CO CO CO CO CO CO CO CO I kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO n n n n n n n n n n in n n n n n fl "O o O o o o o o o o o o o o o O O o O o o O O o o o o o o o o o o o o o o o o 'hi wf > w kO bsx CO n n 3 w ON kO n w n wkO5 w Ok n ko w CO CO ko w OinN ko w l> ko w Ok kO w (N R w CO CO R wn n R wo^: rR w O 00 wn (N 00 w<in> wkO5 r- w ON O 00 00 Ok wO k (N Ok wRin Ok wkO5 w ON 00 O wn o wo^: w w o '-- i w00 w '-- 1 (N wn <N wO k <N w ON CO wk O CO w wn wO k w in w0in0 Ok ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 CO CO CO CO CO CO CO CO CO CO CO CO CO "O bi o o o o o o o o o o o o o o o o o o o o o o o o o o O O o O o o O o o o o O flo s* b( w Ok CO O k w m o www Ok n O k (N CO in w kO kO ww CO O k 00 O www O n '-- i CO in w CO r- w CO Ok ww CO CO '-- 1 CO ww n in rr- w Ok O ww Osi kO (N w O w kO Ok www Osi 00 kO (N r - w Ok O w CO O w kO o w Ok O wwwwww <N kO O k Osi kO Ok '-- 1 '-- i (N (N CO w CO CO w rCO w n n n n n n kO k o k o k o k o k o R R R R 00 00 00 00 00 Ok O k O k fl "O- On) On) On) On) o r bi o o o o o OnI o OnI On) rsl rsl On) On) Osi On) rsl OnI On) Osi Osi Osi Osi Osi Osi Osi Osi On) Osi Osi Osi On) rsl Osi Osi Osi rsl rsl rsl On) o o o o o o Oo o o o o Oo o o OOOo Oo Oo o OOOOOOo fl fl ,--lCO s* bi win 00 w Ok w kO Ok w Osi O w00 w o wwww r- (N (N CO w00 wn w Osi in kO w Ok r- w00 w kO w r - 00 O k w CO o www (N O k (N CO w o wwwwww Ok Ok CO in kO O 00 Ok o 'f i O ,--l ,--l <N <N (N (N (N (N (N (N <N (N <N <N (N <N CO CO CO CO CO CO CO CO CO CO CO w kO w00 w o '-- i <N t J- w CO in w kO kO w Ok r- w CO Ok w00 w Osi o (N w rCO T T n n n On) Osi Osi On) Osi r s l r s l r s l n i On) On) OnI On) On) r s l Osi On) Osi r s l r s l r s l r s l OnI r s l r s l Osi r s l bi o O O o o o o o O o o o o o o o o O O O O O o O O o O o o o o o o o o o o o flW f l wwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwww to 'bi kO kO O 00 Ok 00 CO 00 o <N CO in Ok r- I*-00 o CN O (N 00 in I*-l> 1^ Ok r - 00 O k --- i CO kO CN 00 o r - kO (N n r - O r(N in CN 00 o o I*-- o CO kO o CO l> o o '-- 1 '-- 1 '-- 1 (N (N (N CO CO n n n n n n ko ko ko ko ko ko R R R R 00 00 00 00 Ok Ok Ok Ok I kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO kO flo f i o 'b i w o ON o o o o o o o o o o o o Oo o o o o o o o OOOOo o o o Oo o w CO w00 w www f > kO5 (N w00 w n w CN wO k w kO w CO w wwwwww o^: kC 5 n CO C N w ww fk O ww fk www C N CO w kC 5 w00 w w o Oo O wwww n r - r - 00 00 Ok o o '-- 1 (N CO CO n kO kO r - 00 Ok O '-- i (N CO in kO l> 00 O k o '-- 1 CO in r- 00 Ok 1 1 1 1bx 3 ' ' 1' ' ' (N (N (N (N (N (N <N (N (N (N (N (N <N <N CO CO CO CO CO CO CO CO CO CO TT n This document is a draftfor review purposes only and does not constitute Agency policy. C-199 DRAFT: DO NOT CITE OR QUOTE r r On] r r c r r r On) On) O n) On! r s l r s l r s l Osi Osi r s l Osi Osi Osi r s l (_, (_, (_, (_, (_, (_, (_, (_, (_, (_, (_, (_, (_, (_, (_, (_, (_, (_, (_, "3 b i o o o o o o o o o o o O o o o o O o O o o o o o o o o o o O o o o o o o o o o o o= * w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w b f r~- r r r~- c o o \ > n 0 0 m Osi 00 r - m CO CO rsl 00 o \ Osi n 'O 0 0 o \ Osi n r - 0 0 r o Osi ffl fl i> o <N CO i n o 0 0 o <N in r - CO i n o o o o o o '-- 1 CO F F F F F F'O 'O 0 0 0 0 0 0 0 0 0 0 0 0 cK cK cK cK cK n in in in in in fl r-> r-> r-> r-> r-> r-> r-> r-> r-> r-> r-> r-> r o r o r-> r o r o r o r-> r o r o r o r o r o r-> r-> r-> r o r o r o r o r o r-> r o r o r o r o r o 'b i o o o o O o o o o o 3+ ffl =3 w r-> w r~- w >n w rr www r-> 0 0 'O w ww CO = o o '-- 1 < N CO CO in o oooo oooo wwww wwww r-> o \ 0 0 0 0 O o r~- O 00 00 o '-- 1 (N CO oooo ooo wwww www 00 o \ ro CO o \ in ^ o 0 0 CTs O o o oooo ooo w wwww www o \ r~- n CO o\ O CO a \ o --- i <N CO CO oo ww n n ooo www CO Osi l> 0 0 CTs JO O c o c o c o c o CO CO CO CO CO CO CO CO CO n n n 'O 'O i> l> l> l> l> l> l> l> l> l> ffl 9 r-> r-> r-> r-> r-> r-> r-> r-> r-> r-> r-> r-> ro la i o o o o o o o o o o o o o ro o ro o ro o ro o ro o ro o ro o ro o o o o o o o o o o o o o o o o o o o+ ww www wwwww wwww wwww wwww wwww wwww wwww wwww fa o' b i i n o \ 0 0 CO o \ r-> r~- CO r-> o n CO r o o \ 0 0 O O r - 0 0 CO n 'O 0 0 <N n 'O 0 0 o \ Osi fl oi n i > 0 0 O '-- 1 CO i n 0 0 o '-- 1 CO n r - 0 0 o ( N 0 0 O o o (N CO n in in 'O 'O 'O 'O F F F F F F 0 0 0 0 0 0 0 0 0 0 0 0 cK cK cK cK cK 9 I "3 o o o o o o o o o o o o o o o o o o o o o o o o o o CO CO CO CO CO CO CO CO CO CO CO CO Oo OOOo o o Oo Oo 3 'hi w w 00 w rr w o w CO w o\ w n ww Osi r-> w 00 w kO w n w n w n w 'O w 00 w ro w CO w n w ro w n w w r- w w n w o\ w r- w o\ w w CO w 'O w 00 w r-> w Osi w w r- w o\ w i > 0 0 o '-- 1 CO 0 0 o '-- 1 CO n '-- 1 CO 0 0 o CO n 0 0 o (N CO CO CO i n i n i n i n i n 'O bi3 > n < n > n n n i n ' O ' O ' O ' O ' O ' O l> l> l> l> 0 0 0 0 0 0 0 0 c K c K c K c K ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 On) On) Osi Osi Osi Osi Osi r s l r s l Osi On) On) n i On) On) r s l On) On) On! On) On) Osi Osi Osi r s l r s l Osi r s l r s l n i r s l r s l r s l r s l Osi Osi r s l Osi "3 bi o o o O O o O O O o o o O O O O o o o o o O O O O O O o O o O O O O o O O o O bs*( w CO CO = 0 0 w Osi w oa \\ w o wwwww 00 CO '-- i '-- 1 (N CO w o\ w 'O ww o\ m ^o r - wn w 00 w m o wwww n --- i (N CO w in w wwwwwwww m n r - 00 ro CO r - 00 o (N CO in www 00 00 wwwww r~- ro 0 0 CO o <N CO <N (N (N CO CO CO CO CO CO CO CO CO CO CO CO n n in in n n n n 'O 'O 'O 'O 'O = r o r o r o r o r o r o r-> r o r-> r o r o r-> r o r o r o r o r o r o r o r o r o r o r o r o r o r o r o r o r o r o r o r-> r o r o r o r o r o r o 3 'b i o + o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o 1 1'Q CO = 3 = 3^ O ww 00 '-- 1 (N '' w wO w ro ww CO w o (N (N CO CO CO w CO wO ww Osi in in w0 \ w CO w 00 w Osi w w www n ro in l> r - 00 00 wo \ w a\ o ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' ' ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 (N wo \ w o (N (N w O <N (N wwww n ro (N <N CO CO (N (N (N (N w 00 (N w in (N w ro (N w (N w CO wo \ w l> 00 (N (N (N w CO (N CO r o r o r-> r-> r o r-> r o r-> r o r o r-> r o r o r o r o r o r-> r o r o r-> r-> r-> r-> r o r o r o r o r-> r o r-> r o r o r-> r o r o r-> r o r-> 9 o 3 - la i o + 'bi w ro o w00 o w oo ww o\ = r- r- 00 00 ooooooooooooo w CO O w o '-- 1 wO'-- i w n <N w Osi CO w CO w n w rin w r- w rr- w 00 w n w o oooooooooo wwwwwwwwww CO Osi Osi (N Osi Osi CO n '-- 1 (N CO in l> 00 o ooo wO'-- i w o\ (N w ooooo w CO n w n w00 r- ww CTs o oo wO'-- i w CO (N (N (N (N (N CO CO CO CO CO CO CO CO CO CO CO CO n n n n n n n n 'O 'O 'O I 9 "3 o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o '1'1' 1' 1'1' 1 1'1 13 "h i w w o\ w CO w00 w CO wo \ w www o \ n in in r- 00 00 o wwwwww r~o C'--O1 O (N 'O (N CO CO o wO w n w Osi w r-> w00 w n ww n n r - l> 00 o ww C'--O1 Osi (N w Osi CO w(N w Osi in w CO w r- w n 00 w r- w00 o ww CO (N CO w 'O w ro bi 3 ' ' (N (N (N (N (N (N (N (N (N (N (N (N (N (N CO CO CO CO CO CO CO CO CO CO | 1 .1 2 E + 0 0 | 2 .7 2 E -0 2 || | 1.15E+00 | | 1 .0 9 E + 0 0 | 2 .6 4 E -0 2 II | 1 .0 6 E + 0 0 | 2 .5 8 E -0 2 II | 1 .0 4 E + 0 0 | 2 .5 2 E -0 2 || | 2 .4 6 E -0 2 || | 9 .8 7 E -0 1 | 2 .4 0 E -0 2 || | 9 .6 3 E -0 1 |I 2 .3 4 E -0 2 II | 9 .3 9 E -0 1 | 2 .2 9 E -0 2 || | 9 .1 6 E -0 1 | 2 .2 4 E -0 2 II | 8 .9 3 E -0 1 | 2 .1 8 E -0 2 II | 8 .7 1 E -0 1 | 2 .1 3 E -0 2 || | 8 .4 9 E -0 1 | 2 .0 8 E -0 2 || | 8 .2 8 E -0 1 | 2 .0 3 E -0 2 II | 8 .0 7 E -0 1 | 1 .9 8 E -0 2 II | 7 .8 6 E -0 1 | 1 .9 3 E -0 2 || | 7 .6 6 E -0 1 | 1 .8 8 E -0 2 II | 7 .4 7 E -0 1 | 1 .8 4 E -0 2 II | 7 .2 8 E -0 1 | 1 .7 9 E -0 2 II | 7 .1 0 E -0 1 | 1 .7 5 E -0 2 II | 6 .9 2 E -0 1 | 1 .7 1 E -0 2 II | 6 .7 4 E -0 1 | 1 .6 7 E -0 2 || | 6 .5 7 E -0 1 | 1 .6 3 E -0 2 || | 6 .4 0 E -0 1 | 1 .5 9 E -0 2 || | 6 .2 4 E -0 1 | 1 .5 5 E -0 2 || | 6 .0 8 E -0 1 | 1 .5 1 E -0 2 II | 5 .9 2 E -0 1 | 1 .4 7 E -0 2 || | 5 .7 7 E -0 1 | 1 .4 4 E -0 2 || | 5 .6 2 E -0 1 | 1 .4 0 E -0 2 II | 5 .4 8 E -0 1 | 1 .3 7 E -0 2 II | 5 .3 4 E -0 1 | 1 .3 3 E -0 2 || | 5 .2 0 E -0 1 | 1 .3 0 E -0 2 || | 5 .0 6 E -0 1 | 1 .2 7 E -0 2 || | 1 .1 7 E -0 2 || | 1 .2 1 E -0 2 || | 1 .2 4 E -0 2 || ____(ng/kg)___ | 4 .4 4 E -0 1 | 1 .1 2 E -0 2 || | 4 .5 6 E -0 1 | 1 .1 4 E -0 2 || IBody Burden I1 Blood I 'bi <ONi -bai 3 Woi> F 4 .6 8 E -0 1 4 .8 1 E -0 1 4 .9 3 E -0 1 o o+ 3 = sp | | | 9 5 3 - = roo ro-> ro-> ro-> ro-> ro-> ro-> ro-> ro-> ro-> ro-> ro-> roo ro-> ro-> ro-> ro-> roo ro-> ro-> ro-> ro-> roo ro-> ro-> ro-> ro-> ro-> ro-> ro-> ro-> ro-> ro-> ro-> ro-> ro-> roo ro-> w+ Uo) w O w(N w ---ni ww 0'--01 (N w (N F(N wf > w CO CO CO F CO wO w F w n w n n w in w CO kO F kO w w n w l> w 00 w 00 00 w CO wo^: w COO woo^: www C'--O10'--01 C(NO w 0(N0 w CO wO w n w n w 00 in w CO kO FFFFFFFFFFFF I9 on "3 on on on on on on on on ion on on on on on on on on on on on on on on o o o o o o o o o o o o o o 3 "hi w bsi O n 3 ww k'--Oi CO n n wF n w kO n w 00 n wo^: w k'--Oi w CO n kO kO w in kO w CO rkO w CO kO w F w U) CO F w 00 in F wf > 00 w o F 00 ww 0(N0 CO n 00 00 wo^: r- w on wCN CO wf >w kO 00 00 Q\ o\ Q\ o\ wCN O w on woo^: ww '-- 1---ni w 0'--01 wC(NN w (Nn w (N w CO CO wF CO w w n w < > in ' 1' 1' 1' 1' 1' 1' 1' 1' 1' 1' 1' 1' 1' 1 This document is a draftfor review purposes only and does not constitute Agency policy. C-200 DRAFT: DO NOT CITE OR QUOTE bX Y oOoooOO i i i i i i i i i io O o o o o O =3 W bx w w w w w w w w w w w wwwwww C D > n I"- o \ <N w f l <N <N o O o o o 00 00 00 00 00 cK cK cK cK cK cK OOOOoOoOOOO oOOOoOoOooO wwwwwwwwwww vO 00 O <N r- o \ <N I*-- O --- 1 <N (N (N (N (N C D C D C D OoOoOoooo Ooooooooo wwwwwwwww C D n 00 I*-- o C D v-O Y - Y - n n n VO vO vO a DT DT <N DT DT <N D nI DT DT DT D nI DT DT 'bx o o o o o o o o o o o o o o o o O O o o O O O o O O O O O O O o o o o o O O 3+ ffl=3 wrr wn wo \ w >= in VO l> w00 o wCD w 00 w Y- wr->wv-OwCD wo <N CD in i> 00 O <N wr~CD w>n >n w Yl> wCD w O <N w 00 wr- wvO wvO wn wn w CD in r- --- iC D vO w 00 w O w CD wv-Owo \ w <N wn w 00 w w w o wo w CD wr~- OOo --- i'-- 1(N (N <N C D C D C D JO fOfl >n >n >n >n VO VO vO vO vO vO l> l> l> l> l> l> 00 00 00 00 00 cK cK cK cK ,--l,--l,-- l ,--l,--l,--l,--l,--l,--l,-- l ,-- l ,-- l ,-- l On! On) On) On) rr rr DnI DnI <N DT DnI <N DT DT DT DnI DT DT DnI DT DT DT <N DT DT 9 o IbX O + O O ooO O O oooooo o o o o o o o O o o O O O O o O o o O O O O O O '--1fa 'bi w cd w00 w cd w o\ win w w00 w >n w <N w o\ w o w >n wW Y - r-> wOn) w www r~- o \ wwwww C D n vO 00 wwww C D vO 00 w wwww vO o \ (N w O w r-> wDT w n w o\ fl 00 <N Y- vO i> <N Y- vO 00 o o O o o '-- 1 '-- i --- i '-- i '-- 1 (N (N (N (N C D C D C D C D in in n in l> l> 00 00 00 00 00 00 cK cK cK cK cK I rr On) On) On) On) On) On) On) On) On) O n) O n) O n) O n) O n) O n) O n) r r r r D T D T D T D nI D T D T D T D T D T <N D T D T <N D T D T D T D T <N D T 9 "3 o o o o o o o o o o o o o o o o o o o O o O o O O O O O O O O O O O o o O o 3 'b i w wwwww v-O O n) 00 Y - www 00 Y - wwwwwwwwwww o \ vO Y - On) r-> 00 r- v-O n Y- w w wwwww n v-O r- o \ wwwwww C D v-O o \ C D v-O ww n www v-O DT o \ G\ o o <N <N CD Y - Y- in vO l> 00 00 O --- i (N C D in vO r- 00 a \ '-- i <N C D vO r- CTs o (N C D in VO bx 3 ' 1 ' 1 <N <N <N <N <N <N <N <N <N <N <N <N (N (N CD CD CD C D CD CD CD CD CD CD Tj- Tj- Y- TT TT y - TT n n n n n "3 bi o o O s* b( w in w CD CO = O OOo o o o o o O wwwwwwwww (N o 0 0 r- n 00 00 o \ (N (N C D Y- in in VO 0 0 ooooooooo wwwwwwwww r- O n) o o \ O n) O n) C D C D C D o '-- 1 '-- i C D in vO r - oO ww n 00 o o OOo Oo OOo o o o o o o wwwwwwwwwwwwwwww DT C D n o \ C D v-O o \ v-O vO o \ C D v-O r-> v-O O '-- i ( N vO r - 0 0 '-- i '-- i ( N vO r - O C D TT t J- Tj- TT TT Y - TT Y - n n n n n n n n n n vO vO vO vO VO vO vO vO l > l > l > l > l > i > l > 0 0 = 3 'bx O + o OOo Oo o o o o o o o o o o o o o o o o Oo o Oo o o OOOOo o o o CO w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w n r-> v-O <N 0 0 o v-O o \ v-O r- o v-O C D O n) o 0 0 vO C D C D o \ n C D <N C D C D o\ v-O v-O 0 0 O n) o o\ = (N C D C D Y- in VO vO vO r- CTs o O '-- i <N C D C D n vO vO l > 0 0 (N C D Y- vO vO r- 0 0 o (N C D 3^ O D 4 D 4 (N (N <N D 4 (N (N <N D 4 <N D 4 C D C D C D C D C D C D C D C D C D C D C D C D C D tJ- Tj- Y- Y- Tj- TT Y- n n n n CO 9 o 'bX o + O 3 U. ww 'bi I*-- OO ww CN o Oo OOOo OO wwwwwwww 0 0 VO n C D I"- vO v-O o OOOOO wwwwww o \ v-O n I*-- r- Oo OOo o OOOOo o wwwwwwwwwwww I*-- vO 0 0 I*-- n v-O o \ CN n I*-- Oo o o o o o o wwwwwwww 0 0 C D v-O o \ vO O 00 = r- 0 0 o o '-- 1 (N C D C D vO r- l > 0 0 O --- i <N C D in vO r- 0 0 '-- i C D Y- in vO r- 0 0 o <N in VO CD CD CD Y- TT TT Tj- Y- Y- n n n n n n n n n n vO vO vO vO vO vO vO vO l > l > l > l > l > I C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D O n) O n) O n) dt dt dt dt <N <N DT DT DT DT O n) O n) O n) O n) O n) O n) O n) O n) 9 "3 O O O O o o O O O O O O O O O O O o o o o o o o o O O O O O o o o o o o o o 3 'bi w DT w O wwwwww 0 0 r- r~- 0 0 0 0 O w www CD 00 w w O www r- v-O wwwwwwwww O n) v-O o \ (N n o \ DT vO o w wwww 0 0 (N vO o ww n o w ww o\ w o w n O (N C D in r- '-- 1 n C D in O 0 0 '-- i C D vO oOo --- i '-- i <N (N C D C D C D in n VO vO vO r- 0 0 0 0 bx 3 vO vO vO vO vO vO l> l> l> 00 00 00 00 cK cK cK cK ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' ' ' ' ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 "3 bi o O o o o o o o o o O O O o O O O o O O O O O O o O o O o O o o o o o o o o O w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w wCO bs*( C D = vO rvO 00 VO o r- On) C D O r- m r- rr-- C0D0 00 o (N v'--Oi(N n (N o CD CD o\ CD 00 Y- CiDn 00 in CD VO o\ VO O o \ v-O On) 00 t> o o o (N v-O (N VO 00 in v-O C D VO r- o00 r0-0 o\ (N (N (N (N D 4 (N (N (N (N (N <N D 4 <N (N (N (N (N C D C D C D C D C D C D C D C D C D C D C D C D = o o o r-> r-> r-> r-> r-> r-> 3 'bx o + o o o o o o o o o O O O o O O O O O o o O O O O O O o O O O O o O o o O O w wi w w w w w w wi w1w1w' 1i w' i' w' i1w' 11w' 1w' 1w' 1w' 1w' ' w' 1w' 1w' 1w' 1w' 1w' 1w' w' 1w' 1w' 1w1w' 1w w w1w wi w1 CO 3^ = O o o 00 0\ '-- 00 00 <N 00 00 CD 00 O 00 ri-n 00 rVO 00 rr00 O '-- cK v-O 00 Oo '' '-- CD '-- vO 00 O r-> C D '-- '-- <N (N (N C D C D V.O o \ CD CD (N vO o \ o C D rVO VO VO O n o \ r- ' On) r-> n o o o '-- o \ On) '-- (N D 4 D 4 <N (N <N (N CO ^5 9 'bX o o o o o o o o o o o O O o o O O O o O o O O O o o O o o o o o O O O O O O (J w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w3 O) O) O) vO o O)'bi V.O 0 \ r-> n n r- o \ = r- ' i 1 ' i r-in in vo VO VO VO VO VO 00 n v-O O oO -- no \ C D 00 r- '-- (N <N C D C D vO n in vO v-O C D VO 00 00 00 CD o\ a\ n o-- CD no \ n o \ CD n VO VO (N (N (N (N <N (N (N (N (N (N (N D 4 D 4 (N (N (N <N (N D 4 C D C D C D C D C D C D C D C D C D IC D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D C D 9 "3 O O O O O O O O O o O O O O O O O O O O O O O O O O O O O O O O O O O O O O 3 'bi w ww <T\ w w U) ww <T\ o r w wwwwwww 0 0 n CD wwwwwwwww w w w w w w w w w U ) n CD CD CD CD o r n 00 n n ww CD vO vO r - r - r - r - 0 0 0 0 0 0 CD Y - n VO l > bx3 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ( N <N (N D 4 (N (N 00 (N (N (N o <N CD n VO (N CD CD CD CD CD CD CD CD Y - vO l > T T T j- 00 O n '-- i C D n n n n VO 0 0 n n n This document is a draftfor review purposes only and does not constitute Agency policy. C-201 DRAFT: DO NOT CITE OR QUOTE r-> r-> r-> O r-> r-> r-> r-> r-> r-> r-> r-> r-> ,_,,_,,_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_, 'bx o + o o o o o o o o o o o o o o o o o o o o o o O O o o o o o o o o o o o o o w ww = 00 VO co 00 w a\ wOn) w o\ CO ww r~vO 0 0 wOn) w wO o <N www o \ 00 o\ in i> a \ w o wv.O w 00 w wn w 00 wo wCO wn w 00 w wCO wv-O w o\ w(N wn w 00 O o '--i---i'--1(N (N (N <N CO CO CO CO w wn w 00 w n n in o wn w 00 wOn) wv.O o VO r- r- l> l> l> 0 0 0 0 0 0 0 0 cK cK cK cK cK cK a c o c o c o CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO 3 ffl 'bx =3 = oo + wv.Ow0 0 Oo o o o o o OOOOOOOOo Oo o OO ww 00 '-- 1'-- 1 w <N w <N w 00 <N w CO win CO w 00 CO w<N w vO wr-> w o\ in in wCO w 00 VO VO wOn)wOn)wr~-wOn)w0 0 r- 0 0 0 0 oO wCO w o\ Oo Oo OOO wn w ---i(N wO <N wCO CO wO oooo wVO wCO wro w 00 n o o Oo O wn wCO w r- 0 0 CTs o w o\ a\ O w 00 o O wO '--i JO O '1'1'1'1'1'1'1'1'1'1'1'1'1'1'1'1'''1'1'1' 1 ( N ( N o i < N ( N < N ( N < N < N < N ( N o i ( N ( N < N C O C O CO O n] O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) C O C O C O C O C O C O C O C O C O C O C O C O CO C O C O C O C O C O C O C O C O C O C O CO 9 o Isi o + o o o o o o o o o o o o o O O o o o o o o o o o O O o o o o o o o o o o O wwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwww f a 'bl r r o O n) O o CO 0 0 v.O n C O C O C O C O v O o ( N n o o O n) n O r-> C O n 0 0 O r- 00 fl c o VO '-- 1 < N vO l> '-- 1 C O n O O O '-- 1 --- i '-- 1 < N ( N ( N < N C O C O C O C O n in in o o o l> l> l> l> 00 00 00 00 00 00 cK cK cK cK i 4> 3w 2 fl=b3x >n o o 1 i 1 1 1 i 1 1 1 i 1 i i i i+ w o w <N w vO w w or o w <N w w vO w ww '--1 w vO ww '-- 1 w O W <r> o vO vO vO vO i> l> l> l> l> 00 00 00 cK cK cK oO oO ++ WW v.O o \ Oo o o + W oOooOoOo ooooooOo ++++++++ W HW vO o \ W WW i> WW 00 (N (N CO CO CO ooooo ooooo +++++ W WHWW r - v-O n n o o oo oo ++ WW <r> ooo ooo +++ WW w v.O 00 00 o \ ro ro ro r-> ro ro ro ro ro ro ro r-> r-> ro ro r-> ro r-> ro r-> r-> ro r-> r-> ro ro ro ro ro ro ro ro ro ro ro ro ro ro 'bi o + o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o wwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwww ro r- 00 n CO o \ O n CO (N r-> 00 r~- r- v.O v.O O O r- 00 o \ ro CO n O o \ (N n o \ 00 CO w = in in VO O r- 00 O o '--i <N CO n o o r- 00 O '--i <N CO in l> 00 o '-- i <N in 00 o CO n CO CO CO CO CO CO CO n n n n n n n n n o o O O o o o l> l> l> = O n! O n! O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n! O n! or O n) O n! O n! O n) O n) O n) O n) O n) O n) O n) O n) CO CO 3 'bi CO = o + w ro ooooooo wwwwwww v.O v.O v.O v.O o 00 o r- 00 O '--1(N oo ww O n) in VO oooooooo wwwwwwww v.O o \ O n) n o \ (N v.O r- 00 o ---i (N in o oooo wwww v.O CO o \ n 00 O '--i CO oooooo wwwwww CO ro 00 r~- VO o n l> 00 o (N oooooo wwwwww v.O v.O 00 ro O n) o \ o 00 o CO in a \ oo ww CO o \ <N OO ww ro CO oO 'Q o CO CO CO CO n n n n n n n VO o o o o o i> l> l> l> l> 00 00 00 00 cK cK ' 1' 1 CO O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n! O n) or O n) O n! O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) O n) 04 O n) O n) O n) O n) O n) O n) O n) 9 o Isi o + o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o 3 wwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwww 'bi o o ro 00 n O n) ro 00 n CO O n) ro 00 O n CO CO O n) O n) CO CO CO n v.O r- 00 ro (N ro CO v.O 00 = CO in o o r- 00 o '--1 <N CO CO n o r- 00 o '--1 <N CO n o i> 00 o (N in VO l> o '--1 CO CO CO CO CO CO CO CO CO CO n n n in n n n n n o O O o o o i> l> 9 I "3 o o O ooooooooooooooooooooooooooooO oooooo 3 'hi w ro wv.O w w r- w CO w o\ w n wOn) w 00 w n w (N w o\ ww r~- n wOn) w w o\ w 00 w r- w o w n w n w n wv.O wv.O w r- w o\ w ro w (N w n w r- w ro w wOn) w r- wOn) w w 00 00 CTs O o ---i(N <N CO in o r- 00 00 o '--1<N CO in o r- 00 o (N CO in o o <N in o bx 3 ' 1' 1' '' 1(N (N <N (N (N <N (N (N oi <N (N <N (N (N CO CO CO CO CO CO CO CO CO n n n n ro ro ro ro ro ro ro ro ro ro r-> r-> ro ro r-> r-> ro ro ro ro ro ro ro ro ro ro ro ro r-> ro ro ro r-> r-> r-> ro ro r-> 'bX o + o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o wwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwww CT\ On) n o \ On) v.O o \ CO v.O ro n o \ CO o n o \ o \ CO 00 On) r- On) r- On) 00 CO o \ n ro v.O (N 00 ro r- CO w = O l> l> 00 00 00 o o o '-- 1 '-- 1 (N (N (N CO CO in in o o r- r- 00 00 oO '-- 1 (N CO CO oi oi (N (N <N (N oi (N (N (N (N (N (N oi (N oi (N oi <N (N oi CO CO CO CO CO CO CO CO = On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) or On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) 3 'bx o + o o o o o o o o o o o o o O o O o o o o o o o o o o o o o o o o o o o o o CO w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w o CO o ro 00 On) v.O r-> n o \ o \ CO 00 CO 00 O n v.O (N o \ n On) o \ v.O CO o \ v.O On) ro o \ 00 r- = in in in o o r- l> l> 00 00 Oo (N <N CO CO n o o r- 00 CTs a \ O (N CO CO in O CO ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 (N (N (N (N oi (N (N <N (N <N <N <N oi <N (N <N CO CO CO CO CO CO CO On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On! On) On) On) On) On) On) On) On) On) On) On) On) On! 9 o 3 'bi o + o 'bi w On) w =o o ooo www r- ro o r- oooooooooooooOooo wwwwwwwwwwwwwwwww r- ro r- r - 00 00 00 CTs n o \ a\ CO O ro '-- i n --- i o \ CO '-- 1 (N 00 (N On) v.O O CO CO oooooooooooooooo wwwwwwww n ro n ro n r o n w r- ww On) r - w CO wo \ w n w w o in n o o r - l> 00 00 O o --- i (N (N (N o i (N o i <N <N (N (N <N (N o i o i o i o i <N o i o i (N <N (N <N CO CO CO CO CO I On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On) On! 9 "3 o o o o o o o o o o o o o o o o o O o O O O o o o o O o o O O O O o o o O o 3 'h i w v.O w CO w On) w o bsx 3 00 n O o (N oo wo \ wo \ w ro w ww (N w r- w ww n ro www v.O CO O ww 00 r - w CO w v.O wo \ w (N w O wo \ w CO w O ww ro wwww 00 (N VO wwww n r-> n r o w n in t> o (N VO 00 '-- i CO o 00 '-- 1 VO OOo --- i '-- i (N <N CO CO CO n n o o r- l> ,--lo o (> l> l> i> (> 00 00 00 00 cK cK cK cK ,-- l ,-- l ,-- l ,-- l ,-- l ,-- l ,-- l ,-- l ,-- l ,-- l ,-- l ,-- l ,-- l ,-- l ,-- l ,-- l ,-- l ,-- l This document is a draftfor review purposes only and does not constitute Agency policy. C-202 DRAFT: DO NOT CITE OR QUOTE On) On) On) On) On) On) On) On) On) On) i *. . o o O O o o o o o o o o o o o o o o o o o o o o o o O o o o o ^ yf + o =ffsl w w rr CM ww On) <N CO ww CO >n VO ww >n O w00 wo w CO i> 00 '-- 1 <N ww O o\ CO www On) v-O VO l> CTs www n O o (N CO w vO w00 w ro w n ww On) o \ CO o 00 '--1CO w vO vO ww CO ro 00 o www o CO ro '--1 CO in w r~vO <n >n >n n n n n vO vO vO VO 00 cK oi (N <N oi (N fl qq ^ o s CO CO CO oOoooooooooooooooooooooo + www r-> r~in i> O w CO wO OO wwwww O '-- i CO '-- 1 V'--Oi Q\ '--i CO <N w vO <N w ro CO w CO w00 w <N CO w vO w ro in w in w in w o wwww CO On) On) n o in O cK cK .-- i .-- i .--i .-- i .-- 1 .--i .-- i .-- i .-- i .-- i .-- 1 .--i .-- i .-- i .-- i .--i .--i oi <N CO CO ooooooo w ro w00 w O ww v-O ww CO in n n VO vO l> CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO 9 o ooooooOooO + wOn) w <N wOn) wOn) w <N wCO w w<n wr~- w o\ o <N CO <n VO i> 00 a \ OoOoOooooo w w w r- wro wCO w r- wOn) wn w r- w '--i(N CO in VO r- o o o ooooooooooO wn CO ww o VO '--1(N w (N w00 in w i> wro w ro wCO <N w o\ CO w n <n n n n n n n vO vO l> cK oi (N <N oi & r-> r-> r-> r-> r-> r-> r-> r-> r-> r-> ro ro r-> ro ro ro r-> U o o o o o o o o o o o o o o o o o o o o o o o O o o o o O O o 7 S?5 '--1 6C 5 wr~o win <N w wCO wCO w O 00 o w>n wO <N w o\ VO wOn) wv'--OiwO wv-O vO wOn) w00 w vO w o wro o wr-> w o CO wrvO wro o wCO CO wrvO wro o wCO CO wrvO wro o wCO CO wrvO wro o 0 0 0 1 1 1 1 VO VO i> l> l> l> l> 00 00 00 00 cK cK cK oi (N <N CO CO CO n (_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_,(_, ooOoOOOOOoooOOoooO O ooooooooO oooooooooo + wwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwww CTs CO r- v-O ro o \ CO 00 CO 00 CO 00 o \ n ro v.O O n) o \ n 00 n <N o \ O (N ro 00 v.O n CO O n) (N W= 00 00 CTs o o o '-- 1 '-- 1 (N (N CO CO n VO o r- t > 00 CTs o '-- 1 (N CO in in o r- 00 o oi oi (N (N <N (N (N (N (N (N (N oi oi (N (N (N oi (N (N CO CO CO CO CO CO CO CO CO CO CO CO = CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO 3 'si O + o O O O o o O O O O O o O O O O o O O O O O O O O O O o O O O O O o o o O CO w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w w n n n r- 00 ro (N n 00 n o \ o\ ro v.O CO O 00 v.O n n in n v.O 00 o On) r- CO = <N CO in vO i> 00 o '-- 1 CO n o 00 o <N CO in r- 00 O (N CO in l> --- i CO in r- O (N r- <N O CO CO CO CO CO CO CO CO CO n n n n n n o VO o o o o l> l> l> i> 00 00 00 00 00 cK CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO 9 5 Isi o + o O o O O O O o o o O o o o o o o o o o o o o o o o o o o o o o o o o o o w 'bi = ww in r- w wwww CO n 00 00 G \ a\ o ww w wwwwwwwwwwwww o^: CO CN ts or 00 CO 00 f > V.C5 CN o^: o^: o '-- 1 '-- i (N (N CO CO in o O r- r- 00 G \ o wwwwwwwwwww (N VO (N f > 00 V.C5 n CO '-- 1 (N CO in in VO l> 00 oi (N (N <N (N (N <N (N (N (N oi <N <N oi oi (N (N (N CO CO CO CO CO CO CO CO CO CO CO CO ui 3 ooOOooooooOOOOooOooOOooOOoOOOo ooooooooooooOoooooooOoooooOoOo wwwwwwwwwwwwwwwwwwwwwwwww r- CO o \ vO On) o \ v-O CO r-> 00 m CO On) ro o \ 00 o v.O v.O o o o '-- i (N (N CO in in VO l> 00 a \ o '-- 1 (N CO rin ro 00 r- ro o 1(N (N (N (N (N (N (N oi <N oi (N (N oi oi CO CO CO CO CO CO CO CO CO wwwww CO v.O o \ On) m '-- 1 (N CO in o OO OO ww o\ l> ooooOO oooooO wwwwww 00 o \ v.O On) o \ o (N CO in O 00 n n n n n n This document is a draftfor review purposes only and does not constitute Agency policy. C-203 DRAFT: DO NOT CITE OR QUOTE 1 C .5 . R E F E R E N C E S 2 A m in, S; M oore, R W ; P eterson, RE; et al. (2000) G estational an d lactational exposure to TC D D o r coplanar P C B s 3 alters adult expression o f saccharin preference behavior in fem ale rats. N eurotoxicol Teratol 22(5):675-682. 4 Aylward, LL; Brunet, RC; Carrier, G; et al. (2005a) C oncentration-dependent TCDD elim ination kinetics in 5 humans: toxicokinetic modeling for m oderately to highly exposed adults from Seveso, Italy, and Vienna, Austria, 6 and impact on dose estimates for the NIOSH cohort. J Exp Anal Environ Epidem iol 15:51-65. 7 Aylward, LL; Brunet, RC; Starr, TB; et al. (2005b) Exposure reconstruction for the TCD D -exposed N IO SH cohort 8 using a concentration- and age-dependent model o f elimination. Risk Anal 25(4):945-956. 9 Aylward, LL; Bodner, KM ; Collins, JJ; et al. (2009) TCDD exposure estim ation for w orkers at a New Zealand 10 2,4,5-T m anufacturing facility based on serum sampling data. J Expo Sci Environ Epidemiol. 3 June; 11 doi:10.1038/jes.2009.31. Available online at http://www.nature.com/jes/journal/vaop/ncurrent/full/jes200931a.htm l. 1 2 B ell, D R ; Clode, S; Fan, M Q ; et al. (2007) Toxicity o f 2,3,7,8-tetrachlorodibenzo-p-dioxin in the developing m ale 13 W istar(H an) rat. I: N o decrease in epididym al sperm count after a single acute dose. T oxicol Sci (1):214-23. 14 Bohonowych, JE; Denison, MS. (2007) Persistent binding o f ligands to the aryl hydrocarbon receptor. Toxicological 15 Sciences 98(1):99-109. 16 Boverhof, DR; Burgoon, LD; Tashiro, C; et al. (2005) Tem poral and dose-dependent hepatic gene expression 17 patterns in mice provide new insights into TCDD-mediated hepatotoxicity. Toxicol Sci 85(2): 1048-1063. 18 Cantoni, L; Salmona, M; Rizzardini, M. (1981) Porphyrogenic effect o f chronic treatm ent w ith 2,3,7,8 19 tetrachlorodibenzo-p-dioxin in female rats. D ose-effect relationship following urinary excretion o f porphyrins. 2 0 Toxicol Appl Pharmacol 57:156-163. 2 1 Carrier, G; B runet, RC; B rodeur, J. (1995a) M odeling o f the toxicokinetics o f polychlorinated dibenzo-p-dioxins 2 2 and dibenzofuranes in mammalians, including humans: II kinetics of absorption and disposition of PCDDs/PCDFs. 2 3 Toxicol Appl Pharm acol 131(267):276. 2 4 Carrier, G; Brunet, RC; Brodeur, J. (1995b) M odeling o f the toxicokinetics o f polychlorinated dibenzo-p-dioxins 2 5 and dibenzofurans in mammalians, including humans. Toxicol Appl Pharmacol 131:253-266. 2 6 Chu, I; Valli, VE; R ousseaux, CG. (2007) Com bined effects o f 2,3,7,8-tetrachlorodibenzo-pdioxin and 2 7 polychlorinated biphenyl congeners in rats. Toxicol Environ Chem 89(1):71-87. 2 8 Connor, KT; Aylward, LL. (2006) H um an response to dioxin: aryl hydrocarbon receptor (AHR) m olecular structure, 2 9 function, and dose-response data for enzyme induction indicate an impaired hum an AhR. J Toxicol Environ H ealth 3 0 Part B: Critical Reviews 9(2):147-171. 31 Crofton, KM ; Craft, ES; Hedge, JM ; et al. (2005) Thyroid-horm one-disrupting chemicals: evidence for dose3 2 dependent additivity or synergism. Environ H ealth Perspect 113(11): 1549-1554. 33 Derelanko, MJ; Hollinger, MA. (1995) CRC Handbook of Toxicology, pp.1-948. New York, NY. 3 4 Diliberto, JJ; Burgin, D; Birnbaum, LS. (1997) Role o f CYP1A2 in hepatic sequestration o f dioxin: studies using 3 5 CYP1A2 knock-out mice. Biochem Biophys Res Com m un 236(2):431-433. 3 6 Emond, C; Birnbaum, LS; DeVito, M. (2004) Physiologically based pharmacokinetic model for developmental 3 7 exposures to TCDD in the rat. Toxicol Sci 80(1): 115-133. 3 8 Emond, C; M ichalek, JE; Birnbaum , LS; et al. (2005) Com parison o f the use o f a physiologically based 3 9 pharmacokinetic model and a classical pharmacokinetic model for dioxin exposure assessments. Environ Health 4 0 Perspect 113(12):1666-1668. This document is a draftfor review purposes only and does not constitute Agency policy. C-204 DRAFT--DO NOT CITE OR QUOTE 1 Emond, C; Birnbaum, LS; Devito, MJ. (2006) Use o f a physiologically based pharmacokinetic m odel for rats to 2 study the influence o f body fat mass and induction o f CYP1A2 on the pharmacokinetics o f TCDD. Environ H ealth 3 Perspect 114(9):1394-1400. 4 Fattore, EE; Trossvik, C; Hakansson, H. (2000) Relative potency values derived from hepatic vitam in A reduction in 5 male and female Sprague-D aw ley rats following subchronic dietary exposure to individual polychlorinated dibenzo6 p-dioxin and dibenzofuran congeners and a mixture thereof. Toxicol Appl Pharm acol 165(3):184-194. 7 Haddad, S; Beliveau, M; Tardif, R; Krishnan, K. (2001) A PBPK m odeling-based approach to account for 8 interactions in the health risk assessm ent of chemical mixtures. Toxicol Sci 63:125-131. 9 H assoun, EA; W ilt, SC; Devito, M J; et al. (1998) Induction o f oxidative stress in brain tissues o f mice after 10 subchronic exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Sci 42:23-27. 11 H assoun, EA; Li, F; Abushaban, A; et al. (2000) The relative abilities o f TCDD and its congeners to induce 12 oxidative stress in the hepatic and brain tissues o f rats after subchronic exposure. Toxicology 145:103-113. 13 Heinzl, H; M ittlback, M; Edler, L. (2007) O n the translation o f uncertainty from toxicokinetic to toxicodynam ic 14 models - the TCDD example. Chemosphere 67(9):S365-S374. 15 H ojo, R; Stern, S; Zareba, G; et al. (2002) Sexually dim orphic b ehavioral responses to prenatal d ioxin exposure. 16 Environ Health Perspect 110:247-254. 17 Huh, C; Bloch, WE. (2003) A review o f U.S. anthropom etric reference data (1971-2000) w ith com parisons to both 18 stylized and tomographic anatomic models. Physics in M edicine and Biology 48(20):3411-3429. 19 Ikeda, M ; M itsui, T; Setani, K; et al. (2005) In utero and lactational exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin 2 0 in rats disrupts brain sexual differentiation. Toxicol Appl Pharmacol 205(1):98-105. 21 ILSI (International Life Sciences Institute). (1994) Physiological param eter values for PBPK models. W ashington, 2 2 DC: Risk Science Institute. 23 Irigaray, P; M ejean, L; Laurent, F. (2005) Behaviour o f dioxin in pig adipocytes. Food Chem Toxicol 2 4 43(3):457-460. 2 5 Kattainen, H; Tuukkanen, J; Simanainen, U; et al. (2001) In utero/lactational 2,3,7,8-tetrachlorodibenzo-p-dioxin 2 6 exposure impairs m olar tooth developm ent in rats. Toxicol Appl Pharm acol 174(3):216-224. 2 7 Keller, JM; Huet-Hudson, YM; Leamy, LJ. (2007) Qualitative effects o f dioxin on molars vary among inbred mouse 2 8 strains. Arch Oral Biol 52(5):450-454. 2 9 Kerger, BD; Leung, HW ; Scott, P; et al. (2006) Age- and concentration-dependent elim ination half-life o f 2,3,7,8 3 0 tetrachlorodibenzo-p-dioxin in Seveso children. Environ H ealth Perspect 114(10):1596-1602. 31 Kerger, BD; Leung, HW ; Scott, PK; et al. (2007) Refinem ents on the age-dependent half-life m odel for estim ating 3 2 child body burdens o f polychlorodibenzodioxins and dibenzofurans. Chemosphere 67(9):S272-S278. 33 Kim, AH; Kohn, M C; Nyska, A; et al. (2003) A rea under the curve as a dose m etric for prom otional responses 3 4 following 2,3,7,8-tetrachlorodibenzo-p-dioxin exposure. Toxicol Appl Pharm acol 191(1):12-21. 35 Kitchin, KT; W oods, JS. (1979) 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) effects on hepatic microsomal 3 6 cytochrome P-448-m ediated enzyme activities. Toxicol Appl Pharmacol 47:537-546. 3 7 Kociba, RJ; Keeler, PA; Park, GN; et al. (1976) 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD): results o f a 13-week 3 8 oral toxicity study in rats. Toxicol Appl Pharm acol 35:553-574. 3 9 Kociba, RJ; Keyes, DG; Beyer, JE; et al. (1978) Results o f a tw o-year chronic toxicity and oncogenicity study o f 4 0 2,3,7,8-tetrachlorodibenzo-p-dioxin in rats. Toxicol Appl Pharm acol 46(2):279-303. This document is a draftfor review purposes only and does not constitute Agency policy. C-205 DRAFT--DO NOT CITE OR QUOTE 1 K ociba, RJ; Keyes, DG; Beyer, JE; et al. (1979) Long-term toxicological studies o f 2,3,7,8-tetrachlorodibenzo-p2 dioxin (TCDD) in laboratory animals. A nn N Y Acad Sci 320: 397-404. 3 Korenaga, T; Fukusato, T; Ohta, M ; et al. (2007) Long-term effects o f subcutaneously injected 2,3,7,8 4 tetrachlorodibenzo-p-dioxin on the liver o f rhesus monkeys. Chemosphere 67(9):S399-S404. 5 Korkalainen, M; Tuomisto, J; Pohjanvirta, R. (2004) Primary structure and inducibility by 2,3,7,8 6 tetrachlorodibenzo-p-dioxin (TCDD) of aryl hydrocarbon receptor repressor in a TCDD-sensitive and a TCDD7 resistant rat strain. Biochem Biophys Res Com m un 315(1):123-131. 8 Kransler, KM; M cGarrigle, BP; Olson, JR. (2007) Comparative developmental toxicity o f 2,3,7,8 9 tetrachlorodibenzo-p-dioxin in the hamster, rat, and guinea pig. Toxicology 229(3):214-225. 10 Krishnan, D. (2007) Neurobehavioral and neuroendocrine assessm ent o f rats perinatally exposed to polychlorinated 11 biphenyls: a possible model for autism. Bow ling Green State University. 12 Krishnan, K. (2008) Physiologically based pharm acokinetic m odelling in toxicology. In: Hayes, AW ; ed. Principles 13 and methods o f toxicology. 5th ed. New York, NY: CRC Press, pp. 231-291. 14 Latchoumycandane, C; M athur, PP. (2002) Effects of vitam in E on reactive oxygen species-mediated 2,3,7,8 15 tetrachlorodi-benzo-p-dioxin toxicity in rat testis. J Appl Toxicol 22(5):345-351 16 Latchoumycandane, C; Chitra, KC; M athur, PP. (2002) The effect o f 2,3,7,8-tetrachlorodibenzo-p-dioxin on the 17 antioxidant system in m itochondrial and microsomal fractions o f rat testis. Toxicology 171(2-3): 127-135. 18 Leung, HW ; Poland, A; Paustenbach, D; et al. (1990) Pharm acokinetics o f (125-I)-2-lodo-3,7,8-trichlorodibenzo-p19 dioxin in mice: analysis w ith a physiological modeling approach. Toxicol Appl Pharmacol 103:411-419. 2 0 Li, B; Liu, H; Dai, L; et al. (2006) The early embryo loss caused by 2,3,7,8-tetrachlorodibenzo-p-dioxin may be 21 related to the accum ulation o f this com pound in the uterus. Reprod Toxicol 21(3):301-306. 2 2 Luecke, RH; Pearce, BA; W osilait, W D; et al. (2007) Postnatal grow th considerations for PBPK modeling. J 23 Toxicol Environ Health A 70(12):1027-37. 2 4 M arkowski, VP; Zareba, G; Stern, S; et al. (2001) A ltered operant responding for m otor reinforcem ent and the 2 5 determ ination o f benchm ark doses following perinatal exposure to low-level 2,3,7,8-tetrachlorodibenzo-p-dioxin. 2 6 Environ Health Perspect 109(6):621-627. 2 7 M aruyama, W; Aoki, Y. (2006) Estim ated cancer risk o f dioxins to hum ans using a bioassay and physiologically 2 8 based pharmacokinetic model. Toxicol Appl Pharmacol 214(2):188-198. 2 9 M aruyama, W; Yoshida, K; Tanaka, T; et al. (2002) Determ ination o f tissue-blood partition coefficients for a 3 0 physiological model for humans, and estim ation of dioxin concentration in tissues. Chemosphere 46(7):975-985. 31 M aruyama, W; Yoshida, K; Tanaka, T; et al. (2003) Sim ulation o f dioxin accum ulation in hum an tissues and 3 2 analysis o f reproductive risk. Chemosphere 53(4):301-313. 3 3 M iettinen, H M ; Sovari, R; A laluusua, S; et al. (2006) T he effect o f perinatal T C D D exposure o n caries susceptibility 3 4 in rats. Toxicol Sci 91(2):568-575. 3 5 M ilbrath, M O; W enger, Y; Chang, CW; et al. (2009) A pparent half-lives o f dioxins, furans, and polychlorinated 3 6 biphenyls as a function o f age, body fat, sm oking status, and breast-feeding. Environ H ealth Perspect 3 7 -117(3):417425. 3 8 M oser, GA; M cLachlan, MS. (2002) M odeling digestive tract absorption and desorption o f lipophilic organic 3 9 contaminants in humans. Environ Sci Technol 36(15):3318-25. 4 0 M ullerova, D; K opecky, J. (2007) W hite adipose tissue: storage and effector site fo r environm ental pollutants. 41 Physiol Res 56(4):375-381. This document is a draftfor review purposes only and does not constitute Agency policy. C -206 D R A FT -- D O N O T C IT E O R Q U O T E 1 M urray, FJ; Smith, FA; Nitschke, KD; et al. (1979) Three-generation reproduction study o f rats given 2,3,7,8 2 tetrachlorodibenzo-p-dioxin (TCDD) in the diet. Toxicol Appl Pharm acol 50:241-252. 3 M urray, TJ; Yang, X; Sherr, D. (2006) G row th o f a hum an m am m ary tum or cell line is blocked by galangi, a 4 naturally occurring bioflavonoid, and is accom panied by down-regulation o f cyclins D3, E, and A. Breast Cancer 5 Res 8:R17 (doi:10.1186/bcr1391) 6 Nadal, M ; Perello, G; Schuhmacher, M ; et al. (2008) Concentrations o f PCD D /PCD Fs in plasm a o f subjects living 7 in the vicinity o f a hazardous waste incinerator: Follow-up and modeling validation. Chemosphere 73(6):901-906. 8 Nadal, M; Dom ingo, JL; Garcia, F; et al. (2009) Levels o f PCD D /F in adipose tissue on non-occupationally exposed 9 subjects living near a hazardous waste incinerator in Catalonia, Spain. Chemosphere 74(11):1471-1476. 10 NAS (National Academy o f Sciences). (2006) H ealth risks from dioxin and related compounds: evaluation o f the 11 EPA reassessment. W ashington, DC: National Academies Press. Available online at 12 http://w w w .nap.edu/catalog.php7record_idM 1688. 13 N ohara, K; Fujim aki, H; Tsukum o, S; et al. (2000) T he effects o f perinatal exposure to low doses o f 2 ,3,7,8 14 tetrachlorodibenxo-p-dioxin on immune organs o f rats. Toxicology 154(1-3):123-133 15 Nohara, K; Ao, K; M iyam oto, Y; et al. (2006) Com parison o f the 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)16 induced CYP1A1 gene expression profile in lymphocytes from mice, rats, and humans: M ost potent induction in 17 humans. Toxicology 225(2-3):204-213. 18 NTP (National Toxicology Program). (1982) Bioassay o f 2,3,7,8-tetrachlorodibenzo-p-dioxin for possible 19 carcinogenicity (gavage study). Tech. Rept. Ser. No. 201. U.S. D epartm ent o f H ealth and H um an Services, Public 2 0 H ealth Service, Research Triangle Park, NC. 21 NTP (National Toxicology Program). (2006a) NTP technical report on the toxicology and carcinogenesis studies of 2 2 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (CAS No. 1746-01-6) in female H arlan Sprague-Dawley rats (Gavage 2 3 Studies). Natl Toxicol Program Tech Rep 521. Public H ealth Service, National Institute o f Health, U.S. Departm ent 2 4 o f H ealth and Hum an Services, Research Triangle Park, NC. 2 5 O 'Flaherty, EJ. (1992) M odeling bone m ineral m etabolism , w ith special reference to calcium and lead. 2 6 Neurotoxicity 13(4):789-797. 2 7 O hsako, S; M iyabara, Y; N ishim ura, N; et al. (2001) M aternal exposure to a low dose o f 2,3,7,8-tetrachlorodibenzo2 8 p-dioxin (TCDD) suppressed the developm ent of reproductive organs o f male rats: dose-dependent increase of 2 9 mRNA levels o f 5alpha-reductase type 2 in contrast to decrease o f androgen receptor in the pubertal ventral prostate. 3 0 Toxicol Sci 60(1):132-43. 31 Olsman, H; Engwall, M; Kamm ann, U; et al. (2007) Relative differences in aryl hydrocarbon receptor-m ediated 3 2 response fo r 18 polybrom inated and m ixed halogenated dibenzo-p-dioxins and -furans in cell lines fro m fo u r 33 different species. Environ Toxicol Chem 26(11):2448-2454. 3 4 Pelekis, M; Gephart, LA; Lerman, SE. (2001) Physiological-model-based derivation of the adult and child 35 pharmacokinetic intraspecies uncertainty factors for volatile organic compounds. Reg Toxicol Pharmacol 3 6 33(1):12-20. 3 7 Poulin, P; Theil, FP. (2000) A priori prediction o f tissue:plasm a partition coefficients o f drugs to facilitate the use of 3 8 physiologically-based pharmacokinetic models in drug discovery. J Pharm Sci 89:16-35. 3 9 Saghir, SA; Lebofsky, M ; Pinson, D M ; et al. (2005) V alidation o f H a b er's R ule (dose X tim e = constant) in rats and 4 0 mice for m onochloroacetic acid and 2,3,7,8-tetrachlorodibenzo-p-dioxin under conditions of kinetic steady state. 41 Toxicology 215(1-2):48-56. This document is a draftfor review purposes only and does not constitute Agency policy. C-207 DRAFT--DO NOT CITE OR QUOTE 1 Santostefano, M J; W ang, X; Richardson, VM ; et al. (1998) A pharm acodynam ic analysis o f TCD D -induced 2 cytochrome P450 gene expression in m ultiple tissues: dose- and tim e-dependent effects. Toxicol Appl Pharm acol 3 151:294-310. 4 Schantz, SL; Seo, BW ; M oshtaghian, J; et al. (1996) Effects o f gestational and lactational exposure to TCDD or 5 coplanar PCBs on spatial learning.Neurotoxicol Teratol 18(3):305-313. 6 Schecter, A; Pavuk, M; Ppke, O; et al. (2003) Dioxin, dibenzofuran, and coplanar PCB levels in Laotian blood and 7 milk from agent orange-sprayed and nonsprayed areas, 2001. J Toxicol Environ Health Part A: Current Issues 8 66(21):2067-2075. 9 Sewall, CH; Flagler,N; V anden Heuvel, JP; et al. (1995) Alterations in thyroid function in fem ale Sprague-Dawley 10 rats following chronic treatment w ith 2,3,7,8-tetrachlorodibenzo-p-dioxin. Toxicol Appl Pharm acol 132:237-244. 11 Shi, Z; Valdez, K; Ting, A; et al. (2007) O varian endocrine disruption underlies prem ature reproductive senescence 12 following environmentally relevant chronic exposure to the aryl hydrocarbon receptor agonist 2,3,7,8 13 tetrachlorodibenzo-p-dioxin. Biol Reprod 76(2):198-202. 14 Smialowicz, RJ; D eVito, M J; W illiams, W C; et al. (2008) Relative potency based on hepatic enzyme induction 15 predicts immunosuppressive effects o f a mixture o f PCDDS/PCDFS and PCBS. Toxicol Appl Pharmacol 16 227(3):477-484. 17 Staskal, DF; D iliberto, JJ; D evito, M J; et al. (2005) Inhibition o f hum an and rat CYP1A2 by TCDD and dioxin-like 18 chemicals. Toxicol Sci 84(2):225-231. 1 9 Toth, K ; Som fai-R elle, S; Sugar, J; et al. (1979) C arcinogenicity testin g o f herbicide 2,4,5-trichlorophenoxyethanol 2 0 containing dioxin and o f pure dioxin in Swiss mice. Nature 278:548-549. 21 Toyoshiba, H; W alker, NJ; Bailer, AJ; et al. (2004) Evaluation o f toxic equivalency factors for induction o f 2 2 cytochromes P450 CYP1A1 and CYP1A2 enzyme activity by dioxin-like compounds. Toxicol Appl Pharm acol 23 194(2):156-168. 2 4 U.S. EPA (Environmental Protection Agency). (2003) Exposure and hum an health reassessm ent o f 2,3,7,8 2 5 tetrachlorodibenzo-p-dioxin (TCDD) and related com pounds [NAS review draft]. Volumes 1-3. National Center 2 6 fo r E nvironm ental A ssessm ent, W ashington, D C ; E PA /600/P-00/001 Cb, V olum e 1. A vailable online at 2 7 http://www.epa.gov/nceawww1/pdfs/dioxin/nas-review/. 2 8 V an Birgelen, AP; V an der K olk, J; Fase, KM ; et al. (1995) Subchronic dose-response study o f 2,3,7,8 2 9 tetrachlorodibenzo-p-dioxin in female Sprague-Dawley rats. Toxicol Appl Pharm acol 132:1-13. 3 0 V anden Heuvel, JP; Clark, GC; Tritscher, A; et al. (1994) A ccum ulation o f polychlorinated dibenzo-p-dioxins and 31 dibenzofurans in liver o f control laboratory rats. Fundam Appl Toxicol 23:465-469. 3 2 W ang, X; Santostefano, M J; Evans, M V; et al. (1997) D eterm ination o f param eters responsible for pharm acokinetic 33 behavior o f TCDD in female Sprague-Dawley Rats. Toxicol Appl Pharm acol 147:151-168. 3 4 W ang, X; Santostefano, M J; Devito, M J; et al. (2000) Extrapolation o f a PBPK m odel for dioxins across dosage 3 5 regimen, gender, strain, and species. Toxicol Sci 56(1):49-60. 3 6 W hite, KL, Jr; Lysy, HH; M cCay, JA; et al. (1986) M odulation o f serum com plem ent levels follow ing exposure to 3 7 polychlorinated dibenzo-p-dioxins. Toxicol Appl Pharmacol 84:209-219. 3 8 W ilkes, JG; Hass, BS; Buzatu, DA; et al. (2008) M odeling and assaying dioxin-like biological effects for both 3 9 dioxin-like and certain non-dioxin-like compounds. Toxicol Sci 102(1):187-195. This document is a draftfor review purposes only and does not constitute Agency policy. C-208 DRAFT--DO NOT CITE OR QUOTE [T his p ag e in ten tio n ally left blank.] DRAFT D O N O T C IT E O R Q U O T E M ay 2010 E xternal R ev iew D raft APPENDIX D Epidemiological Kinetic Modeling N O T IC E T H IS D O C U M E N T IS A N E X T E R N A L R E V IE W D R A F T . It has n o t been form ally released by the U .S. E n v iro n m en tal P ro tectio n A g en cy and should n o t at th is stage b e co n stru ed to rep resen t A gency policy. It is b ein g circu lated fo r co m m en t on its technical accuracy and policy im plications. N ational C en ter fo r E n v iro n m en tal A ssessm ent O ffice o f R esearch and D evelopm ent U .S. E n v iro n m en tal P ro tectio n A g en cy C incinnati, O H C O N T E N T S -- A P P E N D IX D : E p id em io lo g ical K in etic M o d elin g L I S T O F T A B L E S .......................................................................................................................................................................D - i i i A P P E N D I X D . E P I D E M I O L O G I C A L K I N E T I C M O D E L I N G ............................................................... D - 1 D .1 . B A C C A R E L L I E T A L . ( 2 0 0 8 ) M O D E L I N G ..................................................................................D - 1 D .1 .1 . I n p u t F i l e f o r E x p o s u r e D u r i n g P r e g n a n c y ................................................................... D - 1 D .1 .2 . T a b l e o f R e s u l t s f o r B a c c a r e l l i e t a l. ( 2 0 0 8 ) ................................................................ D - 1 D .2 . M O C A R E L L I E T A L . ( 2 0 0 8 ) M O D E L I N G ................................................................................. D - 2 D .2 .1 . In p u t F ile fo r E x p o su re fo r P u lse to M e a su re m e n t 0.5 Y e ars A fte r th e S e v e s o P u l s e D o s e ..........................................................................................................................D - 2 D .2.2. In p u t F ile fo r E x p o su re fro m P u lse to th e E n d o f th e C ritical W in d o w 3 .8 Y e a r s A f t e r t h e S e v e s o P u l s e D o s e ............................................................................D - 2 D .2 .3 . I n p u t F i l e f o r C o n t i n u o u s E x p o s u r e f o r 1 0 Y e a r s ....................................................... D - 3 D .2 .4 . T a b l e s o f R e s u l t s f o r M o c a r e l l i e t a l. ( 2 0 0 8 ) ................................................................. D - 4 D .3 . A L A L U U S U A E T A L . ( 2 0 0 4 ) M O D E L I N G ....................................................................................D - 4 D .3 .1 . In p u t F ile fo r E x p o su re fo r P u lse to M e a su re m e n t 0.5 Y e ars A fte r th e S e v e s o P u l s e D o s e ..........................................................................................................................D - 4 D .3.2. In p u t F ile fo r E x p o su re fro m P u lse to th e E n d o f th e C ritical W in d o w 2 .5 Y e a r s A f t e r t h e S e v e s o P u l s e D o s e ............................................................................D - 5 D .3 .3 . I n p u t F i l e f o r C o n t i n u o u s E x p o s u r e f o r 5 Y e a r s .......................................................... D - 5 D .3 .4 . T a b l e s o f R e s u l t s f o r A l a l u u s u a e t a l. ( 2 0 0 4 ) ................................................................ D - 6 D .4 . E S K A N A Z I E T A L . ( 2 0 0 2 ) M O D E L I N G .......................................................................................... D - 7 D .4 .1 . In p u t F ile fo r E x p o su re fo r P u lse to M e a su re m e n t 0.5 Y e ars A fte r th e S e v e s o P u l s e D o s e ..........................................................................................................................D - 7 D .4.2. In p u t F ile fo r E x p o su re fro m P u lse to th e E n d o f th e C ritical W in d o w 6 .7 Y e a r s A f t e r t h e S e v e s o P u l s e D o s e ............................................................................. D - 8 D .4 .3 . I n p u t F i l e f o r C o n t i n u o u s E x p o s u r e f o r 13 Y e a r s ....................................................... D - 8 D .4 .4 . T a b l e s o f R e s u l t s f o r E s k a n a z i e t a l. ( 2 0 0 2 ) ...................................................................D - 9 D .5 . R E F E R E N C E S ....................................................................................................................................................D - 1 0 This document is a draftfo r review purposes only and does not constitute Agency policy. D-ii DRAFT--DO NOT CITE OR QUOTE L IS T O F TA B L E S D -1. E stim ated co n tin u o u s in tak e co rresp o n d in g to m atern al serum co n cen tratio n in F i g u r e 2 A ........................................................................................................................................................................... D - 1 D -2. E stim ated m ax im u m in tak e co rresp o n d in g to m aternal serum co n cen tratio n in F i g u r e 2 A ........................................................................................................................................................................... D - 2 D -3. M atch in g critical w in d o w av erag e after pulse to critical w in d o w av erag e fo r c o n t i n u o u s i n t a k e r u n ..................................................................................................................................................D - 4 D -4. M atch in g critical w in d o w p eak after p u lse to p eak critical w in d o w co n cen tratio n f o r c o n t i n u o u s i n t a k e r u n ......................................................................................................................................... D - 4 D -5. M atch in g critical w in d o w av erag e after pulse to critical w in d o w av erag e fo r c o n t i n u o u s i n t a k e r u n ..................................................................................................................................................D - 6 D -6 . M atch in g critical w in d o w p eak after pulse to p eak critical w in d o w c o n c e n t r a t i o n f o r c o n t i n u o u s i n t a k e r u n ........................................................................................................ D - 7 D -7. M atch in g critical w in d o w av erag e after pulse to critical w in d o w av erag e fo r c o n t i n u o u s i n t a k e r u n ..................................................................................................................................................D - 9 D -8 . M atch in g critical w in d o w p eak after pulse to p eak critical w in d o w c o n c e n t r a t i o n f o r c o n t i n u o u s i n t a k e r u n ........................................................................................................ D - 9 This document is a draftfo r review purposes only and does not constitute Agency policy. D-iii DRAFT--DO NOT CITE OR QUOTE 1 A P P E N D IX D. E P ID E M IO L O G IC A L K IN E T IC M O D E L IN G 2 3 4 D .1. B A C C A R E L L I E T A L . (2008) M O D E L IN G 5 D .1.1. I n p u t F ile fo r E x p o s u re D u rin g P re g n a n c y 6 C IN T = 1 % 168 % 100 % integration tim e 7 % E x p o su re scenario 8 E X P_T IM E _O N =0 % delay b efo re b eg in exposure (H O U R ) 9 E X P_T IM E _O FF = 401190 % T IM E E X P O S U R E STO P (H O U R ) 10 D A Y _C Y C L E = 24 % T IM E 11 B C K _ T IM E _ O N = 401190 % D E L A Y B E FO R E B A C K G R O U N D E X P (H O U R ) 12 B C K _T IM E _O F F = 401190 % T IM E O F B A C K G R O U N D E X P STO P (H O U R ) 13 IV _ L A C K =401190 14 IV _ P E R IO D =401190 15 % G E S T A T IO N C O N T R O L 16 M A T T IN G = 2 6 2800 % B E G IN N IN G M A T T IN G (H O U R )at 30 y ears old 17 T IM E L IM IT = 269184 % S IM U L A T IO N L IM IT T IM E (H O U R ) 18 T R A N S T IM E _ O N = 2 6 4 3 1 2 % E X C H A N G E M O T H E R F E T U S 1512 H O U R P O S T 19 M A T T IN G 20 % E xposure dose 21 M STO T = 0.021 % n g o f T C D D /k g o f B W 22 M STOTBCKGR = 0. % 0.1 % O R A L B A C K G R O U N D E X P O S U R E D O S E (n G /K G ) 23 D O S E IV = 0. % 10 24 D O SE IV L A T E = 0. % 10 25 26 % TR A N FER M O T H E R TO FETU S C LEA R A N C E 27 C LPLA _FET = 0.001 % M O T H E R T O F E T U S T R A N F E R T C L E A R A N C E (L /H R ) 28 29 D .1.2. T a b le o f R e su lts fo r B a c c a re lli e t al. (2008) 30 T ab le D -1. E stim a te d co n tin u o u s in ta k e c o rre sp o n d in g to m a te rn a l seru m 31 c o n c e n tra tio n in F ig u re 2 A 32 V ariab le V alu e N otes Infant b-T SH 5 uU /m L BM R M aternal lipid adjusted serum 270 ng/kg F rom F ig u re 2A In ta k e 33 34 35 36 0.024 n g /kg-day F rom E m o n d m odel, pregnancy at 30 years This document is a draftfo r review purposes only and does not constitute Agency policy. D-1 DRAFT--DO NOT CITE OR QUOTE 1 T ab le D -2. E stim a te d m ax im u m in ta k e co rre sp o n d in g to m a te rn a l seru m 2 c o n c e n tra tio n in F ig u re 2A 3 V ariab le V alu e N otes Infant b-T SH -- -- M aternal lipid adjusted serum 309.5 n g /k g M axim um from F igure 2A In ta k e 4 0.030 n g /kg-day F rom E m o n d m odel, pregnancy at 30 years 5 D .2. M O C A R E L L I E T A L . (2008) M O D E L IN G 6 D .2.1. I n p u t F ile fo r E x p o s u re fo r P u lse to M e a s u re m e n t 0.5 Y e a rs A fte r th e S eveso P u lse 7 D ose 8 C I N T = 1. % 9 E X P _ T IM E _ O N = 54312. % D elay b efo re b eg in ex p o su re (H O U R ) 6.2 y ears 10 E X P _T IM E _O F F = 54335. % 324120 % H O U R /Y E A R !T IM E E X P O S U R E STO P 11 (H O U R ) 6.2 y ears + 23 ho u rs 12 D A Y _C Y C L E = 24. % T IM E 13 B C K _ T IM E _ O N = 0. % D E L A Y B E FO R E B A C K G R O U N D E X P (H O U R ) 14 B C K _ T IM E _ O F F = 6 1 3 2 0 0 % T IM E O F B A C K G R O U N D E X P S T O P (H O U R ) 15 T IM E L IM IT = 5 8 6 9 2 . % h a lf a y e a r (Ju ly 1976 u n til Ja n u a ry 1 977) p a st 6 .2 y e a rs 16 M S T O T B C K G R = 3 .7 E -4 % O R A L B A C K G R O U N D E X P O S U R E D O S E (U G /K G ) 17 18 % oral d o se oral d o se oral d o se 19 M S T O T = 232.4 % S eveso, O R A L D A IL Y E X P O S U R E D O S E (N G /K G ) 20 D O SE IV =0 % 40 % 50 % 5 % 0.5 % 0.3 % 0 .2 % 0 .1 % 0 .0 5 % 0 .3 % N G /K G 21 % oral dose oral dose oral dose 22 23 M E A N L IP ID = 731 % 711 % 664 % 778 % 468 % 671 % 730 % 662 % 592% 615% 730% 24 PA S_IN D U C = 1 % N O N IN D U C T IO N (0) C O N T R O L E D E L 'IN D U C T IO N 25 26 % hum an variable param eter 2 7 M A L E = 1. 28 F E M A L E = 0. 29 Y 0 = 0. % 0 y ears old at th e b eg in n in g o f th e sim ulation 30 31 D .2 .2 . I n p u t F ile fo r E x p o s u re fro m P u lse to th e E n d o f th e C ritic a l W in d o w 3.8 Y e a rs 32 A fte r th e Seveso P u lse D ose 33 C I N T = 1. % 34 E X P _ T IM E _ O N = 54312. % D elay b efo re b eg in ex p o su re (H O U R ) 6.2 y ears 35 E X P _T IM E _O F F = 54335. % 324120 % H O U R /Y E A R !T IM E E X P O S U R E STO P 36 (H O U R ) 6.2 y ears + 23 ho u rs 37 D A Y _C Y C L E = 24. % T IM E This document is a draftfo r review purposes only and does not constitute Agency policy. D-2 DRAFT--DO NOT CITE OR QUOTE 1 B C K _ T IM E _ O N = 0. % D E L A Y B E FO R E B A C K G R O U N D E X P (H O U R ) 2 B C K _T IM E _O F F = 613200. % T IM E O F B A C K G R O U N D E X P STO P (H O U R ) 3 T IM E L IM IT = 87600. % 10 years 4 M S T O T B C K G R = 3.7e-4 % O R A L B A C K G R O U N D E X P O S U R E D O S E (U G /K G ) 5 6 % oral dose oral dose oral dose 7 M STOT = 232.5 % Serveso, O R A L D A IL Y E X P O S U R E D O S E (N G /K G ) 8 D O SE IV = 0 % 40 % 50 % 5 % 0.5 % 0.3 % 0 .2 % 0 .1 % 0 .0 5 % 0 .3 % N G /K G 9 % oral dose oral dose oral dose 10 11 M E A N L IP ID = 730 % 711 % 664 % 778 % 468 % 671 % 730 % 662 % 592% 615% 730% 12 PA S_IN D U C = 1 % N O N IN D U C T IO N (0) C O N T R O L E D E L 'IN D U C T IO N 13 14 % h u m an v ariab le p aram eter 15 M A L E = 1. 16 F E M A L E = 0. 17 Y 0 = 0. % 0 y ears old at th e b eg in n in g o f th e sim ulation 18 19 D .2 .3 . I n p u t F ile fo r C o n tin u o u s E x p o s u re fo r 10 Y e a rs 2 0 C I N T = 1. % 21 E X P _ T IM E _ O N = 0. % D elay b efo re b eg in exposure (H O U R ) 2 2 E X P _T IM E _O F F = 87600. % H O U R /Y E A R !TIM E E X P O S U R E STO P (H O U R ) 23 D A Y _C Y C L E = 24. % T IM E 24 B C K _ T IM E _ O N = 0. % 324120 % D E L A Y B E FO R E B A C K G R O U N D E X P (H O U R ) 25 B C K _T IM E _O FF = 613200 % 324120 % T IM E O F B A C K G R O U N D E X P STO P (H O U R ) 26 T IM E L IM IT = 87600. % 10 years 27 M S T O T B C K G R = 0. % 3.35E -4 % O R A L B A C K G R O U N D E X P O S U R E D O S E (U G /K G ) 28 29 % oral dose oral dose oral dose 30 M ST O T = 3.903 % S eveso, O R A L D A IL Y E X P O S U R E D O S E (N G /K G ) 31 D O S E IV =0 % 40 % 50 % 5 % 0.5 % 0.3 % 0 .2 % 0 .1 % 0 .0 5 % 0 .3 % N G /K G 32 % oral dose oral dose oral dose 33 34 M E A N L IPID = 730 % 711 % 664 % 778 % 468 % 671 % 730 % 662 % 592% 615% 730% 35 P A S _IN D U C = 1 % N O N IN D U C T IO N (0) C O N T R O L E D E L 'IN D U C T IO N 36 37 % hum an variable param eter 3 8 M A L E = 1. 39 F E M A L E = 0. 40 Y 0 = 0. % 0 y ears old at th e b eg in n in g o f th e sim ulation 41 42 43 44 45 This document is a draftfo r review purposes only and does not constitute Agency policy. D-3 DRAFT--DO NOT CITE OR QUOTE 1 D .2.4. T a b le s o f R e su lts fo r M o c a re lli e t al. (2008) 2 3 T ab le D -3. M a tc h in g critica l w in d o w av e ra g e a fte r p u lse to critica l w in d o w 4 av erag e fo r co n tin u o u s in tak e ru n 5 P erson m odeled, beginning at age 0 L ipid adjusted serum (1976) ng/kg from F igure 3E P u lse dose, 0.5 y ear lag tim e (ng/kg) A verage lipid adjusted serum 3.8 y e a rs a fte r incident (ng/kg) C ontinuous intake for 10 years (n g /k g -d a y ) Boy, 1st quartile 68 8.135 57.72 0.008024 Boy, 4th quartile 733 232.5 580.5 0.2128 6 7 8 T ab le D -4. M a tc h in g critica l w in d o w p e a k a fte r pu lse to p e a k critical 9 w in d o w co n c en tratio n fo r co n tin u o u s in tak e ru n 10 P erson m odeled, beginning at age 0 L ipid adjusted serum (1976) ng/kg from F igure 3E P u lse dose, 0.5 y ear lag tim e (ng/kg) P eak lipid adjusted serum after incident (ng/kg) C ontinuous intake for 10 years (n g /k g -d a y ) Boy, 1st quartile 68 8.135 248.0 0.03194 Boy, 4th quartile 733 232.5 6674 3.904 11 12 13 D .3 . A L A L U U S U A E T A L . (2 0 0 4 ) M O D E L I N G 14 D .3 .1 . I n p u t F ile fo r E x p o s u re fo r P u lse to M e a s u re m e n t 0.5 Y e a rs A fte r th e S ev eso P u lse 15 D o se 16 C I N T = 1. % 17 E X P _ T IM E _ O N = 21900. % D elay b efo re b eg in ex p o su re (H O U R ) 2.5 y ears 18 E X P _ T IM E _ O F F = 2 1 9 2 3 . % 2 1 9 0 0 + 2 3 % H O U R /Y E A R !T IM E E X P O S U R E S T O P 19 (H O U R ) 2.5 y ears and 23 h o u rs 20 D A Y _C Y C L E = 24. % TIM E 21 B C K _ T IM E _ O N = 0. % D E L A Y B E FO R E B A C K G R O U N D E X P (H O U R ) 22 B C K _T IM E _O F F = 613200. % T IM E O F B A C K G R O U N D E X P STO P (H O U R ) 23 T IM E L IM IT = 26280. % h a lf a y e a r (Ju ly 1976 until Jan u ary 1977) p ast 2.5 y ears 24 M S T O T B C K G R = 3.7e-4 % O R A L B A C K G R O U N D E X P O S U R E D O S E (U G /K G ) 25 26 % oral dose oral dose oral dose 27 M STO T = 2 4.22 % Seveso, O R A L D A IL Y E X P O S U R E D O S E (N G /K G ) 28 D O SE IV =0 % 40 % 50 % 5 % 0.5 % 0.3 % 0 .2 % 0 .1 % 0 .0 5 % 0 .3 % N G /K G 29 % oral dose oral dose oral dose 30 This document is a draftfo r review purposes only and does not constitute Agency policy. D-4 DRAFT--DO NOT CITE OR QUOTE 1 M E A N L IPID = 730 % 711 % 664 % 778 % 468 % 671 % 730 % 662 % 592% 615% 730% 2 PA S_IN D U C = 1 % N O N IN D U C T IO N (0) C O N T R O L E D E L 'IN D U C T IO N 3 4 % hum an variable param eter 5 M A L E = 1. 6 F E M A L E = 0. 7 Y 0 = 0. % 0 y ears old at th e b eg in n in g o f th e sim ulation 8 9 D .3.2. I n p u t F ile fo r E x p o s u re fro m P u lse to th e E n d o f th e C ritic a l W in d o w 2.5 Y e a rs 10 A fte r th e Seveso P u lse D ose 11 C I N T = 1. % 12 E X P _ T IM E _ O N = 21900. % D elay b efo re b eg in ex p o su re (H O U R ) 2.5 y ears 13 E X P _ T IM E _ O F F = 2 1 9 2 3 . % 3 2 4 1 2 0 % H O U R /Y E A R !T IM E E X P O S U R E S T O P 14 (H O U R ) 2.5 y ears and 23 h o u rs 15 D A Y _ C Y C L E = 24. % TIM E 16 B C K _ T IM E _ O N = 0. % 324120 % D E L A Y B E FO R E B A C K G R O U N D E X P (H O U R ) 17 B C K _ T IM E _ O F F = 613200. % 3 2 4 1 2 0 % T IM E O F B A C K G R O U N D E X P S T O P (H O U R ) 18 T IM E L IM IT = 4 3 8 0 0 . % 5 y ea rs 19 M S T O T B C K G R = 3 .7 e-4 % O R A L B A C K G R O U N D E X P O S U R E D O S E (U G /K G ) 20 21 % oral dose oral dose oral dose 22 M STOT = 24.22 % S eveso, O R A L D A IL Y E X P O S U R E D O S E (N G /K G ) 23 D O S E IV =0 % 40 % 50 % 5 % 0.5 % 0.3 % 0 .2 % 0 .1 % 0 .0 5 % 0 .3 % N G /K G 24 % oral dose oral dose oral dose 25 26 M E A N L IPID = 730 % 711 % 664 % 778 % 468 % 671 % 730 % 662 % 592% 615% 730% 27 PA S_IN D U C = 1 % N O N IN D U C T IO N (0) C O N T R O L E D E L 'IN D U C T IO N 28 29 % hum an variable param eter 3 0 M A L E = 1. 31 F E M A L E = 0. 32 Y 0 = 0. % 0 y ears old at th e b eg in n in g o f th e sim ulation 33 34 D .3.3. I n p u t F ile fo r C o n tin u o u s E x p o s u re fo r 5 Y e a rs 35 C I N T = 1. % 36 E X P _ T IM E _ O N = 0. % D elay b efore b eg in exposure (H O U R ) 37 E X P _T IM E _O F F = 43800. % 324120 % H O U R /Y E A R !TIM E E X P O S U R E STO P (H O U R ) 38 D A Y _C Y C L E = 24. % TIM E 39 B C K _ T IM E _ O N = 0. % 324120 % D E L A Y B E FO R E B A C K G R O U N D E X P (H O U R ) 40 B C K _T IM E _O FF = 613200. % 324120 % T IM E O F B A C K G R O U N D E X P STO P (H O U R ) 41 T IM E L IM IT = 43800. % E n d o f critical w in d o w (5 y ears) 42 M S T O T B C K G R = 0. % O R A L B A C K G R O U N D E X P O S U R E D O S E (U G /K G ) 43 44 % oral dose oral dose oral dose 45 M ST O T = 0.03486 % Seveso, O R A L D A IL Y E X P O S U R E D O S E (N G /K G ) This document is a draftfo r review purposes only and does not constitute Agency policy. D-5 DRAFT--DO NOT CITE OR QUOTE 1 DOSEIV = 0 % 40 %50 %5 %0.5 %0.3 %0.2 %0.1%0.05%0.3 %NG/KG 2 % oral dose oral dose oral dose 3 4 M E A N L IPID = 730 % 711 % 664 % 778 % 468 % 671 % 730 % 662 % 592% 615% 730% 5 PA S_IN D U C = 1 % N O N IN D U C T IO N (0) C O N T R O L E D E L 'IN D U C T IO N 6 7 % hum an variable param eter 8 M A L E = 1. 9 F E M A L E = 0. 10 Y 0 = 0. % 0 y ears old at th e b eg in n in g o f th e sim ulation 11 12 D .3.4. T a b le s o f R e su lts fo r A la lu u s u a e t al. (2004) 13 T a b le D -5 . M a tc h in g c r itic a l w in d o w a v e r a g e a f te r p u ls e to c r itic a l w in d o w 14 a v e ra g e fo r c o n tin u o u s in ta k e r u n 15 P erson m odeled, beginning at age 0 L ipid adjusted serum (1976) ng/kg estim ated from tertile binsa Pulse dose, 0.5 y e a r lag tim e (ng/kg) A verage lipid a d ju ste d seru m 2.5 years after incident (ng/kg) C ontinuous intake for 5 y ears (ng/kg- day) Boy, 1st tertile 130 24.22 110.8 0.03486 Boy, 2nd tertile 383 108.9 322.7 0.1578 Boy, 3rd tertile 1830 1041 1538 1.511 G irl, 1st tertile 130 23.03 110.8 0.03211 G irl, 2nd tertile 383 105.3 324.4 0.1481 G irl, 3rd tertile 1830 1015 1546 1.427 B oy and girl, averaged, 1st tertile 130 - - 0.03349 B oy and girl, averaged, 383 - - 0.1530 2 nd tertile B oy and girl, averaged, 3rd tertile 1830 - 16 17 aM ean o f tertile b in assum ing a lognorm al distribution o f serum concentrations. - 1.469 This document is a draftfo r review purposes only and does not constitute Agency policy. D-6 DRAFT--DO NOT CITE OR QUOTE 1 T ab le D -6. M a tc h in g critica l w in d o w p e a k a fte r pu lse to p e a k critical 2 w in d o w co n c en tratio n fo r co n tin u o u s in tak e ru n 3 P erson m odeled, beginning at age 0 L ipid adjusted serum (1976) ng/kg estim ated from tertile bins P ulse dose, 0.5 y e a r lag tim e (ng/kg) P eak lipid adjusted serum after incident (ng/kg) C ontinuous intake for 5 y ears (ng/kg- day) Boy, 1st tertile Boy, 2nd tertile Boy, 3rd tertile G irl, 1st tertile 130 383 1830 130 24.22 108.9 1041 23.02 618.8 2700 24706 588.0 0.2113 1.783 31.35 0.1882 G irl, 2nd tertile 383 105.3 2610 G irl, 3rd tertile 1830 1015 24113 B oy and girl, averaged, 130 - - 1st tertile B oy and girl, averaged, 383 - - 2 nd tertile B oy and girl, averaged, 3rd tertile 1830 - - 4 5 aM ean o f tertile b in assum ing a lognorm al distribution o f serum concentrations. 6 7 8 D .4. E S K A N A Z I E T A L . (2002) M O D E L IN G 1.642 29.52 0.1998 1.713 30.44 9 D .4.1. I n p u t F ile fo r E x p o s u re fo r P u lse to M e a s u re m e n t 0.5 Y e a rs A fte r th e S eveso P u lse 10 D ose 11 C I N T = 1. % 12 E X P_T IM E _O N = 58692. % D elay b efo re b eg in ex p o su re (H O U R ) 6.7 y ears 13 E X P _ T IM E _ O F F = 5 8 7 1 5 . % H O U R /Y E A R !T IM E E X P O S U R E S T O P (H O U R ) 6 .7 y e a rs + 14 23 h o u rs 15 D A Y _ C Y C L E = 24. % T IM E 16 B C K _ T IM E _ O N = 0. % 324120 % D E L A Y B E FO R E B A C K G R O U N D E X P (H O U R ) 17 B C K _ T IM E _ O F F = 613200. % 3 2 4 1 2 0 % T IM E O F B A C K G R O U N D E X P S T O P (H O U R ) 18 T IM E L IM IT = 6 3 0 7 2 . % h a lf a y e a r (Ju ly 1976 u n til Jan u a ry 1977) p a st 6.7 y ea rs 19 M S T O T B C K G R = 3 .7 e-4 % O R A L B A C K G R O U N D E X P O S U R E D O S E (U G /K G ) 20 21 % oral dose oral dose oral dose 22 M STOT = 7193 % Seveso, O R A L D A IL Y E X P O S U R E D O S E (N G /K G ) 23 D O S E IV =0 % 40 % 50 % 5 % 0.5 % 0.3 % 0 .2 % 0 .1 % 0 .0 5 % 0 .3 % N G /K G 24 % oral dose oral dose oral dose 25 26 M E A N L IPID = 730 % 711 % 664 % 778 % 468 % 671 % 730 % 662 % 592% 615% 730% This document is a draftfo r review purposes only and does not constitute Agency policy. D-7 DRAFT--DO NOT CITE OR QUOTE 1 PAS INDUC= 1 % NON INDUCTION (0) CONTROLE DE L'INDUCTION 2 3 % hum an variable param eter 4 M A L E = 0. 5 F E M A L E = 1. 6 Y 0 = 0. % 0 y ears old at th e b eg in n in g o f th e sim ulation 7 8 D .4 .2 . I n p u t F ile fo r E x p o s u re fro m P u ls e to th e E n d o f th e C ritic a l W in d o w 6.7 Y e a rs 9 A fte r th e Seveso P u lse D ose 1 0 C I N T = 1. % 11 E X P _ T IM E _ O N = 58692. % D elay b efo re b eg in ex p o su re (H O U R ) 6.7 y ears 12 E X P _T IM E _O F F = 58715. % 324120 % H O U R /Y E A R !T IM E E X P O S U R E STO P 13 (H O U R ) 6 .7 y e a rs + 23 h o u rs 14 D A Y _ C Y C L E = 24. % T IM E 15 B C K _ T IM E _ O N = 0. % 324120 % D E L A Y B E FO R E B A C K G R O U N D E X P (H O U R ) 16 B C K _ T IM E _ O F F = 6 1 3 2 0 0 % 3 2 4 1 2 0 % T IM E O F B A C K G R O U N D E X P S T O P (H O U R ) 17 T IM E L IM IT = 113880. % 13 y e a rs 18 M S T O T B C K G R = 3 .7 e-4 % O R A L B A C K G R O U N D E X P O S U R E D O S E (U G /K G ) 19 2 0 % oral dose oral dose oral dose 21 M STO T = 7193 % Seveso, O R A L D A IL Y E X P O S U R E D O S E (N G /K G ) 22 D O SE IV =0 % 40 % 50 % 5 % 0.5 % 0.3 % 0 .2 % 0 .1 % 0 .0 5 % 0 .3 % N G /K G 23 % oral dose oral dose oral dose 24 25 M E A N L IPID = 730 % 711 % 664 % 778 % 468 % 671 % 730 % 662 % 592% 615% 730% 26 PA S_IN D U C = 1 % N O N IN D U C T IO N (0) C O N T R O L E D E L 'IN D U C T IO N 27 28 % hum an variable param eter 29 M A L E = 0. 3 0 F E M A L E = 1. 31 Y 0 = 0. % 0 y ears old at th e b eg in n in g o f th e sim ulation 32 33 D .4.3. I n p u t F ile fo r C o n tin u o u s E x p o s u re fo r 13 Y e a rs 3 4 C I N T = 1. % 35 E X P _ T IM E _ O N = 0. % D elay b efore b eg in exposure (H O U R ) 36 E X P _T IM E _O F F = 113880. % 324120 % H O U R /Y E A R !T IM E E X P O S U R E STO P 37 (H O U R ) 13 y e a rs 38 D A Y _C Y C L E = 24. % T IM E 39 B C K _ T IM E _ O N = 0. % 324120 % D E L A Y B E FO R E B A C K G R O U N D E X P (H O U R ) 40 B C K _T IM E _O FF = 613200. % 324120 % T IM E OF B A C K G R O U N D E X P STO P (H O U R ) 41 T IM E L IM IT = 113880. % 13 y e a rs 42 M S T O T B C K G R = 0. % 3.35E -4 % O R A L B A C K G R O U N D E X P O S U R E D O S E (U G /K G ) 43 44 % oral dose oral dose oral dose 45 M ST O T = 166 % Seveso, O R A L D A IL Y E X P O S U R E D O S E (N G /K G ) This document is a draftfo r review purposes only and does not constitute Agency policy. D-8 DRAFT--DO NOT CITE OR QUOTE 1 DOSEIV = 0 % 40 %50 %5 %0.5 %0.3 %0.2 %0.1%0.05%0.3 %NG/KG 2 % oral dose oral dose oral dose 3 4 M E A N L IPID = 730 % 711 % 664 % 778 % 468 % 671 % 730 % 662 % 592% 615% 730% 5 PA S_IN D U C = 1 % N O N IN D U C T IO N (0) C O N T R O L E D E L 'IN D U C T IO N 6 7 % hum an variable param eter 8 M A L E = 0. 9 F E M A L E = 1. 10 Y 0 = 0. % 0 y ears old at th e b eg in n in g o f th e sim ulation 11 12 D .4.4. T a b le s o f R e su lts fo r E s k a n a z i e t al. (2002) 13 T a b le D -7 . M a tc h in g c r itic a l w in d o w a v e r a g e a f te r p u ls e to c r itic a l w in d o w 14 a v e ra g e fo r c o n tin u o u s in ta k e r u n 15 P erson m odeled, beginning at age 0 L ipid adjusted serum (ad ju sted to 1976 1977 levels) ng/kg from F igure 1A Pulse dose, 0.5 y e a r lag tim e (ng/kg) A verage lipid a d ju ste d seru m 6.7 years after incident (ng/kg) C ontinuous in tak e fo r 13 y ears (ng/kg-day) G irl, estrous cycle 28.5 days 166 28.40 114.0 0.01660 G irl, estrous cycle 29 days 693 215.5 455.1 0.1224 G irl, estrous cycle 29.5 days 2020 1008 1295 0.5693 G irl, estrous cycle 30 days 8450 7193 5179 4.054 16 17 T a b le D -8 . M a tc h in g c ritic a l w in d o w p e a k a f te r p u lse to p e a k c ritic a l 18 w in d o w c o n c e n tra tio n fo r c o n tin u o u s in ta k e r u n 19 P erson m odeled, beginning at age 0 L ipid adjusted serum (ad ju sted to 1976 1977 levels) ng/kg from F igure 1A P ulse dose, 0.5 y e a r lag tim e (ng/kg) P eak lipid adjusted serum after incident (ng/kg) C ontinuous intake fo r 13 years (n g /k g -d a y ) G irl, estrous cycle 28.5 days 166 28.40 838.2 0.1800 G irl, estrous cycle 29 days 693 215.5 6183 3.148 G irl, estrous cycle 29.5 days 2020 1008 28316 20.86 G irl, estrous cycle 30 days 8450 7193 198240 166.6 This document is a draftfo r review purposes only and does not constitute Agency policy. D-9 DRAFT--DO NOT CITE OR QUOTE 1 2 D .5. R E F E R E N C E S 3 A laluusua, S; Calderara, P; G erthoux, PM ; et al. (2004) D evelopm ental dental aberrations after the d ioxin accident 4 in Seveso. Environ H ealth Perspect 112(13):1313-1318. 5 Baccarelli, A; Giacomini, SM; Corbetta, C; et al. (2008) N eonatal thyroid function in Seveso 25 years after m aternal 6 exposure to dioxin. PLoS M ed 5(7):e161. 7 Eskenazi, B; M ocarelli, P; W arner, M ; et al. (2002). Serum dioxin concentrations and endom etriosis: a cohort study 8 in Seveso, Italy. Environ H ealth Perspect 110(7): 629-634. 9 M ocarelli, P; Gerthoux, PM ; Patterson, DG, Jr.; et al. (2008) D ioxin exposure, from infancy through puberty, 10 produces endocrine disruption and affects hum an semen quality. Environ H ealth Perspect 116(1):70-77. 11 This document is a draftfo r review purposes only and does not constitute Agency policy. D-10 DRAFT--DO NOT CITE OR QUOTE [T his p ag e in ten tio n ally left blank.] DRAFT D O N O T C IT E O R Q U O T E M ay 2010 E xternal R ev iew D raft APPENDIX E Noncancer Benchmark Dose Modeling N O T IC E T H IS D O C U M E N T IS A N E X T E R N A L R E V IE W D R A F T . It has n o t been form ally released by the U .S. E n v iro n m en tal P ro tectio n A g en cy and should n o t at th is stage b e co n stru ed to rep resen t A gency policy. It is b ein g circu lated fo r co m m en t on its technical accuracy and policy im plications. N ational C en ter fo r E n v iro n m en tal A ssessm ent O ffice o f R esearch and D evelopm ent U .S. E n v iro n m en tal P ro tectio n A g en cy C incinnati, O H C O N T E N T S -- A P P E N D IX E : N o n can cer B e n c h m a rk D ose M o d elin g A P P E N D I X E . N O N C A N C E R B E N C H M A R K D O S E M O D E L I N G ............................................... E -1 E .1 . B M D S I N P U T T A B L E S .................................................................................................................................. E -1 E .1 .1 . A m i n e t a l. ( 2 0 0 0 ) .............................................................................................................................. E -1 E .1 .2 . B e l l e t a l. ( 2 0 0 7 ) ..................................................................................................................................E -1 E .1 .3 . C a n t o n i e t a l. ( 1 9 8 1 ) .........................................................................................................................E - 2 E .1 .4 . C r o f t o n e t a l. ( 2 0 0 5 ) .......................................................................................................................... E - 2 E .1 .5 . D e C a p r i o e t a l. ( 1 9 8 6 ) ...................................................................................................................... E - 3 E .1 .6 . F r a n c e t a l. ( 2 0 0 1 ) ............................................................................................................................... E - 4 E .1 .7 . H o j o e t a l. ( 2 0 0 2 ) .................................................................................................................................E - 4 E .1 .8 . K a t t a i n e n e t a l. ( 2 0 0 1 ) ...................................................................................................................... E - 5 E .1 .9 . K e l l e r e t a l. ( 2 0 0 7 , 2 0 0 8 a , b ) ....................................................................................................... E - 5 E . 1 .1 0 . K o c i b a e t a l. ( 1 9 7 8 ) ............................................................................................................................ E - 6 E . 1 .1 1 . L a t c h o u m y c a n d a n e a n d M a t h u r ( 2 0 0 2 ) ................................................................................ E - 6 E . 1 .1 2 . L i e t a l. ( 1 9 9 7 ) .......................................................................................................................................E - 7 E . 1 .1 3 . L i e t a l. ( 2 0 0 6 ) .......................................................................................................................................E - 7 E . 1 .1 4 . M a r k o w s k i e t a l. ( 2 0 0 1 ) .................................................................................................................. E - 8 E . 1 .1 5 . M i e t t i n e n e t a l. ( 2 0 0 6 ) .......................................................................................................................E - 8 E . 1 .1 6 . N a t i o n a l T o x i c o l o g y P r o g r a m ( 1 9 8 2 ) .................................................................................... E - 9 E . 1 .1 7 . N a t i o n a l T o x i c o l o g y P r o g r a m ( 2 0 0 6 ) .................................................................................E - 1 0 E . 1 .1 8 . O h s a k o e t a l. ( 2 0 0 1 ) ........................................................................................................................ E - 1 1 E . 1 .1 9 . S h i e t a l. ( 2 0 0 7 ) .................................................................................................................................. E - 1 1 E . 1 .2 0 . S m i a l o w i c z e t a l. ( 2 0 0 8 ) .............................................................................................................. E - 1 2 E . 1 .2 1 . T o t h e t a l. ( 1 9 7 9 ) ............................................................................................................................... E - 1 2 E . 1 .2 2 . V a n B i r g e l e n e t a l. ( 1 9 9 5 ) ........................................................................................................... E - 1 3 E . 1 .2 3 . W h i t e e t a l. ( 1 9 8 6 ) ...........................................................................................................................E - 1 3 E .2 . A L T E R N A T E D O S E : W H O L E B L O O D B M D S R E S U L T S ............................................... E - 1 4 E .2 .1 . A m i n e t a l., 2 0 0 0 : 0 .2 5 % S a c c h a r i n C o n s u m e d , F e m a l e ........................................ E - 1 4 E .2 .1 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ........................................ E - 1 4 E .2 .1 .2 . O u t p u t f o r S e l e c t e d M o d e l : L i n e a r .................................................................E - 1 4 E .2 .1 .3 . F i g u r e f o r S e l e c t e d M o d e l : L i n e a r ................................................................ E - 1 7 E .2.1.4. O u tp u t fo r A d d itio n al M o d el P resen ted : P ow er, U n re stric te d ....E -17 E .2 .1 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : P o w e r , U n r e s t r i c t e d ......E - 2 0 E .2 .2 . A m i n e t a l., 2 0 0 0 : 0 .2 5 % S a c c h a r i n P r e f e r e n c e R a t i o , F e m a l e ..........................E - 2 0 E .2 .2 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ........................................ E - 2 0 E .2 .2 .2 . O u t p u t f o r S e l e c t e d M o d e l : L i n e a r .................................................................E - 2 1 E .2 .2 .3 . F i g u r e f o r S e l e c t e d M o d e l : L i n e a r ................................................................ E - 2 3 E .2 .3 . A m i n e t a l., 2 0 0 0 : 0 .5 0 % S a c c h a r i n C o n s u m e d , F e m a l e ........................................ E - 2 4 E .2 .3 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ........................................ E - 2 4 E .2 .3 .2 . O u t p u t f o r S e l e c t e d M o d e l : L i n e a r .................................................................E - 2 4 E .2 .3 .3 . F i g u r e f o r S e l e c t e d M o d e l : L i n e a r ................................................................ E - 2 7 E .2.3.4. O u tp u t fo r A d d itio n al M o d el P resen ted : P ow er, U n restricted ....E -27 E .2 .3 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : P o w e r , U n r e s t r i c t e d ......E - 3 0 This document is a draftfor review purposes only and does not constitute Agency policy. E-ii DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) E .2 .4 . A m i n e t a l., 2 0 0 0 : 0 .5 0 % S a c c h a r i n P r e f e r e n c e R a t i o , F e m a l e ..........................E - 3 1 E .2.4.1. S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .......................... E - 3 1 E .2 .4 .2 . O u t p u t f o r S e l e c t e d M o d e l : L i n e a r .................................................................. E - 3 1 E .2 .4 .3 . F i g u r e f o r S e l e c t e d M o d e l : L i n e a r ..................................................................E - 3 4 E .2.4.4.O u tp u t fo r A d d itio n al M o d el P resen ted : P ow er, U n restricted ....E -34 E .2 .4 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : P o w e r , U n r e s t r i c t e d .........E - 3 7 E .2 .5 . B e l l e t a l., 2 0 0 7 a : B a l a n o - P r e p u t i a l S e p a r a t i o n , P o s t n a t a l D a y 4 9 .....................E - 3 8 E .2.5.1. S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .......................... E - 3 8 E .2 .5 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - L o g i s t i c .................................................. E - 3 8 E .2 .5 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - L o g i s t i c ....................................................E - 4 0 E .2.5.4. O u tp u t fo r A d d itio n al M o d el P resen ted : L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................... E - 4 1 E .2.5.5. F ig u re fo r A d d itio n al M o d el P resented: L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................... E - 4 3 E .2 .6 . C a n t o n i e t a l., 1 9 8 1 : U r i n a r y C o p r o p o r h y r i n s , 3 M o n t h s ....................................... E - 4 4 E .2 .6 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ....................................... E - 4 4 E .2 .6 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ........................................ E - 4 4 E .2 .6 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ........................................ E - 4 7 E .2.6.4.O u tp u t fo r A d d itio n al M o d el P resen ted : P ow er, U n restricted ....E -47 E .2 .6 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : P o w e r , U n r e s t r i c t e d .........E - 5 0 E .2.7. C antoni e t a l., 1 9 8 1 : U r i n a r y P o r p h y r i n s ......................................... E - 5 1 E .2.7.1. S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .......................... E - 5 1 E .2 .7 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 2 ) ........................................ E - 5 1 E .2 .7 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 2 ) ........................................ E - 5 4 E .2.8. C ro fto n e t a l., 2 0 0 5 : S e r u m , T 4 ...................................................E - 5 5 E .2.8.1. S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .......................... E - 5 5 E .2 .8 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ........................................ E - 5 5 E .2 .8 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ........................................ E - 5 8 E .2 .9 . F r a n c e t a l., 2 0 0 1 : S - D R a t s , R e l a t i v e L i v e r W e i g h t .................................................E - 5 9 E .2.9.1. S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .......................... E - 5 9 E .2 .9 .2 . O u t p u t f o r S e l e c t e d M o d e l : P o w e r ............................................................... E - 5 9 E .2 .9 .3 . F i g u r e f o r S e l e c t e d M o d e l : P o w e r .................................................................E - 6 2 E . 2 .1 0 . F r a n c e t a l., 2 0 0 1 : L - E R a t s , R e l a t i v e L i v e r W e i g h t .................................................E - 6 3 E .2 .1 0 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ........................................ E - 6 3 E .2 .1 0 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ..................................................................... E - 6 3 E .2 .1 0 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ...................................................................... E - 6 6 E .2 .1 0 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ......... E - 6 6 E .2 .1 0 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d .......... E - 6 9 E . 2 .1 1 . F r a n c e t a l., 2 0 0 1 : S - D R a t s , R e l a t i v e T h y m u s W e i g h t ...........................................E - 7 0 E .2 .1 1 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ........................................ E - 7 0 E .2 .1 1 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ......................................E - 7 0 E .2 .1 1 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ......................................E - 7 3 E .2 .1 1 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o ly n o m ial, 3 -d eg ree ..E -73 This document is a draftfor review purposes only and does not constitute Agency policy. E-iii DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) E .2 .1 1 .5 . F ig u re fo r A d d itio n al M o d el P resen ted : P o ly n o m ial, 3 -d eg ree ..E -7 6 E . 2 .1 2 . F r a n c e t a l., 2 0 0 1 : L - E R a t s , R e l a t i v e T h y m u s W e i g h t ...........................................E - 7 7 E .2 .1 2 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ........................................ E - 7 7 E .2 .1 2 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ....................................... E - 7 7 E .2 .1 2 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ....................................... E - 8 0 E . 2 .1 3 . F r a n c e t a l., 2 0 0 1 : H / W R a t s , R e l a t i v e T h y m u s W e i g h t ........................................ E - 8 1 E .2 .1 3 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ........................................ E - 8 1 E .2 .1 3 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 2 ) ....................................... E - 8 1 E .2 .1 3 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 2 ) ....................................... E - 8 4 E . 2 .1 4 . H o j o e t a l., 2 0 0 2 : D R L R e i n f o r c e P e r M i n u t e ............................................................... E - 8 5 E .2 .1 4 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ........................................ E - 8 5 E .2 .1 4 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ....................................... E - 8 5 E .2 .1 4 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ....................................... E - 8 8 E . 2 .1 5 . H o j o e t a l., 2 0 0 2 : D R L R e s p o n s e P e r M i n u t e ............................................................... E - 8 9 E .2 .1 5 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ........................................ E - 8 9 E .2 .1 5 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ....................................... E - 8 9 E .2 .1 5 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ....................................... E - 9 2 E . 2 .1 6 . K a t t a i n e n e t a l., 2 0 0 1 : 3 r d M o l a r E r u p t i o n , F e m a l e ................................................. E - 9 3 E .2 .1 6 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .........................................E - 9 3 E .2 .1 6 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - L o g i s t i c ................................................. E - 9 3 E .2 .1 6 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - L o g i s t i c ...................................................E - 9 5 E .2 .1 6 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................... E - 9 5 E .2 .1 6 .5 . F ig u re fo r A d d itio n al M o d el P resented: L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................... E - 9 7 E . 2 .1 7 . K a t t a i n e n e t a l., 2 0 0 1 : 3 r d M o l a r L e n g t h , F e m a l e ......................................................E - 9 8 E .2 .1 7 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .........................................E - 9 8 E .2 .1 7 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l .......................................................................E - 9 8 E .2 .1 7 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ..................................................................... E - 1 0 1 E .2 .1 7 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ....... E - 1 0 1 E .2 .1 7 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d .........E - 1 0 4 E . 2 .1 8 . K e l l e r e t a l., 2 0 0 7 : M i s s i n g M a n d i b u l a r M o l a r s , C B A J ..................................... E - 1 0 5 E .2 .1 8 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .................................... E - 1 0 5 E .2 .1 8 .2 . O u t p u t f o r S e l e c t e d M o d e l : M u l t i s t a g e , 1 - D e g r e e ............................ E - 1 0 5 E .2 .1 8 .3 . F i g u r e f o r S e l e c t e d M o d e l : M u l t i s t a g e , 1 - D e g r e e ............................. E - 1 0 7 E . 2 .1 9 . K o c i b a e t a l., 1 9 7 8 : U r i n a r y C o p r o p o r p h y r i n , F e m a l e s ....................................... E - 1 0 8 E .2 .1 9 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ...................................... E - 1 0 8 E .2 .1 9 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) .....................................E - 1 0 8 E .2 .1 9 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) .....................................E - 1 1 1 E . 2 .2 0 . K o c i b a e t a l., 1 9 7 8 : U r o p o r p h y r i n p e r C r e a ti n i n e , F e m a l e .................................E - 1 1 2 E .2 .2 0 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .................................... E - 1 1 2 E .2 .2 0 .2 . O u t p u t f o r S e l e c t e d M o d e l : L i n e a r .............................................................. E - 1 1 2 E .2 .2 0 .3 . F i g u r e f o r S e l e c t e d M o d e l : L i n e a r .............................................................. E - 1 1 5 This document is a draftfor review purposes only and does not constitute Agency policy. E-iv DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) E . 2 .2 1 . L a t c h o u m y c a n d a n e a n d M a t h u r , 2 0 0 2 : S p e r m P r o d u c t i o n .................................E - 1 1 6 E .2 .2 1 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 1 6 E .2 .2 1 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ....................................................................E - 1 1 6 E .2 .2 1 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ..................................................................... E - 1 1 9 E .2 .2 1 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ...... E - 1 1 9 E .2 .2 1 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ...... E - 1 2 2 E . 2 .2 2 . L i e t a l., 1 9 9 7 : F S H .......................................................................................................................E - 1 2 3 E .2 .2 2 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 2 3 E .2 .2 2 .2 . O u t p u t f o r S e l e c t e d M o d e l : P o w e r .............................................................. E - 1 2 3 E .2 .2 2 .3 . F i g u r e f o r S e l e c t e d M o d e l : P o w e r ................................................................E - 1 2 6 E .2 .2 2 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o w er, U n restricted ..E -1 2 6 E .2 .2 2 .5 . F ig u re fo r A d d itio n al M o d el P resented: P o w er, U n re stricted ...E -1 2 9 E . 2 .2 3 . L i e t a l., 2 0 0 6 : E s t r a d i o l , 3 - D a y ........................................................................................... E - 1 3 0 E .2 .2 3 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 3 0 E .2 .2 3 .2 . O u t p u t f o r S e l e c t e d M o d e l : L i n e a r .............................................................. E - 1 3 0 E .2 .2 3 .3 . F i g u r e f o r S e l e c t e d M o d e l : L i n e a r .............................................................. E - 1 3 3 E . 2 .2 4 . L i e t a l., 2 0 0 6 : P r o g e s t e r o n e , 3 - D a y ................................................................................. E - 1 3 4 E .2 .2 4 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 3 4 E .2 .2 4 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ....................................................................E - 1 3 4 E .2 .2 4 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ..................................................................... E - 1 3 7 E . 2 .2 5 . M a r k o w s k i e t a l., 2 0 0 1 : F R 1 0 R u n O p p o r t u n i t i e s ................................................... E - 1 3 8 E .2 .2 5 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 3 8 E .2 .2 5 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 2 ) .....................................E - 1 3 8 E .2 .2 5 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 2 ) .....................................E - 1 4 1 E . 2 .2 6 . M a r k o w s k i e t a l., 2 0 0 1 : F R 2 R e v o l u t i o n s .....................................................................E - 1 4 2 E .2 .2 6 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 4 2 E .2 .2 6 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ....................................................................E - 1 4 2 E .2 .2 6 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ..................................................................... E - 1 4 5 E .2 .2 6 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o w er, U n restricted ..E -145 E .2 .2 6 .5 . F ig u re fo r A d d itio n al M o d el P resented: P o w er, U n re stricted ...E -148 E . 2 .2 7 . M a r k o w s k i e t a l., 2 0 0 1 : F R 5 R u n O p p o r t u n i t i e s ...................................................... E - 1 4 9 E .2 .2 7 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 4 9 E .2 .2 7 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ....................................................................E - 1 4 9 E .2 .2 7 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ..................................................................... E - 1 5 2 E .2 .2 7 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o w er, U n restricted ..E -1 5 2 E .2 .2 7 .5 . F ig u re fo r A d d itio n al M o d el P resented: P o w er, U n re stricted ...E -155 E . 2 .2 8 . M i e t t i n e n e t a l., 2 0 0 6 : C a r i o g e n i c L e s i o n s , P u p s .....................................................E - 1 5 6 E .2 .2 8 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 5 6 E .2 .2 8 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - L o g i s t i c .............................................. E - 1 5 6 E .2 .2 8 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - L o g i s t i c ................................................E - 1 5 8 E .2 .2 8 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : L o g -L o g istic, U n r e s t r i c t e d ............................................................................................................... E - 1 5 8 This document is a draftfor review purposes only and does not constitute Agency policy. E-v DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) E .2 .2 8 .5 . F ig u re fo r A d d itio n al M o d el P resented: L o g -L o g istic, U n r e s t r i c t e d ............................................................................................................... E - 1 6 0 E . 2 .2 9 . M u r r a y e t a l., 1 9 7 9 : F e r t i l i t y i n F 2 G e n e r a t i o n ...........................................................E - 1 6 1 E .2 .2 9 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 6 1 E .2 .2 9 .2 . O u t p u t f o r S e l e c t e d M o d e l : M u l t i s t a g e , 2 - D e g r e e ............................ E - 1 6 1 E .2 .2 9 .3 . F i g u r e f o r S e l e c t e d M o d e l : M u l t i s t a g e , 2 - D e g r e e ............................. E - 1 6 3 E . 2 .3 0 . N a t i o n a l T o x i c o l o g y P r o g r a m , 1 9 8 2 : T o x i c H e p a t i t i s , M a l e M i c e .............. E - 1 6 4 E .2 .3 0 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 6 4 E .2 .3 0 .2 . O u t p u t f o r S e l e c t e d M o d e l : M u l t i s t a g e , 3 - D e g r e e ............................ E - 1 6 4 E .2 .3 0 .3 . F i g u r e f o r S e l e c t e d M o d e l : M u l t i s t a g e , 3 - D e g r e e ............................. E - 1 6 6 E . 2 .3 1 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : A l v e o l a r M e t a p l a s i a ............................... E - 1 6 7 E .2 .3 1 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 6 7 E .2 .3 1 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - L o g i s t i c .............................................. E - 1 6 7 E .2 .3 1 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - L o g i s t i c ................................................E - 1 6 9 E . 2 .3 2 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : E o s i n o p h i l i c F o c u s , L i v e r ...................E - 1 7 0 E .2 .3 2 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 7 0 E .2 .3 2 .2 . O u t p u t f o r S e l e c t e d M o d e l : P r o b i t .............................................................. E - 1 7 0 E .2 .3 2 .3 . F i g u r e f o r S e l e c t e d M o d e l : P r o b i t ................................................................E - 1 7 2 E . 2 .3 3 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : F a t t y C h a n g e D i f f u s e , L i v e r .............E - 1 7 3 E .2 .3 3 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 7 3 E .2 .3 3 .2 . O u t p u t f o r S e l e c t e d M o d e l : W e i b u l l ..........................................................E - 1 7 3 E .2 .3 3 .3 . F i g u r e f o r S e l e c t e d M o d e l : W e i b u l l ........................................................... E - 1 7 5 E .2.34. N atio n al T o x ico lo g y P ro g ram , 2006: G ingival H y p erp lasia, S quam ous, 2 Y e a r s .......................................................................................................................................................E - 1 7 6 E .2 .3 4 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 7 6 E .2 .3 4 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - L o g i s t i c .............................................. E - 1 7 6 E .2 .3 4 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - L o g i s t i c ................................................E - 1 7 8 E .2 .3 4 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 1 7 8 E .2 .3 4 .5 . F ig u re fo r A d d itio n al M o d el P resented: L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 1 8 0 E .2.35. N atio n al T o x ico lo g y P ro g ram , 2006: H ep ato cy te H y p ertro p h y , 2 Y e a r s ................................................................................................................................................E - 1 8 1 E .2 .3 5 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 8 1 E .2 .3 5 .2 . O u t p u t f o r S e l e c t e d M o d e l : M u l t i s t a g e , 5 - D e g r e e ............................ E - 1 8 1 E .2 .3 5 .3 . F i g u r e f o r S e l e c t e d M o d e l : M u l t i s t a g e , 5 - D e g r e e ............................. E - 1 8 3 E . 2 .3 6 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : N e c r o s i s , L i v e r ...................................E - 1 8 4 E .2 .3 6 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .......................................E - 1 8 4 E .2 .3 6 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - P r o b i t , U n r e s t r i c t e d .....................E - 1 8 4 E .2 .3 6 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - P r o b i t , U n r e s t r i c t e d .....................E - 1 8 6 E . 2 .3 7 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : O v a l C e ll H y p e r p l a s i a ....................E - 1 8 7 E .2 .3 7 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .......................................E - 1 8 7 E .2 .3 7 .2 . O u t p u t f o r S e l e c t e d M o d e l : P r o b i t .............................................................. E - 1 8 7 This document is a draftfor review purposes only and does not constitute Agency policy. E-vi DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) E .2 .3 7 .3 . F i g u r e f o r S e l e c t e d M o d e l : P r o b i t ...............................................................E - 1 8 9 E .2 .3 7 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n t e d : W e i b u l l ............................. E - 1 8 9 E .2 .3 7 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : W e i b u l l ...............................E - 1 9 1 E . 2 .3 8 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : P ig m e n t a t i o n , L i v e r ............................... E - 1 9 2 E .2 .3 8 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 9 2 E .2 .3 8 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - P r o b i t .................................................... E - 1 9 2 E .2 .3 8 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - P r o b i t .................................................... E - 1 9 4 E . 2 .3 9 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : T o x i c H e p a t o p a t h y .................................. E - 1 9 5 E .2 .3 9 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 1 9 5 E .2 .3 9 .2 . O u t p u t f o r S e l e c t e d M o d e l : M u l t i s t a g e , 5 - D e g r e e ............................ E - 1 9 5 E .2 .3 9 .3 . F i g u r e f o r S e l e c t e d M o d e l : M u l t i s t a g e , 5 - D e g r e e ............................. E - 1 9 7 E . 2 .4 0 . O h s a k o e t a l., 2 0 0 1 : A n o - G e n i t a l L e n g t h , P N D 1 2 0 ............................................. E - 1 9 8 E .2 .4 0 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .................................... E - 1 9 8 E .2 .4 0 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ....................................................................E - 1 9 8 E .2 .4 0 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ..................................................................... E - 2 0 1 E .2 .4 0 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ........E - 2 0 1 E .2 .4 0 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ......... E - 2 0 4 E . 2 .4 1 . S e w a ll e t a l., 1 9 9 5 : T 4 I n S e r u m ..........................................................................................E - 2 0 5 E .2 .4 1 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 2 0 5 E .2 .4 1 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ....................................................................E - 2 0 5 E .2 .4 1 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ..................................................................... E - 2 0 8 E .2 .4 1 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ........E - 2 0 8 E .2 .4 1 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ......... E - 2 1 1 E . 2 .4 2 . S h i e t a l., 2 0 0 7 : E s t r a d i o l 1 7 B , P E 9 ...................................................................................E - 2 1 2 E .2 .4 2 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 2 1 2 E .2 .4 2 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) .....................................E - 2 1 2 E .2 .4 2 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) .....................................E - 2 1 5 E . 2 .4 3 . S m i a l o w i c z e t a l., 2 0 0 8 : P F C p e r 1 0 A6 C e l l s ............................................................. E - 2 1 6 E .2 .4 3 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 2 1 6 E .2 .4 3 .2 . O u t p u t f o r S e l e c t e d M o d e l : P o w e r , U n r e s t r i c t e d ...............................E - 2 1 6 E .2 .4 3 .3 . F i g u r e f o r S e l e c t e d M o d e l : P o w e r , U n r e s t r i c t e d ................................ E - 2 1 9 E . 2 .4 4 . S m i a l o w i c z e t a l., 2 0 0 8 : P F C p e r S p l e e n ........................................................................E - 2 2 0 E .2 .4 4 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .......................................E - 2 2 0 E .2 .4 4 .2 . O u t p u t f o r S e l e c t e d M o d e l : P o w e r , U n r e s t r i c t e d ...............................E - 2 2 0 E .2 .4 4 .3 . F i g u r e f o r S e l e c t e d M o d e l : P o w e r , U n r e s t r i c t e d ................................ E - 2 2 3 E . 2 .4 5 . T o t h e t a l., 1 9 7 9 : A m y l o i d o s i s .............................................................................................. E - 2 2 4 E .2 .4 5 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .......................................E - 2 2 4 E .2 .4 5 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - L o g i s t i c .............................................. E - 2 2 4 E .2 .4 5 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - L o g i s t i c ................................................E - 2 2 6 E .2 .4 5 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 2 2 6 E .2 .4 5 .5 . F ig u re fo r A d d itio n al M o d el P resented: L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 2 2 8 This document is a draftfor review purposes only and does not constitute Agency policy. E-vii DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) E . 2 .4 6 . T o t h e t a l., 1 9 7 9 : S k in L e s i o n s .............................................................................................E - 2 2 9 E .2 .4 6 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .....................................E - 2 2 9 E .2 .4 6 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - L o g i s t i c .............................................E - 2 2 9 E .2 .4 6 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - L o g i s t i c .............................................. E - 2 3 1 E .2 .4 6 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 2 3 1 E .2 .4 6 .5 . F ig u re fo r A d d itio n al M o d el P resented: L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 2 3 3 E . 2 .4 7 . V a n B i r g e l e n e t a l., 1 9 9 5 a : H e p a t i c R e t i n o l ................................................................ E - 2 3 4 E .2 .4 7 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 2 3 4 E .2 .4 7 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ...................................E - 2 3 4 E .2 .4 7 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ...................................E - 2 3 7 E .2 .4 7 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o w er, U n restricted ..E -2 3 7 E .2 .4 7 .5 . F ig u re fo r A d d itio n al M o d el P resented: P o w er, U n re stricted ...E -2 4 0 E . 2 .4 8 . V a n B i r g e l e n e t a l., 1 9 9 5 a : H e p a t i c R e t i n o l P a l m i t a t e ......................................... E - 2 4 1 E .2 .4 8 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .................................... E - 2 4 1 E .2 .4 8 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ...................................E - 2 4 1 E .2 .4 8 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ...................................E - 2 4 4 E .2 .4 8 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o w er, U n restricted ..E -2 4 4 E .2 .4 8 .5 . F ig u re fo r A d d itio n al M o d el P resented: P o w er, U n re stricted ...E -2 4 7 E . 2 .4 9 . W h i t e e t a l., 1 9 8 6 : C H 5 0 .......................................................................................................... E - 2 4 8 E .2 .4 9 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 2 4 8 E .2 .4 9 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l .................................................................. E - 2 4 8 E .2 .4 9 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ................................................................... E - 2 5 1 E .2 .4 9 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d .........E - 2 5 1 E .2 .4 9 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d .........E - 2 5 4 E .3 . A D M I N I S T E R E D D O S E B M D S R E S U L T S .................................................................................E - 2 5 5 E .3 .1 . A m i n e t a l., 2 0 0 0 : 0 .2 5 % S a c c h a r i n C o n s u m e d , F e m a l e ..................................... E - 2 5 5 E .3 .1 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 2 5 5 E .3 .1 .2 . O u t p u t f o r S e l e c t e d M o d e l : L i n e a r ............................................................... E - 2 5 5 E .3 .1 .3 . F i g u r e f o r S e l e c t e d M o d e l : L i n e a r ............................................................... E - 2 5 8 E .3 .1 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o w er, U n restricted ..E -2 5 8 E .3.1.5. F ig u re fo r A d d itio n al M o d el P resented: P o w er, U n re stricted ...E -261 E .3 .2 . A m i n e t a l., 2 0 0 0 : 0 .2 5 % S a c c h a r i n P r e f e r e n c e R a t i o , F e m a l e .......................E - 2 6 2 E .3 .2 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 2 6 2 E .3 .2 .2 . O u t p u t f o r S e l e c t e d M o d e l : L i n e a r ............................................................... E - 2 6 2 E .3 .2 .3 . F i g u r e f o r S e l e c t e d M o d e l : L i n e a r ............................................................... E - 2 6 5 E .3 .3 . A m i n e t a l., 2 0 0 0 : 0 .5 0 % S a c c h a r i n C o n s u m e d , F e m a l e ..................................... E - 2 6 6 E .3 .3 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .................................... E - 2 6 6 E .3 .3 .2 . O u t p u t f o r S e l e c t e d M o d e l : L i n e a r ............................................................... E - 2 6 6 E .3 .3 .3 . F i g u r e f o r S e l e c t e d M o d e l : L i n e a r ............................................................... E - 2 6 9 E .3 .3 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o w er, U n restricted ..E -2 6 9 E .3.3.5. F ig u re fo r A d d itio n al M o d el P resented: P o w er, U n re stricted ...E -2 7 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-viii DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) E .3 .4 . A m i n e t a l., 2 0 0 0 : 0 .5 0 % S a c c h a r i n P r e f e r e n c e R a t i o , F e m a l e .......................E - 2 7 3 E .3 .4 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 2 7 3 E .3 .4 .2 . O u t p u t f o r S e l e c t e d M o d e l : L i n e a r ..............................................................E - 2 7 3 E .3 .4 .3 . F i g u r e f o r S e l e c t e d M o d e l : L i n e a r ..............................................................E - 2 7 6 E .3 .4 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o w er, U n restricted ..E -2 7 6 E .3.4.5. F ig u re fo r A d d itio n al M o d el P resented: P o w er, U n re stricted ...E -2 7 9 E .3 .5 . B e l l e t a l., 2 0 0 7 a : B a l a n o - P r e p u t i a l S e p a r a t i o n , P o s t n a t a l D a y 4 9 ..................E - 2 8 0 E .3 .5 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ...................................... E - 2 8 0 E .3.5.2. E -2 8 0 E .3 .5 .3 . O u t p u t f o r S e l e c t e d M o d e l : L o g - L o g i s t i c .............................................. E - 2 8 0 E .3 .5 .4 . F i g u r e f o r S e l e c t e d M o d e l : L o g - L o g i s t i c ............................................... E - 2 8 2 E .3.5.5. O u tp u t fo r A d d itio n al M o d el P resen ted : L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 2 8 3 E .3.5.6. F ig u re fo r A d d itio n al M o d el P resented: L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 2 8 5 E .3 .6 . C a n t o n i e t a l., 1 9 8 1 : U r i n a r y C o p r o p o r h y r i n s , 3 M o n t h s .....................................E - 2 8 6 E .3 .6 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ...................................... E - 2 8 6 E .3 .6 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) .................................... E - 2 8 6 E .3 .6 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ....................................E - 2 8 9 E .3 .6 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o w er, U n restricted ..E -2 8 9 E .3.6.5. F ig u re fo r A d d itio n al M o d el P resented: P o w er, U n re stricted ...E -2 9 2 E .3 .7 . C a n t o n i e t a l., 1 9 8 1 : U r i n a r y P o r p h y r i n s ..........................................................................E - 2 9 3 E .3 .7 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ...................................... E - 2 9 3 E .3 .7 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 2 ) .................................... E - 2 9 3 E .3 .7 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 2 ) ....................................E - 2 9 6 E .3 .8 . C r o f t o n e t a l., 2 0 0 5 : S e r u m , T 4 ..............................................................................................E - 2 9 7 E .3 .8 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ...................................... E - 2 9 7 E .3 .8 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) .................................... E - 2 9 7 E .3 .8 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ....................................E - 3 0 0 E .3 .9 . F r a n c e t a l., 2 0 0 1 : S - D R a t s , R e l a t i v e L i v e r W e i g h t ................................................E - 3 0 1 E .3 .9 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ...................................... E - 3 0 1 E .3 .9 .2 . O u t p u t f o r S e l e c t e d M o d e l : P o w e r .............................................................. E - 3 0 1 E .3 .9 .3 . F i g u r e f o r S e l e c t e d M o d e l : P o w e r ................................................................E - 3 0 4 E .3 .9 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o w er, U n restricted ..E -3 0 4 E .3.9.5. F ig u re fo r A d d itio n al M o d el P resented: P o w er, U n re stricted ...E -3 0 7 E . 3 .1 0 . F r a n c e t a l., 2 0 0 1 : L - E R a t s , R e l a t i v e L i v e r W e i g h t ..............................................E - 3 0 8 E .3 .1 0 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 3 0 8 E .3 .1 0 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ....................................................................E - 3 0 8 E .3 .1 0 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ..................................................................... E - 3 1 1 E .3 .1 0 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ........E - 3 1 1 E .3 .1 0 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ......... E - 3 1 4 E . 3 .1 1 . F r a n c e t a l., 2 0 0 1 : S - D R a t s , R e l a t i v e T h y m u s W e i g h t ........................................ E - 3 1 5 E .3 .1 1 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 3 1 5 This document is a draftfor review purposes only and does not constitute Agency policy. E-ix DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) E .3 .1 1 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ................................... E - 3 1 5 E .3 .1 1 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ................................... E - 3 1 8 E .3 .1 1 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P oly n o m ial, 3 - D e g r e e ....................................................................................................................... E - 3 1 8 E .3 .1 1 .5 . F ig u re fo r A d d itio n al M o d el P resented: P o ly n o m ial, 3 - D e g r e e ....................................................................................................................... E - 3 2 1 E . 3 .1 2 . F r a n c e t a l., 2 0 0 1 : L - E R a t s , R e l a t i v e T h y m u s W e i g h t ........................................ E - 3 2 2 E .3 .1 2 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 3 2 2 E .3 .1 2 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ................................... E - 3 2 2 E .3 .1 2 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ................................... E - 3 2 5 E . 3 .1 3 . F r a n c e t a l., 2 0 0 1 : H / W R a t s , R e l a t i v e T h y m u s W e i g h t .....................................E - 3 2 6 E .3 .1 3 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .....................................E - 3 2 6 E .3 .1 3 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 2 ) ................................... E - 3 2 6 E .3 .1 3 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 2 ) ................................... E - 3 2 9 E .3 .1 3 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n t e d : E x p o n e n t i a l ( M 4 ) ..... E - 3 2 9 E .3 .1 3 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : E x p o n e n t i a l ( M 4 ) .......E - 3 3 2 E . 3 .1 4 . H o j o e t a l., 2 0 0 2 : D R L R e i n f o r c e P e r M i n u t e ............................................................ E - 3 3 3 E .3 .1 4 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 3 3 3 E .3 .1 4 .2 . O u t p u t f o r S e l e c t e d M o d e l : L i n e a r ............................................................. E - 3 3 3 E .3 .1 4 .3 . F i g u r e f o r S e l e c t e d M o d e l : L i n e a r ............................................................. E - 3 3 6 E .3 .1 4 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n t e d : E x p o n e n t i a l ( M 4 ) ..... E - 3 3 6 E .3 .1 4 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : E x p o n e n t i a l ( M 4 ) .......E - 3 3 9 E . 3 .1 5 . H o j o e t a l., 2 0 0 2 : D R L R e s p o n s e P e r M i n u t e ............................................................ E - 3 4 0 E .3 .1 5 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 3 4 0 E .3 .1 5 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ................................... E - 3 4 0 E .3 .1 5 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ................................... E - 3 4 3 E . 3 .1 6 . K a t t a i n e n e t a l., 2 0 0 1 : 3 r d M o l a r E r u p t i o n , F e m a l e ................................................ E - 3 4 4 E .3 .1 6 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 3 4 4 E .3 .1 6 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - L o g i s t i c ............................................. E - 3 4 4 E .3 .1 6 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - L o g i s t i c ...............................................E - 3 4 6 E .3 .1 6 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 3 4 6 E .3 .1 6 .5 . F ig u re fo r A d d itio n al M o d el P resented: L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 3 4 8 E . 3 .1 7 . K a t t a i n e n e t a l., 2 0 0 1 : 3 r d M o l a r L e n g t h , F e m a l e .................................................... E - 3 4 9 E .3 .1 7 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 3 4 9 E .3 .1 7 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ...................................................................E - 3 4 9 E .3 .1 7 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l .................................................................... E - 3 5 2 E .3 .1 7 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d .......E - 3 5 2 E .3 .1 7 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ........ E - 3 5 5 E . 3 .1 8 . K e l l e r e t a l., 2 0 0 7 : M i s s i n g M a n d i b u l a r M o l a r s , C B A J ..................................... E - 3 5 6 E .3 .1 8 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 3 5 6 E .3 .1 8 .2 . O u t p u t f o r S e l e c t e d M o d e l : M u l t i s t a g e , 1 - D e g r e e ...........................E - 3 5 6 This document is a draftfor review purposes only and does not constitute Agency policy. E-x DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) E .3 .1 8 .3 . F i g u r e f o r S e l e c t e d M o d e l : M u l t i s t a g e , 1 - D e g r e e .............................E - 3 5 8 E . 3 .1 9 . K o c i b a e t a l., 1 9 7 8 : U r i n a r y C o p r o p o r p h y r i n , F e m a l e s ........................................ E - 3 5 9 E .3 .1 9 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .................................... E - 3 5 9 E .3 .1 9 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ................................... E - 3 5 9 E .3 .1 9 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ...................................E - 3 6 2 E . 3 .2 0 . K o c i b a e t a l., 1 9 7 8 : U r o p o r p h y r i n p e r C r e a ti n i n e , F e m a l e .................................E - 3 6 3 E .3 .2 0 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .................................... E - 3 6 3 E .3 .2 0 .2 . O u t p u t f o r S e l e c t e d M o d e l : L i n e a r .............................................................. E - 3 6 3 E .3 .2 0 .3 . F i g u r e f o r S e l e c t e d M o d e l : L i n e a r .............................................................. E - 3 6 6 E . 3 .2 1 . L a t c h o u m y c a n d a n e a n d M a t h u r , 2 0 0 2 : S p e r m P r o d u c t i o n .................................E - 3 6 7 E .3 .2 1 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .................................... E - 3 6 7 E .3 .2 1 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ....................................................................E - 3 6 7 E .3 .2 1 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ..................................................................... E - 3 7 0 E .3 .2 1 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ........E - 3 7 0 E .3 .2 1 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ......... E - 3 7 3 E . 3 .2 2 . L i e t a l., 1 9 9 7 : F S H .......................................................................................................................E - 3 7 4 E .3 .2 2 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .................................... E - 3 7 4 E .3 .2 2 .2 . O u t p u t f o r S e l e c t e d M o d e l : P o w e r .............................................................. E - 3 7 4 E .3 .2 2 .3 . F i g u r e f o r S e l e c t e d M o d e l : P o w e r ................................................................E - 3 7 7 E .3 .2 2 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o w er, U n restricted ..E -3 7 7 E .3 .2 2 .5 . F ig u re fo r A d d itio n al M o d el P resented: P o w er, U n re stricted ...E -3 8 0 E . 3 .2 3 . L i e t a l., 2 0 0 6 : E s t r a d i o l , 3 - D a y ........................................................................................... E - 3 8 1 E .3 .2 3 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .................................... E - 3 8 1 E .3 .2 3 .2 . O u t p u t f o r S e l e c t e d M o d e l : L i n e a r .............................................................. E - 3 8 1 E .3 .2 3 .3 . F i g u r e f o r S e l e c t e d M o d e l : L i n e a r .............................................................. E - 3 8 4 E . 3 .2 4 . L i e t a l., 2 0 0 6 : P r o g e s t e r o n e , 3 - D a y ................................................................................. E - 3 8 5 E .3 .2 4 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .................................... E - 3 8 5 E .3 .2 4 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) .....................................E - 3 8 5 E .3 .2 4 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) .....................................E - 3 8 8 E .3 .2 4 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ........E - 3 8 8 E .3 .2 4 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ......... E - 3 9 1 E . 3 .2 5 . M a r k o w s k i e t a l., 2 0 0 1 : F R 1 0 R u n O p p o r t u n i t i e s ................................................... E - 3 9 2 E .3 .2 5 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .................................... E - 3 9 2 E .3 .2 5 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 2 ) .....................................E - 3 9 2 E .3 .2 5 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 2 ) .....................................E - 3 9 5 E . 3 .2 6 . M a r k o w s k i e t a l., 2 0 0 1 : F R 2 R e v o l u t i o n s .....................................................................E - 3 9 6 E .3 .2 6 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .................................... E - 3 9 6 E .3 .2 6 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ....................................................................E - 3 9 6 E .3 .2 6 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ..................................................................... E - 3 9 9 E .3 .2 6 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o w er, U n restricted ..E -3 9 9 E .3 .2 6 .5 . F ig u re fo r A d d itio n al M o d el P resented: P o w er, U n re stricted ...E -4 0 2 E . 3 .2 7 . M a r k o w s k i e t a l., 2 0 0 1 : F R 5 R u n O p p o r t u n i t i e s ...................................................... E - 4 0 3 E .3 .2 7 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .................................... E - 4 0 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-xi DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) E .3 .2 7 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ....................................................................E - 4 0 3 E .3 .2 7 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ..................................................................... E - 4 0 6 E .3 .2 7 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o w er, U n restricted ..E -4 0 7 E .3 .2 7 .5 . F ig u re fo r A d d itio n al M o d el P resented: P o w er, U n re stricted ...E -4 0 9 E . 3 .2 8 . M i e t t i n e n e t a l., 2 0 0 6 : C a r i o g e n i c L e s i o n s , P u p s .....................................................E - 4 1 0 E .3 .2 8 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 4 1 0 E .3 .2 8 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - L o g i s t i c .............................................. E - 4 1 0 E .3 .2 8 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - L o g i s t i c ................................................E - 4 1 2 E .3 .2 8 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 4 1 2 E .3 .2 8 .5 . F ig u re fo r A d d itio n al M o d el P resented: L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 4 1 4 E . 3 .2 9 . M u r r a y e t a l., 1 9 7 9 : F e r t i l i t y i n F 2 G e n e r a t i o n .......................................................... E - 4 1 5 E .3 .2 9 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 4 1 5 E .3 .2 9 .2 . O u t p u t f o r S e l e c t e d M o d e l : M u l t i s t a g e , 2 - D e g r e e ............................ E - 4 1 5 E .3 .2 9 .3 . F i g u r e f o r S e l e c t e d M o d e l : M u l t i s t a g e , 2 - D e g r e e ............................. E - 4 1 7 E . 3 .3 0 . N a t i o n a l T o x i c o l o g y P r o g r a m , 1 9 8 2 : T o x i c H e p a t i t i s , M a l e M i c e .............. E - 4 1 8 E .3 .3 0 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 4 1 8 E .3 .3 0 .2 . O u t p u t f o r S e l e c t e d M o d e l : M u l t i s t a g e , 3 - D e g r e e ............................ E - 4 1 8 E .3 .3 0 .3 . F i g u r e f o r S e l e c t e d M o d e l : M u l t i s t a g e , 3 - D e g r e e ............................. E - 4 2 0 E .3.31. N atio n al T o x ico lo g y P ro g ram , 2006: A l v e o l a r M e t a p l a s i a ..................E - 4 2 1 E .3 .3 1 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .......................................E - 4 2 1 E .3 .3 1 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - L o g i s t i c .............................................. E - 4 2 1 E .3 .3 1 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - L o g i s t i c ................................................E - 4 2 3 E .3 .3 1 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 4 2 3 E .3 .3 1 .5 . F ig u re fo r A d d itio n al M o d el P resented: L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 4 2 5 E .3.32. N atio n al T o x ico lo g y P ro g ram , 2006: E o s i n o p h i l i c F o c u s , L i v e r .....E - 4 2 6 E .3 .3 2 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .......................................E - 4 2 6 E .3 .3 2 .2 . O u t p u t f o r S e l e c t e d M o d e l : P r o b i t .............................................................. E - 4 2 6 E .3 .3 2 .3 . F i g u r e f o r S e l e c t e d M o d e l : P r o b i t ................................................................E - 4 2 8 E .3.33. N atio n al T o x ico lo g y P ro g ram , 2006: F atty C hange D iffuse, L iv e rE -429 . E .3 .3 3 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .......................................E - 4 2 9 E .3 .3 3 .2 . O u t p u t f o r S e l e c t e d M o d e l : W e i b u l l ..........................................................E - 4 2 9 E .3 .3 3 .3 . F i g u r e f o r S e l e c t e d M o d e l : W e i b u l l ........................................................... E - 4 3 1 E .3.34. N atio n al T o x ico lo g y P ro g ram , 2006: G ingival H y p erp lasia, S quam ous, 2 Y e a r s .......................................................................................................................................................E - 4 3 2 E .3 .3 4 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .......................................E - 4 3 2 E .3 .3 4 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - L o g i s t i c .............................................. E - 4 3 2 E .3 .3 4 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - L o g i s t i c ................................................E - 4 3 4 E .3 .3 4 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 4 3 4 This document is a draftfor review purposes only and does not constitute Agency policy. E-xii DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) E .3 .3 4 .5 . F ig u re fo r A d d itio n al M o d el P resented: L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 4 3 6 E .3.35. N atio n al T o x ico lo g y P ro g ram , 2006: H ep ato cy te H y p ertro p h y , 2 Y e a r s ................................................................................................................................................E - 4 3 7 E .3 .3 5 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 4 3 7 E .3 .3 5 .2 . O u t p u t f o r S e l e c t e d M o d e l : M u l t i s t a g e , 5 - D e g r e e ............................ E - 4 3 7 E .3 .3 5 .3 . F i g u r e f o r S e l e c t e d M o d e l : M u l t i s t a g e , 5 - D e g r e e ............................. E - 4 3 9 E . 3 .3 6 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : N e c r o s i s , L i v e r ...........................................E - 4 4 0 E .3 .3 6 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 4 4 0 E .3 .3 6 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - P r o b i t , U n r e s t r i c t e d .....................E - 4 4 0 E .3 .3 6 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - P r o b i t , U n r e s t r i c t e d .....................E - 4 4 2 E . 3 .3 7 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : O v a l C e ll H y p e r p l a s i a ........................... E - 4 4 3 E .3 .3 7 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 4 4 3 E .3 .3 7 .2 . O u t p u t f o r S e l e c t e d M o d e l : P r o b i t .............................................................. E - 4 4 3 E .3 .3 7 .3 . F i g u r e f o r S e l e c t e d M o d e l : P r o b i t ................................................................E - 4 4 5 E .3 .3 7 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n t e d : W e i b u l l ............................. E - 4 4 5 E .3 .3 7 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : W e i b u l l ...............................E - 4 4 7 E . 3 .3 8 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : P ig m e n t a t i o n , L i v e r ............................... E - 4 4 8 E .3 .3 8 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 4 4 8 E .3 .3 8 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - P r o b i t .................................................... E - 4 4 8 E .3 .3 8 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - P r o b i t .................................................... E - 4 5 0 E . 3 .3 9 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : T o x i c H e p a t o p a t h y ..................................E - 4 5 1 E .3 .3 9 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .......................................E - 4 5 1 E .3 .3 9 .2 . O u t p u t f o r S e l e c t e d M o d e l : M u l t i s t a g e , 5 - D e g r e e ............................ E - 4 5 1 E .3 .3 9 .3 . F i g u r e f o r S e l e c t e d M o d e l : M u l t i s t a g e , 5 - D e g r e e ............................. E - 4 5 3 E . 3 .4 0 . O h s a k o e t a l., 2 0 0 1 : A n o - G e n i t a l L e n g t h , P N D 1 2 0 .......................................E - 4 5 4 E .3 .4 0 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .......................................E - 4 5 4 E .3 .4 0 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ....................................................................E - 4 5 4 E .3 .4 0 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ..................................................................... E - 4 5 7 E .3 .4 0 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ........E - 4 5 7 E .3 .4 0 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ......... E - 4 6 0 E . 3 .4 1 . S e w a ll e t a l., 1 9 9 5 : T 4 I n S e r u m ..........................................................................................E - 4 6 1 E .3 .4 1 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 4 6 1 E .3 .4 1 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ....................................................................E - 4 6 1 E .3 .4 1 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ..................................................................... E - 4 6 4 E .3 .4 1 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ........E - 4 6 4 E .3 .4 1 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ......... E - 4 6 7 E . 3 .4 2 . S h i e t a l., 2 0 0 7 : E s t r a d i o l 1 7 B , P E 9 ...................................................................................E - 4 6 8 E .3 .4 2 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 4 6 8 E .3 .4 2 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) .....................................E - 4 6 8 E .3 .4 2 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) .....................................E - 4 7 1 E . 3 .4 3 . S m i a l o w i c z e t a l., 2 0 0 8 : P F C p e r 1 0 A6 C e l l s ............................................................. E - 4 7 2 E .3 .4 3 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 4 7 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-xiii DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) E .3 .4 3 .2 . O u t p u t f o r S e l e c t e d M o d e l : P o w e r , U n r e s t r i c t e d ...............................E - 4 7 2 E .3 .4 3 .3 . F i g u r e f o r S e l e c t e d M o d e l : P o w e r , U n r e s t r i c t e d ................................ E - 4 7 5 E .3 .4 3 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n te d : P o w e r ..................................E - 4 7 5 E .3 .4 3 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : P o w e r ..................................E - 4 7 8 E . 3 .4 4 . S m i a l o w i c z e t a l., 2 0 0 8 : P F C p e r S p l e e n ........................................................................E - 4 7 9 E .3 .4 4 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 4 7 9 E .3 .4 4 .2 . O u t p u t f o r S e l e c t e d M o d e l : P o w e r , U n r e s t r i c t e d ...............................E - 4 7 9 E .3 .4 4 .3 . F i g u r e f o r S e l e c t e d M o d e l : P o w e r , U n r e s t r i c t e d ................................ E - 4 8 2 E .3 .4 4 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n te d : P o w e r ..................................E - 4 8 2 E .3 .4 4 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : P o w e r ..................................E - 4 8 5 E . 3 .4 5 . T o t h e t a l., 1 9 7 9 : A m y l o i d o s i s .............................................................................................. E - 4 8 6 E .3 .4 5 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ..................................... E - 4 8 6 E .3 .4 5 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g - L o g i s t i c .............................................. E - 4 8 6 E .3 .4 5 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g - L o g i s t i c ................................................E - 4 8 8 E .3 .4 5 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 4 8 8 E .3 .4 5 .5 . F ig u re fo r A d d itio n al M o d el P resented: L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 4 9 0 E . 3 .4 6 . T o t h e t a l., 1 9 7 9 : S k in L e s i o n s .............................................................................................E - 4 9 1 E .3 .4 6 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ...................................... E - 4 9 1 E .3 .4 6 .2 . O u t p u t f o r S e l e c t e d M o d e l : L o g i s t i c ..........................................................E - 4 9 1 E .3 .4 6 .3 . F i g u r e f o r S e l e c t e d M o d e l : L o g i s t i c ........................................................... E - 4 9 3 E .3 .4 6 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 4 9 3 E .3 .4 6 .5 . F ig u re fo r A d d itio n al M o d el P resented: L o g -L o g istic, U n r e s t r i c t e d ................................................................................................................ E - 4 9 5 E . 3 .4 7 . V a n B i r g e l e n e t a l., 1 9 9 5 a : H e p a t i c R e t i n o l ................................................................ E - 4 9 6 E .3 .4 7 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ...................................... E - 4 9 6 E .3 .4 7 .2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ..................................E - 4 9 6 E .3 .4 7 .3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ..................................E - 4 9 9 E .3 .4 7 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o w er, U n restricted ..E -4 9 9 E .3 .4 7 .5 . F ig u re fo r A d d itio n al M o d el P resented: P o w er, U n re stricted ...E -5 0 2 E . 3 .4 8 . V a n B i r g e l e n e t a l., 1 9 9 5 a : H e p a t i c R e t i n o l P a l m i t a t e ......................................... E - 5 0 3 E .3 .4 8 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ...................................... E - 5 0 3 E .3 .4 8 .2 . O u t p u t f o r S e l e c t e d M o d e l : L i n e a r .............................................................. E - 5 0 3 E .3 .4 8 .3 . F i g u r e f o r S e l e c t e d M o d e l : L i n e a r .............................................................. E - 5 0 6 E .3 .4 8 .4 . O u tp u t fo r A d d itio n al M o d el P resen ted : P o w er, U n restricted ..E -5 0 6 E .3 .4 8 .5 . F ig u re fo r A d d itio n al M o d el P resented: P o w er, U n re stricted ...E -5 0 9 E . 3 .4 9 . W h i t e e t a l., 1 9 8 6 : C H 5 0 ......................................................................................................... E - 5 1 0 E .3 .4 9 .1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ...................................... E - 5 1 0 E .3 .4 9 .2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ....................................................................E - 5 1 0 E .3 .4 9 .3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ..................................................................... E - 5 1 3 E .3 .4 9 .4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d ....... E - 5 1 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-xiv DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) E .3 .4 9 .5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n te d : H i l l , U n r e s t r i c t e d .........E - 5 1 6 E .4 . R E F E R E N C E S ................................................................................................................................................... E - 5 1 7 This document is a draftfor review purposes only and does not constitute Agency policy. E-xv DRAFT--DO NOT CITE OR QUOTE 1 A P P E N D IX E . N O N C A N C E R B E N C H M A R K D O S E M O D E L IN G 2 3 4 E .1. B M D S IN P U T T A B L E S 5 E .1 .1 . A m in et al. (2000) Endpoint c Saccharin consumed, female rats (0.25%) (ml saccharin solution/1 0 0 g body weight) c Saccharin consumed, female rats (0.50%) (ml saccharin solution/1 0 0 g body weight) c Saccharin preference ratio, female rats (0.25%) (ratio o f saccharin solution consumed to total fluid consumed) d Saccharin preference ratio, female rats (0.50%) (ratio o f saccharin solution consumed to total fluid consumed) d Administered Dose (ng/kg-day) 0 25 a 100 Internal Dose (ng/kg blood) b 0 3.38 10.57 (n = 10) (n = 10) (n = 10) 31.67 6.53 24.60 3.79 10.70 1.68 22.40 5.05 11.38 2.42 4.54 1.05 82.14 4.22 58.12 10.71 54.87 6.17 72.73 7.79 44.48 10.39 33.77 7.79 aLOAEL identified. bFrom the Em ond PB PK m odel described in 3.3. c Values are the m ean SE. D ata obtained from Figure 2 in A m in et al. 2000. d V alues are the ratio SE. D ata obtained from Figure 3 in A m in et al. 2000. 6 7 8 E .1 .2 . B ell et al. (2007) 0 Endpoint Proportion of male rat pups that had not undergone balano-preputial separation on PND 49 c 0 (n = 30) 1/30 (3%) Administered Dose (ng/kg-day) 2.4 a 8 Internal Dose (ng/kg blood) b 2.20 5.14 (n = 30) (n = 30) 46 18.41 (n = 30) 5/30 (17%) 6/30 (20%) 15/30 (50%) aLOAEL identified. b From the Em ond PB PK m odel described in 3.3. c D ata obtained from Figure 2 in Bell et al. 2007. This document is a draftfor review purposes only and does not constitute Agency policy. E-1 DRAFT--DO NOT CITE OR QUOTE 1 E.1.3. Cantoni et al. (1981) 0 Endpoint Urinary coproporphyrins in female rats (pg coproporphyrin methyl ester/24 hr) at 3 m onths c Urinary porphyrins in rats (nmol/24 hr) after 45 weeks c 0 (n = 4) 0.74 0.17 2.27 0.49 Administered Dose (ng/kg-day) 1.43 a 14.3 Internal Dose (ng/kg blood) b 1.85 8.84 (n = 4) (n = 3) 1.81 0.42 d 2.73 0.75 e 143 50.05 (n = 3) 3.00 1.30 e 5.55 0.85 d 7.62 1.79 d 196.89 63.14 e aLOAEL identifed. bFrom the Em ond PB PK m odel described in 3.3. cValues are the m ean SE. D ata for urinary coproporphyrins and urinary porphyrins obtained from Figure 1 and Table 1, respectively, in C antoni et al. 1981. d Statistically significant as com pared to control (p < 0.05). e Statistically significant as com pared to control (p < 0.01). 2 3 4 E .1 .4 . C ro fto n et al. (2005) Endpoint Serum T4 in female rats (% control) d Administered Dose (ng/kg-day) 0 0.1 3 10 30 a 100 b 300 1,000 3,000 10,000 Internal Dose (ng/kg blood) c 0 0.02 0.49 1.38 3.46 9.26 23.07 65.65 180.90 583.48 (n = 14) (n = 6) (n = 12) (n = 6) (n = 6) (n = 6) (n = 6) (n = 6) (n = 6) (n = 4) 100.00 96.27 98.57 99.76 93.32 70.94 62.52 52.68 54.66 49.15 15.44 14.98 18.11 19.04 12.11 12.74 14.75 22.73 19.71 11.15 aNOAEL identifed. bLOAEL identifed. cFrom the Em ond PB PK m odel described in 3.3. d Values are the m ean SD. D ata were obtained from a Crofton et al. supplem ental file, available at http://ehp.niehs.nih.gov/docs/2005/8195/supplem ental.pdf. 5 This document is a draftfor review purposes only and does not constitute Agency policy. E-2 DRAFT--DO NOT CITE OR QUOTE 1 E.1.5. DeCaprio et al. (1986) Administered Dose (ng/kg-day) 0 0.12 0.61 a 4.9 b Internal Dose (ng/kg blood) c Endpoint n/a (n = 10) n/a (n = 10) n/a (n = 11) n/a (n = 10) Absolute kidney w eight (g), males d 5.49 0.17 5.14 0.12 4.71 0.12 4.3 0.15 f Absolute thym us weight (g), males d 0.56 0.050 0.45 0.022 0.44 0.034 0.35 0.167 g Body w eight (g), m ales e 713 15 682 16 651 19 603 20 f Relative brain weight, males d 0.54 0.015 0.56 0.016 0.6 0.016 0.65 0.016 f Relative liver weight, males d 4.54 0.23 4.1 0.14 5.36 0.61 5.630.29 f Relative thymus weight, males d 0.078 0.006 0.066 0.003 0.068 0.004 0.060.003 f Administered Dose (ng/kg-day) 0 0.12 0.68 4.86 Internal Dose (ng/kg blood) c Endpoint 0 (n = 8) n/a (n = 10) n/a (n = 9) n/a (n = 10) Body w eight (g), fem ales e 602 12 583 22 570 22 531 14f Relative liver w eight, fem ales d 4.3 0.26 4.49 0.35 4.27 0.16 5.54 0.43 f 26 n/a (n = 4) - 433 38h - - 31 n/a (n = 4) 351 49h - aNOAEL identified. bLOAEL identified. cInternal dose not calculated using the Em ond PBPK (guinea pigs). dO rgan w eight data in guinea pigs obtained from Table 2 o f D eCaprio et al. 1986. Values are the m ean SE. R elative organs w eights w ere calculated as organ w eight (g) / body w eight (g) X 100. eBody w eight data in guinea pigs obtained from Table 1 o f D eCaprio et al. 1986. V alues are the m ean SE. f Statistically significant as com pared to control (p < 0.05). g Statistically significant as com pared to control (p < 0 .0 1 ). h Statistically significant as com pared to control (p < 0.001). 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-3 DRAFT--DO NOT CITE OR QUOTE 1 E.1.6. Franc et al. (2001) Endpoint S-D rats, relative liver w eight d L-E rats, relative liver w eight d S-D rats, relative thym us w eight d L-E rats, relative thym us w eight d H/W rats, relative thym us w eight d 0 0 (n = 8) 100.0 5.0 100.0 3.5 100.2 29.4 103.4 19.3 101.2 12.7 Administered Dose (ng/kg-day) 10 a 30 b Internal Dose (ng/kg blood) c 6.59 14.48 (n = 8) (n = 8) 108.1 6.0 e 116.8 9.2 e 106.3 6.3 116.8 3.2 e 91.2 17.0 51.4 15.4 e 95.4 24.9 38.7 17.0 e 97.5 11.7.0 71.0 8.5 e 100 36.43 (n = 8) 155.3 10.9 e 122.2 7.0 e 22.8 10.6 e 35.0 27.6 e 49.3 15.4 e aNOAEL identified. bLOAEL identified. cFrom the Em ond PB PK m odel described in 3.3. d V alues are the m ean SE. D ata obtained from Figure 5 in Franc et al. 2001. e Statistically significant as com pared to control (p < 0.05). 2 3 4 E .1 .7 . H o jo e t al. (2002) 0 Endpoint DRL reinforcements/min, rat litters c DRL responses/min, rat litters c 0 (n = 5) -0.814 0.45 18.44 7.99 Administered Dose (ng/kg-day) 20 a 60 Internal Dose (ng/kg blood) b 1.62 4.17 (n = 5) (n = 6) -0.364 0.82 0.374 0.54 -0.99 10.96 -4.52 7.19 180 10.70 (n = 5) -0.163 0.44 -0.41 15.23 aLOAEL identified. bFrom the Em ond PB PK m odel described in 3.3. c DRL = differential reinforcem ent o f low rate. Values are the m ean SD. D ata obtained from Table 5 in Hojo et al. 2 0 0 2 . 5 This document is a draftfor review purposes only and does not constitute Agency policy. E-4 DRAFT--DO NOT CITE OR QUOTE 1 E.1.8. Kattainen et al. (2001) 0 Endpoint 3rd m olar m esio-distal length in female rat offspring (molar development) (mm) c Proportion of female rat offspring w ithout 3rd m olar eruption on PND 35 d 0 (n = 16) 1.86 0.017 1/16 (10%) Administered Dose (ng/kg-day) 30 a 100 300 Internal Dose (ng/kg blood) b 2.23 6.25 16.08 (n = 17) (n = 15) (n = 12) 1,000 46.86 (n = 19) 1.58 0.045 e 1.6 0.069 e 1.5 0.064 e 1.35 0.118 e 3/17 (20%) 4/15 (30%) 6/12 (50% ) e 13/19 (70%) e aLOAEL identified. bFrom the Em ond PB PK m odel described in 3.3. c Values are the m ean SE. D ata were obtained from Figure 3 in K attainen et al. 2001. d D ata w ere obtained from Figure 2 in K attainen et al. 2001. e Statistically significant as com pared to control (p < 0.05). 2 3 4 E .1 .9 . K e lle r et al. (2007, 2 0 0 8 a, b) Endpoint Frequency o f m issing 3rd m andibular m olars in C BA J mice c Administered Dose (ng/kg-day) 0 10 a 100 1,000 Internal Dose (ng/kg blood) b 0 0.54 4.29 34.06 0/29 (0%) 2/23 (10%) 6/29 (20%) 30/30 (100%) aLOAEL identified. bFrom the Em ond PB PK m odel described in 3.3. c D ata obtained from Table 1 in K eller et al. 2007. 5 This document is a draftfor review purposes only and does not constitute Agency policy. E-5 DRAFT--DO NOT CITE OR QUOTE 1 E.1.10. Kociba et al. (1978) Endpoint Urinary coproporphyrin (pg/48 h), female rats d pg uroporphyrin per mg creatinine, female rats d 0 0 (n = 5) 9.8 1.3 Administered Dose (ng/kg-day) 1 a 10 b Internal Dose (ng/kg blood) c 1.55 7.15 (n = 5) (n = 5) 8.6 2 16.4 4.7 e 100 38.56 (n = 5) 17.4 4 e 0.157 0.05 0.143 0.037 0.181 0.053 0.296 0.074 e aNOAEL identified. bLOAEL identified. cFrom the Em ond PB PK m odel described in 3.3. d V alues are the m ean SD. D ata obtained from Table 2 in K ociba et al. 1978. e Statistically significant as com pared to control (p < 0.05). 2 3 4 E .1 .1 1 . L a tc h o u m y c a n d a n e a n d M a th u r (2002) Endpoint D aily sperm production (*106) in adult male rats (mg) c 0 0 (n = 6) 22.19 2.67 Administered Dose (ng/kg-day) 1 a 10 Internal Dose (ng/kg blood) b 0.78 4.65 (n = 6) (n = 6) 15.67 2.65 d 13.65 2.19 d 100 27.27 (n = 6) 13.1 3.16 d aLOAEL identified. bFrom the Em ond PB PK m odel described in 3.3. c Values are the m ean SD. D ata obtained from Table 1 in Latchoumycandane and M athur 2002. d Statistically significant as com pared to control (p < 0.05). 5 6 This document is a draftfor review purposes only and does not constitute Agency policy. E-6 DRAFT--DO NOT CITE OR QUOTE 1 E.1.12. Li et al. (1997) Endpoint Serum FSH (ng/ml) in female rats d Administered Dose (ng/kg-day) 0 3 a 10 b 30 100 300 1,000 3,000 10,000 30,000 Internal Dose (ng/kg blood) c 0 0.27 0.80 2.1 5.87 15 43.33 119.94 385.96 1171.90 (n = 10) (n =10) (n = 10) (n = 10) (n = 10) (n = 10) (n = 10) (n = 10) (n = 10) (n = 10) 23.86 9.38 22.16 15.34 85.23 29.83 73.30 126.14 132.10 116.76 304.26 346.88 455.11 15.34 50.28 36.65 16.19 48.58 47.73 90.34 aNOAEL identified. bLOAEL identified. cFrom the Em ond PBPK m odel described in 3.3. d V alues are the m ean SE. D ata obtained from Figure 3 in Li et al. 1997. 2 3 4 E .1 .1 3 . L i et al. (2006) Endpoint Serum estradiol/(pg-m l)-1 in fem ale m ice (1~3d) c Serum progesterone (ng-ml)"1 in fem ale mice (1~3d) c 0 0 (n = 10) Administered Dose (ng/kg-day) 2 a 50 Internal Dose (ng/kg blood) b 0.16 2.84 (n = 10) (n = 10) 10.17 3.85 19.91 6.31 24.72 4.60 61.74 3.51 30.56 12.80 d 16.93 10.53 100 5.12 (n = 10) 18.09 5.57 11.36 13.83 aLOAEL identified. bFrom the Emond PBPK m odel described in 3.3. c V alues are the m ean SE. Data obtained from Figures 3 (estradiol) and 4 (progesterone) in L i et al. 2006. d Statistically sign ifican t as com pared to control (p < 0.01). 5 This document is a draftfor review purposes only and does not constitute Agency policy. E-7 DRAFT--DO NOT CITE OR QUOTE 1 E.1.14. Markowski et al. (2001) 0 Endpoint FR10 earned run opportunities, adult female offspring c FR2 total revolutions, adult female offspring c FR5 earned run opportunities, adult female offspring c 0 (n = 7) 13.29 8.65 119.29 69.9 26.14 12.28 Administered Dose (ng/kg-day) 20 a 60 Internal Dose (ng/kg blood) b 1.56 4.03 (n = 4) (n = 6) 11.25 5.56 5.75 3.53 108.5 61 56.5 31.21 23.5 7.04 12.8 6.17 180 10.32 (n = 7) 7 6.01 68.14 33.23 13.14 7.14 aLOAEL identified. bFrom the Em ond PB PK m odel described in 3.3. c Values are the m ean SD. D ata obtained from Table 3 in M arkow ski et al. 2001. 2 3 4 E .1 .1 5 . M ie ttin e n et al. (2006) Endpoint Cariogenic lesions in rat pups c Administered Dose (ng/kg-day) 0 30 a 100 300 Internal Dose (ng/kg blood) b 0 2.22 6.23 16.01 (n = 42) II (n = 15) (n = 24) 25/42 (60%) 23/29 (79%) d 19/25 (76%) 20/24 (83%) d 1,000 46.64 (n = 32) 29/32 (91%) d aLOAEL identified. bFrom the Em ond PB PK m odel described in 3.3. c D ata obtained from Table 2 in M iettinen et al. 2006. d Statistically significant as com pared to control (p < 0.05). 5 6 This document is a draftfor review purposes only and does not constitute Agency policy. E-8 DRAFT--DO NOT CITE OR QUOTE 1 E.1.16. National Toxicology Program (1982) 0 Endpoint Numbers o f male m ice w ith toxic hepatitis c I ' ll 0 1/73 (1.4% ) Administered Dose (ng/kg-day) 1.43 a 7.14 Internal Dose (ng/kg blood) b 0.77 2.27 (n = 49) (n = 49) 5/49 (10%) 3/49 (6.1% ) aLOAEL identified. bFrom the Emond PBPK m odel described in 3.3. cD ata obtained from Table 11 in N T P 1982. 2 71.4 11.24 (n = 50) 44/50 (88%) This document is a draftfor review purposes only and does not constitute Agency policy. E-9 DRAFT--DO NOT CITE OR QUOTE 1 E.1.17. National Toxicology Program (2006) Endpoint e Gingival squamous hyperplasia Liver, hepatocyte hypertrophy Heart, cardiomyopathy Liver, eosinophilic focus, multiple Liver, fatty change, diffuse Liver, necrosis Liver, pigmentation Liver, toxic hepatopathy Oval cell hyperplasia Lung, alveolar to bronchiolar epithelial metaplasia (Alveolar epithelium, metaplasia, bronchiolar) 0 0 (n = 10) 1/53 (2%) 0/53 (0%) 10/53 (19% ) 3/53 (6%) 0/53 (0%) 1/53 (2%) 4/53 (8%) 0/53 (0%) 0/53 (0%) 2/53 (4%) Administered Dose (ng/kg-day) 2.14 a 7.14 15.7 32.9 Internal Dose (ng/kg blood) b 2.56 5.69 9.79 16.57 (n = 10) (n = 10) (n = 10) (n = 10) 7/54 (13% )d 14/53 (26% )c 13/53 (25% )c 15/53 (28% )c 19/54 (40% )c' 19/53 (40% )c 42/53 (80% )c 41/53 (80% )c 12/54 (22% ) 22/53c (42% ) 25/52c (48% ) 32/53c (60% ) 8/54 (15% ) 14/53 (26% ) 17/53 (32% ) 22/53 (42% ) 2/54 (4%) 12/53c (23% ) 17/53c (32% ) 30/53c (57% ) 4/54 (7%) 4/53 (8%) 8/53d (15% ) 10/53c (19% ) 9/54 (17% ) 34/53c (64% ) 48/53c (91% ) 52/53c (98% ) 2/54 (4%) 8/53 (15% ) 30/53 (57% ) 45/50 (85% ) 4/54 (10% )d 3/53 (10% ) 20/53 (40% )c 38/53 (70% )d 19/54 c (35% ) 33/53c (62% ) 35/52c (67% ) 45/53c (85% ) 71.4 29.70 (n = 10) 16/53 (30% )c 52/53 (100% )c 36/52c (69% ) 42/53 (79% ) 48/53c (91% ) 17/53c (32% ) 53/53c (100% ) 53/53 (100% ) 53/53 (100% )c 46/52c (89% ) aLOAEL identified. bFrom the Emond PBPK m odel described in 3.3. c Statistically sign ifican t as com pared to control (p < 0.01). d Statistically sign ifican t as com pared to control (p < 0.05). e Data are for fem ale rats in 2-year gavage study. Data for all endpoints obtained from Table A 5b in N TP 2006. 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-10 DRAFT--DO NOT CITE OR QUOTE 1 E.1.18. Ohsako et al. (2001) 0 Endpoint 0 (n = 12) Anogenital distance (mm) in male 28.91 0.90 rat offspring, PND 120 d Administered Dose (ng/kg-day) 12.5 a 50 b 200 Internal Dose (ng/kg blood) c 1.04 3.47 11.36 (n = 10) (n = 10) (n = 10) 800 38.42 (n = 12) 27.94 0.79 25.17 1.02 e 26.01 0.90 f 23.80 0.45 e aNOAEL for selected endpoint. bLOAEL for selected endpoint. cFrom the Em ond PB PK m odel described in 3.3. d V alues are the m ean SE. D ata obtained from Figure 7 in Ohsako et al. 2001. e Statistically significant as com pared to control (p < 0.01). f Statistically significant as com pared to control (p < 0.05). 2 3 4 E .1 .1 9 . S h i et al. (2007) Administered Dose (ng/kg-day) 0 0.143 a 0.714 b 7.14 28.6 Internal Dose (ng/kg blood) c Endpoint 0 (n = 10) 0.34 (n = 10) 1.07 (n = 10) 5.23 (n = 10) 13.91 (n = 10) Serum estradiol - 17p at proestrus 9 in female rats at 9 m o.of age 102.86 13.10 (pg/ml) d 86.19 6.19 63.33 9.29 e 48.1 5.95 e 38.57 7.14 e aNOAEL identified. bLOAEL identified. cFrom the Em ond PB PK m odel described in 3.3. d V alues are the m ean SE. D ata obtained from Figure 4 in Shi et al. 2007. e Statistically significant as com pared to control (p < 0.05). 5 This document is a draftfor review purposes only and does not constitute Agency policy. E-11 DRAFT--DO NOT CITE OR QUOTE 1 E.1.20. Smialowicz et al. (2008) 0 Endpoint PFC per 106 cells in fem ale m ice c PFC x 104 per spleen in fem ale m ice c 0 (n = 15) 1491 716 27.8 13.4 Administered Dose (ng/kg-day) 1.07 a 10.7 107 Internal Dose (ng/kg blood) b 0.44 2.46 13.40 (n = 14) (n = 15) (n = 15) 321 31.65 (n = 8) 1129 1 7 1 d 9 4 5 5 1 6 d 677 465 d 161 1 1 7 d 21 13.6 d 17.6 9.4 d 12.6 8.7 d 3.0 3.1 d aLOAEL identified. bFrom the Emond PBPK m odel described in 3.3. c V alues are the m ean SD. D ata obtained from Table 4 in S m ialow icz et al. 2008. d Statistically significant as com pared to control (p < 0.05). 2 3 4 E .1 .2 1 . T o th e t al. (1979) Endpoint Num ber w ith am yloidosis plus skin lesions in m ice c Num ber with skin lesions in m ice c 0 0 (n =38) 0/38 (0%) 0/38 (0%) Administered Dose (ng/kg-day) 1a 100 Internal Dose (ng/kg blood) b 0.57 14.21 (n = 44) (n = 44) 5/44 (11%) 10/44 (23%) 5/44 (11%) 13/44 (30%) 1,000 91.21 (n = 43) 17/43 (40%) 25/43 (58%) aLOAEL identified. bFrom the Emond PBPK m odel described in 3.3. c Data obtained from Table 2 in Toth et al. 1979. 5 6 This document is a draftfor review purposes only and does not constitute Agency policy. E-12 DRAFT--DO NOT CITE OR QUOTE 1 E.1.22. Van Birgelen et al. (1995) Administered Dose (ng/kg-day) 0 14 a 26 47 320 1,024 Internal Dose (ng/kg blood) b Endpoint Hepatic retinol (m g/g liver) in fem ale rats c Hepatic retinol palmitate (m g/g liver) in fem ale rats c Plasm a FT4 (pmol/liter) in fem ale rats c Plasm a TT4 (nmol/liter) in fem ale rats c 0 n=8 14.9 3.1 472 96 2 3 .4 1.1 40.9 2.4 7.20 n=8 8.4 1.2 d 94 24 d 24.5 2.0 41.4 1.9 11.76 n=8 8.2 0.8 d 107 27 d 22.4 1.0 41.4 2.3 18.09 n=8 86.41 n=8 250.16 n=8 5.1 0.3 d 2.2 0.3 d 0.6 0.2 d 74 14 d 22 8 d 3 1d 19.3 3.3 16.3 1.5 d 10.3 1.7 d 32.3 2.6 d 33.6 2.2 d 25.5 2.7 d aLOAEL identified. bFrom the Emond PBPK m odel described in 3.3. c V alues are the m ean SE. Data obtained from Table 3 in V an B irgelen et al. 1995. d Statistically significant as com pared to control (p < 0.05). 2 3 4 E .1 .2 3 . W h ite e t al. (1986) Endpoint CH50 (U/m l) in fem ale m ice c 0 0 (n = 8) 91 5 10 a 1.09 (n = 8) 54 3 d Administered Dose (ng/kg-day) 50 100 500 Internal Dose (ng/kg blood) b 4.08 7.14 26.81 (n = 8) (n = 8) (n = 8) 63 4d 56 9 d 41 6 d 1,000 48.72 (n = 8) 32 6 d 2,000 90.56 (n = 8) 17 6 d aLOAEL identified. bFrom the Emond PBPK m odel described in 3.3. c V alues are the m ean SE. Data obtained from Table 1 in W hite et al. 1986. dStatistically significant as com pared to control (p < 0.05). 5 This document is a draftfor review purposes only and does not constitute Agency policy. E-13 DRAFT--DO NOT CITE OR QUOTE 1 E.2. A L T E R N A T E D O SE: W H O L E BL O O D B M D S R E SU L T S 2 E.2.1. A m in et al., 2000: 0.25% Saccharin C onsum ed, Fem ale 3 E .2 .1 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes linear b 1 0.551 179.214 9.147E+00 6.094E+00 polynom ial, 2degree 1 0.551 179.214 9.147E+00 6.094E+00 power 1 0.551 179.214 9 .1 4 7 E + 0 0 6 .0 9 4 E + 0 0 p ow er b oun d hit (p ow er = 1) power, unrestricted c 0 N /A 180.858 8.367E+00 3.419E+00 unrestricted (power = 0.736) a N on-constant variance m odel selected (p = 0 .0005) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 4 5 6 E .2 .1 .2 . Outputfor Selected Model: Linear 7 A m in et al., 20 0 0 : 0 .2 5 % S acc h arin C o n su m ed , F em ale 8 9 10 11 P o l y n o m i a l M o d e l . ( V e r s i o n : 2 . 1 3 ; D a t e : 0 4 / 0 8 / 2 0 0 8 ) 12 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 1 _ A m i n _ 2 0 0 0 _ 2 5 _ S C _ L i n e a r _ 1 . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 1 _ A m i n _ 2 0 0 0 _ 2 5 _ S C _ L i n e a r _ 1 . p l t 14 M o n F e b 08 1 0 : 4 4 : 2 2 2 0 1 0 15 16 17 18 19 20 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 21 22 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 23 24 25 D e p e n d e n t v a r i a b l e = M e a n 26 I n d e p e n d e n t v a r i a b l e = D o s e 27 S i g n s o f t h e p o l y n o m i a l c o e f f i c i e n t s a r e n o t r e s t r i c t e d 28 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 29 30 T o t a l n u m b e r o f d o s e g r o u p s = 3 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 33 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 34 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 35 36 37 38 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 39 lalpha = 5.29482 40 rho = 0 41 beta_0 = 31.5112 42 beta_1 = -1.97726 43 This document is a draftfor review purposes only and does not constitute Agency policy. E-14 DRAFT--DO NOT CITE OR QUOTE 1 2 Asymptotic Correlation Matrix of Parameter Estimates 3 4 lalpha rho beta 0 beta 1 5 6 lalpha 1 -0.99 -0.029 0.044 7 o o o o 8 rho -0.99 1 0.026 9 10 beta 0 -0.029 0.026 1 -0.94 11 12 b e t a 1 0.044 -0.94 1 13 14 15 16 P a r a m e t e r E s t i m a t e s 17 18 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 19 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 20 lalpha -2.54215 1.65048 -5.77702 0.692726 21 rho 2.40985 0.541771 1.34799 3.4717 22 beta 0 31.2644 4.1929 23.0464 39.4823 23 beta 1 -1.9414 0.436071 -2.79609 -1.08672 24 25 26 27 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 28 29 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 30 31 32 0 10 31.7 31.3 20.6 17.8 0.0727 33 3 . 3 7 8 10 24.6 24.7 12 13.4 -0.0264 34 1 0 . 5 7 10 10.7 10.8 5.33 4.91 -0.0362 35 36 37 38 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 39 40 41 M o d e l A1: Yij Mu(i) + e(ij) 42 Var{e(ij)} Sigma^2 43 44 M o d e l A2: Yij Mu(i) + e(ij) 45 Var{e(ij)} Sigma(i)^2 46 47 M o d e l A3: Yij = Mu(i) + e(ij) 48 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 49 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 50 w e r e s p e c i f i e d b y t h e u s e r 51 52 M o d e l R: Yi = M u + e(i) 53 V a r { e ( i ) } = S i g m a ^ 2 54 55 56 L i k e l i h o o d s o f I n t e r e s t 57 58 Model Log(likelihood) # Param's AIC 59 A1 -92.841935 4 193.683870 60 A2 -85.255316 6 182.510632 61 A3 -85.429148 5 180.858295 62 fitted -85.606998 4 179.213995 63 R -98.136607 2 200.273213 64 65 66 E x p l a n a t i o n o f T e s t s 67 68 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 69 (A2 vs. R) 70 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) This document is a draftfor review purposes only and does not constitute Agency policy. E-15 DRAFT--DO NOT CITE OR QUOTE 1 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 2 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 3 (Note: When rho=0 the results of Test 3 and Test 2 will be the same.) 4 5 Tests of Interest 6 7 Test -2*log(Likelihood Ratio) Test df p-value 8 9 Test 1 25.7626 4 <.0001 10 T e s t 2 15.1732 2 0.0005072 11 T e s t 3 0.347663 1 0.5554 12 T e s t 4 0.3557 1 0.5509 13 14 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 15 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 16 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 17 18 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 19 m o d e l a p p e a r s t o b e a p p r o p r i a t e 20 21 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 22 t o b e a p p r o p r i a t e h e r e 23 24 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 25 t o a d e q u a t e l y d e s c r i b e t h e d a t a 26 27 28 B e n c h m a r k D o s e C o m p u t a t i o n 29 30 S p e c i f i e d e f f e c t = 1 31 32 R i s k T y p e = Estimated standard deviations from the control mean 33 34 C o n f i d e n c e l e v e l = 0.95 35 36 BMD = 9.14709 37 38 39 BMDL = 6.09414 40 This document is a draftfor review purposes only and does not constitute Agency policy. E-16 DRAFT--DO NOT CITE OR QUOTE 1 E.2.1.3. F igure f o r Selected M odel: L in ea r Linear Model with 0.95 Confidence Level 2 10:44 02/08 2010 3 4 5 E .2.1.4. Outputfor Additional Model Presented: Power, Unrestricted 6 A m in et al., 2000: 0.25% Saccharin Consum ed, Fem ale 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 1 _ A m i n _ 2 0 0 0 _ 2 5 _ S C _ P w r _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 1 _ A m i n _ 2 0 0 0 _ 2 5 _ S C _ P w r _ U _ 1 . p l t 13 M o n F e b 08 1 0 : 4 4 : 2 2 2 0 1 0 14 15 16 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 T h e p o w e r is n o t r e s t r i c t e d 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-17 DRAFT--DO NOT CITE OR QUOTE 1 Total number of dose groups = 3 2 Total number of records with missing values = 0 3 Maximum number of iterations = 250 4 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 5 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 6 7 8 9 Default Initial Parameter Values 10 lalpha = 5.29482 11 rho = 0 12 control = 31.6727 13 slope = -2.2195 14 power = 0.952715 15 16 17 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 18 19 lalpha rho control slope power 20 21 l a l p h a 1 -0.99 0.34 -0.17 -0.061 22 23 rho -0.99 1 -0.42 0.19 0.068 24 Lf) o Lf) o 25 c o n t r o l 0.34 -0.42 1 -0.72 26 27 slope -0.17 0.19 -0.72 1 0.97 28 29 power -0.061 0.068 0.97 1 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 36 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 37 lalpha -2.48291 2.08669 -6.57274 1.60693 38 rho 2.38455 0.692047 1.02817 3.74094 39 c o n t r o l 32.99 5.40754 22.3914 43.5886 40 slope -3.91099 3.83883 -11.435 3.61299 41 power 0.735877 0.350669 0.0485775 1.42318 42 43 44 45 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 46 47 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 48 49 50 0 10 31.7 33 20.6 18.7 -0.223 51 3 . 3 7 8 10 24.6 23.4 12 12.4 0.302 CO o o 52 1 0 . 5 7 10 10.7 10.8 5.33 4.94 53 54 W a r n i n g : L i k e l i h o o d f o r f i t t e d m o d e l l a r g e r t h a n t h e L i k e l i h o o d f o r m o d e l A3. 55 56 57 58 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 59 60 61 M o d e l A1: Yij Mu(i) + e(ij) 62 Var{e(ij)} Sigma^2 63 64 M o d e l A2: Yij Mu(i) + e(ij) 65 Var{e(ij)} Sigma(i)^2 66 67 M o d e l A3: Yij = Mu(i) + e(ij) 68 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 69 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 70 w e r e s p e c i f i e d b y t h e u s e r This document is a draftfor review purposes only and does not constitute Agency policy. E-18 DRAFT--DO NOT CITE OR QUOTE 1 2 M o d e l R: Yi = M u + e(i) 3 Var{e(i)} = Sigma^2 4 5 6 Likelihoods of Interest 7 8 Model Log(likelihood) # Param's AIC 9 A1 -92.841935 4 193.683870 10 A2 -85.255316 6 182.510632 11 A3 -85.429148 5 180.858295 12 fitted -85.429148 5 180.858295 13 R -98.136607 2 200.273213 14 15 16 E x p l a n a t i o n o f T e s t s 17 18 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 19 (A2 vs. R) 20 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 21 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 22 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 23 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 24 25 T e s t s o f I n t e r e s t 26 27 Test -2*log(Likelihood Ratio) Test df p-value 28 29 T e s t 1 25.7626 4 <.0001 30 T e s t 2 15.1732 2 0.0005072 31 T e s t 3 0.347663 1 0.5554 32 T e s t 4 -8.2423e-013 0 NA 33 34 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 35 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 36 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 37 38 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 39 m o d e l a p p e a r s t o b e a p p r o p r i a t e 40 41 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 42 t o b e a p p r o p r i a t e h e r e 43 44 N A - D e g r e e s o f f r e e d o m f o r T e s t 4 a r e l e s s t h a n o r e q u a l t o 0. T h e C h i - S q u a r e 45 t e s t f o r f i t is n o t v a l i d 46 47 48 B e n c h m a r k D o s e C o m p u t a t i o n 49 50 S p e c i f i e d e f f e c t = 1 51 52 R i s k T y p e = Estimated standard deviations from the control mean 53 54 C o n f i d e n c e l e v e l = 0.95 55 56 B M D = 8 . 3 6 6 7 8 57 58 59 B M D L = 3 . 4 1 9 0 6 60 61 This document is a draftfor review purposes only and does not constitute Agency policy. E-19 DRAFT--DO NOT CITE OR QUOTE 1 E.2.1.5. F igure f o r A d d itio n a l M o d el Presented: Power, U nrestricted Power Model with 0.95 Confidence Level 2 10:44 02/08 2010 3 4 5 E.2.2. A m in et al., 2000: 0.25% Saccharin P reference R atio, Fem ale 6 E .2 .2 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2 pValue AIC BMD (ng/kg) BMDL (ng/kg) Notes linear b 1 0.002 227.807 1.162E+01 5.572E+00 polynom ial, 2degree 1 0.002 227.807 1.162E+01 5.572E +00 power 1 0 .0 0 2 2 2 7 .8 0 7 1.162E +01 5 .5 7 2 E + 0 0 p o w er b oun d hit (p ow er = 1) a N on-con stant variance m odel selected (p = 0 .0 1 3 5 ) b Best-fitting m odel, BM D S output presented in this appendix 7 8 This document is a draftfor review purposes only and does not constitute Agency policy. E-20 DRAFT--DO NOT CITE OR QUOTE 1 E .2 .2 .2 . Outputfor Selected Model: Linear 2 A m in et al., 2000: 0.25% Saccharin Preference Ratio, Fem ale 3 4 5 6 Polynomial Model. (Version: 2.13; Date: 04/08/2008) 7 Input Data File: C:\1\Blood\2_Amin_2000_25_SP_Linear_1.(d) 8 Gnuplot Plotting File: C:\1\Blood\2_Amin_2000_25_SP_Linear_1.plt 9 M o n Feb 08 1 0 : 4 4 : 4 9 2010 10 11 12 13 14 15 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 16 17 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 18 19 20 D e p e n d e n t v a r i a b l e = M e a n 21 I n d e p e n d e n t v a r i a b l e = D o s e 22 S i g n s o f t h e p o l y n o m i a l c o e f f i c i e n t s a r e n o t r e s t r i c t e d 23 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 24 25 T o t a l n u m b e r o f d o s e g r o u p s = 3 26 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 27 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 28 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 29 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 30 31 32 33 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 34 lalpha = 6.34368 35 rho = 0 36 beta_0 = 75.4888 37 beta_1 = -2.24733 38 39 40 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 41 42 lalpha rho beta_0 beta_1 43 44 l a l p h a 1 -1 0.22 -0.31 45 46 rh o -1 1 -0.22 0.31 47 48 b e t a _ 0 0.22 -0.22 1 -0.77 49 50 b e t a _ 1 -0.31 0.31 -0.77 1 51 52 53 54 P a r a m e t e r E s t i m a t e s 55 56 95..0% W a l d C o n f i d e n c e I n t e r v a l 57 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 58 lalpha 3.00523 9.2122 -15.0503 21.0608 59 rho 0.797764 2.21138 -3.53646 5.13199 60 beta_0 75.1087 6.74312 61.8924 88.3249 61 beta_1 -2.16469 1.00825 -4.14082 -0.188553 62 63 64 65 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 66 67 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res. 68 This document is a draftfor review purposes only and does not constitute Agency policy. E-21 DRAFT--DO NOT CITE OR QUOTE 1 2 0 10 82.1 75.1 13.3 25.2 0 884 3 3.378 10 58.1 67.8 33.9 24.2 - .27 4 10.57 10 54.9 52.2 19.5 21.8 0 383 5 6 7 8 Model Descriptions for likelihoods calculated 9 10 11 M o d e l A1: Yij Mu(i) + e(ij) 12 Var{e(ij)} Sigma^2 13 14 M o d e l A2: Yij Mu(i) + e(ij) 15 Var{e(ij)} Sigma(i)^2 16 17 M o d e l A3: Yij = Mu(i) + e(ij) 18 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 19 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 20 w e r e s p e c i f i e d b y t h e u s e r 21 22 M o d e l R: Yi = M u + e(i) 23 V a r { e ( i ) } = S i g m a ^ 2 24 25 26 L i k e l i h o o d s o f I n t e r e s t 27 28 Model Log(likelihood) # Param's AIC 29 A1 -108.574798 4 225.149597 30 A2 -104.269377 6 220.538754 31 A3 -105.147952 5 220.295903 32 fitted -109.903705 4 227.807410 33 R -112.382522 2 228.765045 34 35 36 E x p l a n a t i o n o f T e s t s 37 38 T e s t 1 Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 39 (A2 vs. R) 40 T e s t 2 A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 41 T e s t 3 A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 42 T e s t 4 D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 43 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 44 45 T e s t s o f I n t e r e s t 46 47 Test -2*log(Likelihood Ratio) Test df p-value 48 49 T e s t 1 16.2263 4 0.00273 50 T e s t 2 8.61084 2 0.0135 51 T e s t 3 1.75715 1 0.185 52 T e s t 4 9.51151 1 0.002042 53 54 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 55 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 56 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 57 58 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 59 m o d e l a p p e a r s t o b e a p p r o p r i a t e 60 61 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 62 t o b e a p p r o p r i a t e h e r e 63 64 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 65 m o d e l 66 67 68 B e n c h m a r k D o s e C o m p u t a t i o n 69 70 S p e c i f i e d e f f e c t = 1 This document is a draftfor review purposes only and does not constitute Agency policy. E-22 DRAFT--DO NOT CITE OR QUOTE 1 2 Risk Type 3 4 Confidence level = 5 6 BMD = 7 8 9 BMDL = 10 11 Estimated standard deviations from the control mean 0.95 11 .6241 5. 5 7 2 1 5 12 E .2 .2 .3 . Figurefor Selected Model: Linear Linear Model with 0.95 Confidence Level Mean Response 13 10:44 02/08 2010 14 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-23 DRAFT--DO NOT CITE OR QUOTE 1 E.2.3. Amin et al., 2000: 0.50% Saccharin Consumed, Female 2 E.2.3.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes linear b 1 0.060 158.591 1.016E+01 6.567E+00 polynom ial, 2degree 1 0.060 158.591 1.016E+01 6.567E +00 power 1 0.060 158.591 1.016E +01 6 .5 6 7 E + 0 0 p ow er b oun d hit (p ow er = 1) power, unrestricted c 0 N /A 157.060 6.567E +00 1.155E+00 unrestricted (power = 0.396) a N on-con stant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .3 .2 . Outputfor Selected Model: Linear 6 A m in et al., 2000: 0.50% Saccharin Consum ed, Fem ale 7 8 9 10 P o l y n o m i a l M o d e l . ( V e r s i o n : 2 . 1 3 ; D a t e : 0 4 / 0 8 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 3 _ A m i n _ 2 0 0 0 _ 5 0 _ S C _ L i n e a r _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 3 _ A m i n _ 2 0 0 0 _ 5 0 _ S C _ L i n e a r _ 1 . p l t 13 M o n F e b 08 1 0 : 4 5 : 2 0 2 0 1 0 14 15 16 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 S i g n s o f t h e p o l y n o m i a l c o e f f i c i e n t s a r e n o t r e s t r i c t e d 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 28 29 T o t a l n u m b e r o f d o s e g r o u p s = 3 30 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 31 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 32 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 33 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 34 35 36 37 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 38 lalpha = 4.68512 39 rho = 0 40 beta_0 = 20.0631 41 beta_1 = -1.57142 42 43 44 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 45 This document is a draftfor review purposes only and does not constitute Agency policy. E-24 DRAFT--DO NOT CITE OR QUOTE 1 lalpha rho beta 0 beta 1 2 3 lalpha 1 -0.96 0.019 -0.0016 4 5 rho -0.96 1 -0.031 0.015 6 7 beta 0 0.019 -0.031 1 -0.96 8 9 beta 1 -0.0016 0.015 -0.96 1 10 11 12 13 P a r a m e t e r E s t i m a t e s 14 15 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 16 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 17 lalpha -0.982115 0.982262 -2.90731 0.943084 18 rho 2.11808 0.401166 1.33181 2.90435 19 beta 0 18.6171 3.1782 12.3879 24.8462 20 beta 1 -1.33226 0.322037 -1.96344 -0.70108 21 22 23 24 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 25 26 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 27 28 29 0 10 22.4 18.6 16 13.5 0.873 30 3 . 3 7 8 10 11.4 14.1 7.66 10.1 -0.856 31 1 0 . 5 7 10 4.54 4.54 3.33 3.04 -0.00339 32 33 34 35 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 36 37 38 M o d e l A1: Yij Mu(i) + e(ij) 39 Var{e(ij)} Sigma^2 40 41 M o d e l A2: Yij Mu(i) + e(ij) 42 Var{e(ij)} Sigma(i)^2 43 44 M o d e l A3: Yij = Mu(i) + e(ij) 45 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 46 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 47 w e r e s p e c i f i e d b y t h e u s e r 48 49 M o d e l R: Yi = M u + e(i) 50 V a r { e ( i ) } = S i g m a ^ 2 51 52 53 L i k e l i h o o d s o f I n t e r e s t 54 55 Model Log(likelihood) # Param's AIC 56 A1 -83.696404 4 175.392808 57 A2 -73.511830 6 159.023660 58 A3 -73.530233 5 157.060467 59 fitted -75.295363 4 158.590726 60 R -90.294746 2 184.589492 61 62 63 E x p l a n a t i o n o f T e s t s 64 65 T e s t 1 Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 66 (A2 vs. R) 67 T e s t 2 A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 68 T e s t 3 A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 69 T e s t 4 D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 70 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) This document is a draftfor review purposes only and does not constitute Agency policy. E-25 DRAFT--DO NOT CITE OR QUOTE 1 2 Tests of Interest 3 4 Test -2*log(Likelihood Ratio) Test df p-value 5 6 Test 1 33.5658 4 <.0001 7 Test 2 20.3691 2 <.0001 8 Test 3 0.0368066 1 0.8479 9 Test 4 3.53026 1 0.06026 10 11 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 12 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 13 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 14 15 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 16 m o d e l a p p e a r s t o b e a p p r o p r i a t e 17 18 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 19 t o b e a p p r o p r i a t e h e r e 20 21 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 22 m o d e l 23 24 25 B e n c h m a r k D o s e C o m p u t a t i o n 26 27 S p e c i f i e d e f f e c t = 1 28 29 R i s k T y p e = Estimated standard deviations from the control mean 30 31 C o n f i d e n c e l e v e l = 0.95 32 33 BMD = 10.1633 34 35 36 BMDL = 6.56742 37 This document is a draftfor review purposes only and does not constitute Agency policy. E-26 DRAFT--DO NOT CITE OR QUOTE 1 E.2.3.3. F igure f o r Selected M odel: L in ea r Linear Model with 0.95 Confidence Level 2 10:45 02/08 2010 3 4 5 E .2.3.4. Outputfor Additional Model Presented: Power, Unrestricted 6 A m in et al., 2000: 0.50% Saccharin Consum ed, Fem ale 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 3 _ A m i n _ 2 0 0 0 _ 5 0 _ S C _ P w r _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 3 _ A m i n _ 2 0 0 0 _ 5 0 _ S C _ P w r _ U _ 1 . p l t 13 M o n F e b 08 1 0 : 4 5 : 2 0 2 0 1 0 14 15 16 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 T h e p o w e r is n o t r e s t r i c t e d 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-27 DRAFT--DO NOT CITE OR QUOTE 1 Total number of dose groups = 3 2 Total number of records with missing values = 0 3 Maximum number of iterations = 250 4 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 5 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 6 7 8 9 Default Initial Parameter Values 10 lalpha = 4.68512 11 rho = 0 12 control = 22.3564 13 slope = -6.53901 14 power = 0.425213 15 16 17 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 18 19 lalpha rho control slope power 20 21 l a l p h a 1 -0.96 0.34 -0.31 -0.15 22 23 rho -0.96 1 -0.47 0.36 0.15 24 25 c o n t r o l 0.34 -0.47 1 -0.81 -0.52 26 27 slope -0.31 0.36 -0.81 1 0.92 28 29 power -0.15 0.15 -0.52 0.92 1 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 36 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 37 lalpha -0.708629 1.298 -3.25267 1.83541 38 rho 1.96142 0.529653 0.923323 2.99953 39 control 22.6293 4.48416 13.8405 31.4181 40 slope -7.10123 4.04394 -15.0272 0.824743 41 power 0.395571 0.168677 0.0649698 0.726173 42 43 44 45 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 46 47 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 48 49 50 0 10 22.4 22.6 16 15 -0.0577 51 3 . 3 7 8 10 11.4 11.1 7.66 7.46 0.105 52 1 0 . 5 7 10 4.54 4.58 3.33 3.12 -0.0475 53 54 D e g r e e s o f f r e e d o m f o r T e s t A 3 v s f i t t e d < = 0 55 56 57 58 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 59 60 61 M o d e l A1: Yij Mu(i) + e(ij) 62 Var{e(ij)} Sigma^2 63 64 M o d e l A2: Yij Mu(i) + e(ij) 65 Var{e(ij)} Sigma(i)^2 66 67 M o d e l A3: Yij = Mu(i) + e(ij) 68 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 69 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 70 w e r e s p e c i f i e d b y t h e u s e r This document is a draftfor review purposes only and does not constitute Agency policy. E-28 DRAFT--DO NOT CITE OR QUOTE 1 2 M o d e l R: Yi = M u + e(i) 3 Var{e(i)} = Sigma^2 4 5 6 Likelihoods of Interest 7 8 Model Log(likelihood) # Param's AIC 9 A1 -83.696404 4 175.392808 10 A2 -73.511830 6 159.023660 11 A3 -73.530233 5 157.060467 12 fitted -73.530233 5 157.060467 13 R -90.294746 2 184.589492 14 15 16 E x p l a n a t i o n o f T e s t s 17 18 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 19 (A2 vs. R) 20 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 21 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 22 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 23 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 24 25 T e s t s o f I n t e r e s t 26 27 Test -2*log(Likelihood Ratio) Test df p-vali 28 29 T e s t 1 33.5658 4 <.0001 30 T e s t 2 20.3691 2 <.0001 31 T e s t 3 0.0368066 1 0.8479 32 T e s t 4 00 NA 33 34 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 35 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 36 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 37 38 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 39 m o d e l a p p e a r s t o b e a p p r o p r i a t e 40 41 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 42 t o b e a p p r o p r i a t e h e r e 43 44 N A - D e g r e e s o f f r e e d o m f o r T e s t 4 a r e l e s s t h a n o r e q u a l t o 0. T h e C h i - S q u a r e 45 t e s t f o r f i t is n o t v a l i d 46 47 48 B e n c h m a r k D o s e C o m p u t a t i o n 49 50 S p e c i f i e d e f f e c t = 1 51 52 R i s k T y p e = Estimated standard deviations from the control mean 53 54 C o n f i d e n c e l e v e l = 0.95 55 56 B M D = 6 . 5 6 7 1 9 57 58 59 B M D L = 1 . 1 5 4 7 6 60 61 This document is a draftfor review purposes only and does not constitute Agency policy. E-29 DRAFT--DO NOT CITE OR QUOTE 1 E.2.3.5. F igure f o r A d d itio n a l M o d el Presented: Power, U nrestricted Power Model with 0.95 Confidence Level Mean Response 2 10:45 02/08 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-30 DRAFT--DO NOT CITE OR QUOTE 1 E.2.4. Amin et al., 2000: 0.50% Saccharin Preference Ratio, Female 2 E.2.4.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes linear b 1 0.135 234.250 8.144E+00 5.105E+00 polynom ial, 2degree 1 0.135 234.250 8.144E+00 5.105E+00 power 1 0.1 3 5 2 3 4 .2 5 0 8 .1 4 4 E + 0 0 5 .1 0 5 E + 0 0 p ow er b oun d hit (p ow er = 1) power, unrestricted c 0 N /A 234.020 2.598E +00 1.057E-14 unrestricted (power = 0.282) a C onstant variance m odel selected (p = 0 .5 5 9 3 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .4 .2 . Outputfor Selected Model: Linear 6 A m in et al., 2000: 0.50% Saccharin Preference Ratio, Fem ale 7 8 9 10 P o l y n o m i a l M o d e l . ( V e r s i o n : 2 . 1 3 ; D a t e : 0 4 / 0 8 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 4 _ A m i n _ 2 0 0 0 _ 5 0 _ S P _ L i n e a r C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 4 _ A m i n _ 2 0 0 0 _ 5 0 _ S P _ L i n e a r C V _ 1 . p l t 13 M o n F e b 08 1 0 : 4 5 : 5 0 2 0 1 0 14 15 16 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 S i g n s o f t h e p o l y n o m i a l c o e f f i c i e n t s a r e n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t 29 30 T o t a l n u m b e r o f d o s e g r o u p s = 3 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 33 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 34 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 35 36 37 38 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 39 alpha = 764.602 40 rho = 0 Specified 41 beta_0 = 65.8627 42 beta_1 = -3.34297 43 44 This document is a draftfor review purposes only and does not constitute Agency policy. E-31 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 ( *** The model parameter(s) -rho 4 have been estimated at a boundary point, or have been specified by the user 5 and do not appear in the correlation matrix ) 6 7 alpha beta 0 beta 1 8 9 alpha 1 2.6e-008 2.1e-009 10 CO r o CO r o 11 beta 0 2.6e-008 1 12 13 beta 1 2.1e-009 1 14 15 16 17 P a r a m e t e r E s t i m a t e s 18 19 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 20 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 21 alpha 741.255 191.391 366.135 1116.38 22 beta_0 65.8627 7.22524 51.7015 80.0239 23 beta 1 -3.34297 1.12815 -5.55412 -1.13183 24 25 26 27 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 28 29 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 30 31 32 0 10 72.7 65.9 24.6 27.2 0.797 33 3 . 3 7 8 10 44.5 54.6 32.9 27.2 -1.17 34 1 0 . 5 7 10 33.8 30.5 24.6 27.2 0.375 35 36 37 38 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 39 40 41 M o d e l A1: Yij Mu(i) + e(ij) 42 Var{e(ij)} Sigma^2 43 44 M o d e l A2: Yij Mu(i) + e(ij) 45 Var{e(ij)} Sigma(i)^2 46 47 M o d e l A3: Yij = Mu(i) + e(ij) 48 V a r { e ( i j ) } = S i g m a ^ 2 49 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 50 w e r e s p e c i f i e d b y t h e u s e r 51 52 M o d e l R: Yi = M u + e(i) 53 V a r { e ( i ) } = S i g m a ^ 2 54 55 56 L i k e l i h o o d s o f I n t e r e s t 57 58 Model Log(likelihood) # Param's AIC 59 A1 -113.009921 4 234.019841 60 A2 -112.428886 6 236.857773 61 A3 -113.009921 4 234.019841 62 fitted -114.125184 3 234.250368 63 R -117.976057 2 239.952114 64 65 66 E x p l a n a t i o n o f T e s t s 67 68 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e levels' 69 (A2 vs. R) 70 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) This document is a draftfor review purposes only and does not constitute Agency policy. E-32 DRAFT--DO NOT CITE OR QUOTE 1 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 2 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 3 (Note: When rho=0 the results of Test 3 and Test 2 will be the same.) 4 5 Tests of Interest 6 7 Test -2*log(Likelihood Ratio) Test df p-value 8 9 Test 1 11.0943 4 0.02552 10 T e s t 2 1.16207 2 0.5593 11 T e s t 3 1.16207 2 0.5593 12 T e s t 4 2.23053 1 0.1353 13 14 p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e 15 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 16 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 17 18 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 19 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 20 21 22 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 23 t o b e a p p r o p r i a t e h e r e 24 25 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 26 t o a d e q u a t e l y d e s c r i b e t h e d a t a 27 28 29 B e n c h m a r k D o s e C o m p u t a t i o n 30 31 S p e c i f i e d e f f e c t = 1 32 33 R i s k T y p e = Estimated standard deviations from the control mean 34 35 C o n f i d e n c e l e v e l = 0.95 36 37 BMD = 8.14425 38 39 40 BMDL = 5.10523 41 This document is a draftfor review purposes only and does not constitute Agency policy. E-33 DRAFT--DO NOT CITE OR QUOTE 1 E.2.4.3. F igure f o r Selected M odel: L in ea r Linear Model with 0.95 Confidence Level 2 3 4 5 E .2.4.4. Outputfor Additional Model Presented: Power, Unrestricted 6 A m in et al., 2000: 0.50% Saccharin Preference Ratio, Fem ale 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 4 _ A m i n _ 2 0 0 0 _ 5 0 _ S P _ P w r C V _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 4 _ A m i n _ 2 0 0 0 _ 5 0 _ S P _ P w r C V _ U _ 1 . p l t 13 M o n F e b 08 1 0 : 4 5 : 5 0 2 0 1 0 14 15 16 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 T h e p o w e r is n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t This document is a draftfor review purposes only and does not constitute Agency policy. E-34 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 3 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha = 764.602 12 rho = 0 Specified 13 control = 72.7273 14 slope = -20.0402 15 power = 0.281985 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 21 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 22 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 23 24 alpha control slope power 25 26 a l p h a 1 -1.2e-009 -1.2e-009 -2.2e-010 27 28 control 1.2e-009 1 -0.51 -0.22 29 30 slope 1.2e-009 -0.51 1 0.92 31 32 power 2.2e-010 -0.22 0.92 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 39 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 40 alpha 688.142 177.677 339.9 1036.38 41 control 72.7273 8.29543 56.4686 88.986 42 slope -20.0402 15.0576 -49.5526 9.47219 43 power 0.281985 0.325861 -0.35669 0.920661 44 45 46 47 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 48 49 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 50 51 52 0 10 72.7 72.7 24.6 26.2 4.67e-009 53 3 . 3 7 8 10 44.5 44.5 32.9 26.2 1.52e-008 54 1 0 . 5 7 10 33.8 33.8 24.6 26.2 1.77e-008 55 56 W a r n i n g : L i k e l i h o o d f o r f i t t e d m o d e l l a r g e r t h a n t h e L i k e l i h o o d f o r m o d e l A3. 57 58 59 60 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 61 62 63 M o d e l A1: Yij = M u ( i ) + e( 64 V a r { e i ;ij)} = S i g m a ^ 2 65 66 M o d e l A2: Yij = M u ( i ) + e( 67 V a r { e i :ij)} = S i g m a ( i ) ^ 2 68 69 M o d e l A3: Yij = M u ( i ) + e( 70 V a r { e i :ij)} = S i g m a ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-35 DRAFT--DO NOT CITE OR QUOTE 1 Model A3 uses any fixed variance parameters that 2 were specified by the user 3 4 M o d e l R: Yi = M u + e(i) 5 Var{e(i)} = Sigma^2 6 7 8 Likelihoods of Interest 9 10 Model Log(likelihood) # Param's AIC 11 A1 -113.009921 4 234.019841 12 A2 -112.428886 6 236.857773 13 A3 -113.009921 4 234.019841 14 fitted -113.009921 4 234.019841 15 R -117.976057 2 239.952114 16 17 18 E x p l a n a t i o n o f T e s t s 19 20 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 21 (A2 vs. R) 22 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 23 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 24 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 25 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 26 27 T e s t s o f I n t e r e s t 28 29 Test -2*log(Likelihood Ratio) Test df p-vali 30 31 T e s t 1 11.0943 4 0.02552 32 T e s t 2 1.16207 2 0.5593 33 T e s t 3 1.16207 2 0.5593 34 T e s t 4 -2.84217e-014 0 NA 35 36 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 37 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 38 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 39 40 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 41 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 42 43 44 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 45 t o b e a p p r o p r i a t e h e r e 46 47 N A - D e g r e e s o f f r e e d o m f o r T e s t 4 a r e l e s s t h a n o r e q u a l t o 0. T h e C h i - S q u a r e 48 t e s t f o r f i t is n o t v a l i d 49 50 51 B e n c h m a r k D o s e C o m p u t a t i o n 52 53 S p e c i f i e d e f f e c t = 1 54 55 R i s k T y p e = Estimated standard deviations from the control mean 56 57 C o n f i d e n c e l e v e l = 0.95 58 59 B M D = 2 . 5 9 8 3 1 60 61 62 B M D L = 1 . 0 5 6 6 1 e - 0 1 4 63 This document is a draftfor review purposes only and does not constitute Agency policy. E-36 DRAFT--DO NOT CITE OR QUOTE 1 E.2.4.5. F igure f o r A d d itio n a l M o d el Presented: Power, Unrestricted Power Model with 0.95 Confidence Level Mean Response 2 10:45 02/08 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-37 DRAFT--DO NOT CITE OR QUOTE 1 E.2.5. Bell et al., 2007a: B alano-P reputial Separation, Postnatal D ay 49 2 E .2 .5 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 2 0.684 1 12.136 2 .8 6 7 E + 0 0 1 .943E + 00 p o w er b oun d hit (p ow er = 1) logistic negative intercept (intercept = 2 0.342 113.915 6.159E+00 4.746E+00 -2.246) log-logistic a 2 0.777 111.908 2.246E+00 1.394E+00 slope bound hit (slope = 1) log-probit multistage, 3degree probit 2 0.269 1 14.254 5 .3 2 2 E + 0 0 3 .5 1 2 E + 0 0 slop e b ound hit (slop e = 1) 2 0.684 112.136 2.867E+00 1.943E+00 final B = 0 2 0.367 113.713 5.715E+00 4.422E+00 W eibull 2 0.684 1 12.136 2 .8 6 7 E + 0 0 1 .943E + 00 p o w er b oun d hit (p ow er = 1) gamma, unrestricted log-logistic, unrestricted b log-probit, unrestricted W eibull, unrestricted 1 0.566 113.746 1.862E+00 1.829E-01 unrestricted (pow er = 0.741) 1 0.501 113.871 1.998E+00 2.795E-01 unrestricted (slope = 0.93) 1 0.456 113.977 2.038E+00 3.250E-01 unrestricted (slope = 0.54) 1 0.551 113.771 1.914E+00 2.346E-01 unrestricted (pow er = 0.795) aBest-fitting m odel, BM D S output presented in this appendix b Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .5 .2 . Outputfor Selected Model: Log-Logistic 6 B ell et al., 2007a: Balano-Preputial Separation, Postnatal D ay 49 7 8 9 10 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 5 _ B e l l _ 2 0 0 7 _ B P S _ L o g L o g i s t i c _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 5 _ B e l l _ 2 0 0 7 _ B P S _ L o g L o g i s t i c _ 1 . p l t 13 M o n F e b 08 1 0 : 4 6 : 1 8 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e This document is a draftfor review purposes only and does not constitute Agency policy. E-38 D RAFT-- DO N O T CITE OR QUOTE 1 Slope pa r a m e t e r is r estricted as slope >= 1 2 3 Total number of observations = 4 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 12 13 14 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 15 background = 0.0333333 16 intercept = -2.99896 17 slope = 1 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( * * * T h e m o d e l p a r a m e t e r ( s ) - s l o p e 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 24 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 25 26 background intercept 27 28 b a c k g r o u n d 1 -0.49 29 30 i n t e r c e p t -0.49 1 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 37 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 38 b a c k g r o u n d 0.038005 39 intercept -3.00658 40 s l o p e 1 41 42 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 43 44 45 46 A n a l y s i s o f D e v i a n c e T a b l e 47 48 Model Log(likelihood) # Param's Deviance Test d.f. P-value 49 Full model -53.7077 4 50 F i t t e d m o d e l -53.954 2 0.492596 2 0.7817 51 R e d u c e d m o d e l -63.9797 1 20.544 3 0.0001309 52 53 AIC: 111.908 54 55 56 G o o d n e s s o f F i t 57 S c a l e d 58 Dose Est. Prob. Expected Observed Size Residual 59 60 0.0000 0.0380 1.140 1.000 30 -0.134 61 2.2040 0.1326 3.977 5.000 30 0.551 62 5.1378 0.2329 6.988 6.000 30 -0.427 63 18.4110 0.4965 14.895 15.000 30 0.038 64 65 C h i ^ 2 = 0 . 5 0 d.f. = 2 P-value = 0.7769 66 67 68 B e n c h m a r k D o s e C o m p u t a t i o n 69 70 S p e c i f i e d e f f e c t = 0.1 This document is a draftfor review purposes only and does not constitute Agency policy. E-39 DRAFT--DO NOT CITE OR QUOTE 1 2 Risk Type 3 4 Confidence level 5 6 BMD 7 8 BMDL 9 10 Extra risk 0.95 2.24647 1.39385 11 E .2 .5 .3 . Figurefor Selected Model: Log-Logistic Log-Logistic Model with 0.95 Confidence Level Fraction Affected 12 10:46 02/08 2010 13 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-40 DRAFT--DO NOT CITE OR QUOTE 1 E .2.5.4. Outputfor Additional Model Presented: Log-Logistic, Unrestricted 2 B ell et al., 2007a: Balano-Preputial Separation, Postnatal D ay 49 3 4 5 6 Logistic Model. (Version: 2.12; Date: 05/16/2008) 7 Input Data File: C:\1\Blood\5_Bell_2007_BPS_LogLogistic_U_1.(d) 8 Gnuplot Plotting File: C:\1\Blood\5_Bell_2007_BPS_LogLogistic_U_1.plt 9 M o n Feb 08 1 0 : 4 6 : 1 8 2010 10 11 12 0 13 14 15 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 16 17 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 18 19 20 D e p e n d e n t v a r i a b l e = D i c h E f f 21 I n d e p e n d e n t v a r i a b l e = D o s e 22 S l o p e p a r a m e t e r is n o t r e s t r i c t e d 23 24 T o t a l n u m b e r o f o b s e r v a t i o n s = 4 25 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 26 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 27 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 28 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 29 30 31 32 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 33 34 35 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 36 background = 0.0333333 37 intercept = -2.68464 38 slope = 0.858398 39 40 41 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 42 43 background intercept slope 44 CO o CO o 45 b a c k g r o u n d 1 0.35 46 47 i n t e r c e p t 1 -0.94 48 49 s l o p e 0.35 -0.94 1 50 51 52 53 P a r a m e t e r E s t i m a t e s 54 55 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 56 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 57 b a c k g r o u n d 1.0353402 * 58 intercept -2.84051 * 59 slope 0.929645 * 60 61 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 62 63 64 65 A n a l y s i s o f D e v i a n c e T a b l e 66 67 Model Log(likelihood) # Param's Deviance Test d.f. P-value 68 Full model -53.7077 4 This document is a draftfor review purposes only and does not constitute Agency policy. E-41 DRAFT--DO NOT CITE OR QUOTE 1 Fitted model -53.9354 3 0.455534 1 0.4997 2 Reduced model -63.9797 1 20.544 3 0.0001309 3 4 AIC: 113.871 5 6 7 Goodness of Fit 8 Scaled 9 Dose Est. Prob. Expected Observed Size Residual 10 11 0.0000 0.0353 1.060 1.000 30 -0.060 12 2.2040 0.1400 4.201 5.000 30 0.420 13 5.1378 0.2389 7.166 6.000 30 -0.499 14 18.4110 0.4858 14.573 15.000 30 0.156 15 16 C h i ^ 2 = 0 . 4 5 d.f. = 1 P-value = 0.5005 17 18 19 B e n c h m a r k D o s e C o m p u t a t i o n 20 21 S p e c i f i e d e f f e c t = 0.1 22 23 R i s k T y p e = Extra risk 24 25 C o n f i d e n c e l e v e l = 0.95 26 27 BMD = 1.99765 28 29 BMDL = 0.279534 30 31 This document is a draftfor review purposes only and does not constitute Agency policy. E-42 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 E.2.5.5. F igure f o r A d d itio n a l M o d el Presented: Log-Logistic, U nrestricted Log-Logistic Model with 0.95 Confidence Level 2 10:46 02/08 2010 3 4 This document is a draftfor review purposes only and does not constitute Agency policy. E-43 DRAFT--DO NOT CITE OR QUOTE 1 E.2.6. Cantoni et al., 1981: Urinary Coproporhyrins, 3 Months 2 E.2.6.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 2 0.003 32.882 3.209E+01 1.567E+01 Notes exponential (M 3) exponential (M4) b exponential (M 5) 2 0.003 3 2 .8 8 2 3.2 0 9 E + 0 1 1.567E +01 p ow er hit bound (d = 1) 1 0.486 23.459 5.339E-01 1.803E-01 1 0 .4 8 6 2 3 .4 5 9 5 .339E -01 1.803E -01 p ow er hit bound (d = 1) H ill 1 0.788 23.047 4.333E -01 error n lo w er b oun d hit (n = 1) linear polynom ial, 3degree power power, unrestricted c H ill, unrestricted 2 0.005 31.595 1.464E+01 2.753E+00 2 0.005 31.595 1.464E+01 2.753E+00 2 0.0 0 5 31 .5 9 5 1.464E +01 2 .7 5 3 E + 0 0 p ow er b oun d hit (p ow er = 1) 1 0.610 23.235 2.766E -02 2.031E-05 unrestricted (power = 0.304) 0 N /A 24.974 2.602E -01 error unrestricted (n = 0.739) a N on-constant variance m odel selected (p = 0 .0039) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2.6.2. Outputfor Selected Model: Exponential (M4) 6 Cantoni et al., 1981: Urinary Coproporhyrins, 3 M onths 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a Fil e : C : \ 1 \ B l o o d \ 6 C a n t o n i 1 9 8 1 U r i C o p r o E x p 1 . ( d 12 G n u p l o t P l o t t i n g Fil e : 13 M o n F e b 08 10 :4 6 : 14 15 16 F i g u r e l - U r i n a r y C o p r o p o r p h y r i n 3 m o n t h s 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-l) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-l) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; This document is a draftfor review purposes only and does not constitute Agency policy. E-44 DRAFT--DO NOT CITE OR QUOTE 1 sign = -1 for d e c r e a s i n g trend. 2 3 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 4 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 5 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 6 7 8 Dependent variable = Mean 9 Independent variable = Dose 10 D a t a a r e a s s u m e d t o b e d i s t r i b u t e d : n o r m a l l y 11 V a r i a n c e M o d e l : e x p ( l n a l p h a + r h o * l n ( Y [ d o s e ] ) ) 12 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 13 14 T o t a l n u m b e r o f d o s e g r o u p s = 4 15 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 16 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 17 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 18 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 19 20 M L E s o l u t i o n p r o v i d e d : E x a c t 21 22 23 I n i t i a l P a r a m e t e r V a l u e s 24 25 Variable Model 4 26 27 lnalpha -1.50063 28 rho 2.60979 29 a 0 . 7 0 4 3 0 3 30 b 0 . 0 6 0 4 9 6 1 31 c 4 . 4 7 2 6 8 32 d 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 Variable Model 4 39 40 lnalpha -1.75302 41 rho 2.6322 42 a 0 . 7 6 1 2 1 8 43 b 0 . 2 4 1 5 6 1 44 c 4 . 1 5 5 9 7 45 d 1 46 47 48 T a b l e o f S t a t s F r o m I n p u t D a t a 49 50 Dose N Obs Mean Obs Std 51 52 04 0.7414 0.3475 53 1.847 4 1.807 0.8341 54 8.839 4 2.734 1.506 55 50.05 4 3 2.6 56 57 58 E s t i m a t e d V a l u e s o f I n t e r e s t 59 60 Dose Est Mean Est Std Scaled Residual 61 62 0 0.7612 0.2907 -0.1366 63 1 . 8 4 7 1.626 0.7892 0.4588 64 8 . 8 3 9 2.88 1.674 -0.1743 65 5 0 . 0 5 3.164 1.895 -0.1725 66 67 68 69 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d : 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-45 DRAFT--DO NOT CITE OR QUOTE 1 Model A1: Yij = Mu(i) + e(ij) 2 V a r { e i ;ij)} = S i g m a ^ 2 3 4 Model A2: Yij = Mu(i) + e(ij) 5 V a r { e i :ij)} = S i g m a ( i ) ^ 2 6 7 Model A3: Yij = Mu(i) + e(ij) 8 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 9 10 M o d e l R: Yij = M u + e(i) 11 V a r { e i ij)} = S i g m a ^ 2 12 13 14 L i k e l i h o o d s o f I n t e r e s t 15 16 Model Log(likelihood) DF AIC 17 18 A1 -12.90166 5 35.80333 19 A2 -6.203643 8 28.40729 20 A3 -6.487204 6 24.97441 21 R -15.73713 2 35.47427 22 4 -6.729737 5 23.45947 23 24 25 Additive constant for all log-likelihoods = -14.7. This constant added to the 26 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 27 d e p e n d o n t h e m o d e l p a r a m e t e r s . 28 29 30 E x p l a n a t i o n o f T e s t s 31 32 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 33 T e s t 2: A r e H o m o g e n e o u s ? (A2 vs. A1) 34 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 35 36 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 37 38 39 T e s t s o f I n t e r e s t 40 41 T e s t -2*log(Likelihood Ratio) D. F. p-value 42 43 T e s t 1 19.07 6 0.004052 44 T e s t 2 13.4 3 0.003854 45 T e s t 3 0.5671 2 0.7531 46 T e s t 6a 0.4851 1 0.4861 47 48 49 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 50 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 51 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 52 53 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 54 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e . 55 56 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 57 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 58 59 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 60 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 61 62 63 B e n c h m a r k D o s e C o m p u t a t i o n s : 64 65 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 66 67 R i s k T y p e = E s t i m a t e d s t a n d a r d d e v i a t i o n s f r o m c o n t r o l 68 69 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-46 DRAFT--DO NOT CITE OR QUOTE 1 BMD = 0.533855 2 3 BMDL = 0.180293 4 5 6 E .2.6.3. Figurefor Selected Model: Exponential (M4) Exponential Model 4 with 0.95 Confidence Level 7 8 9 10 E .2.6.4. Outputfor Additional Model Presented: Power, Unrestricted 11 Cantoni et al., 1981: Urinary Coproporhyrins, 3 M onths 12 13 14 15 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 16 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 6 _ C a n t o n i _ 1 9 8 1 _ U r i C o p r o _ P w r _ U _ 1 . ( d ) 17 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 6 _ C a n t o n i _ 1 9 8 1 _ U r i C o p r o _ P w r _ U _ 1 . p l t 18 M o n F e b 08 1 0 : 4 6 : 4 7 2 0 1 0 19 20 21 F i g u r e 1 - U r i n a r y C o p r o p o r p h y r i n 3 m o n t h s 22 23 24 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 25 26 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 27 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-47 DRAFT--DO NOT CITE OR QUOTE 1 Dependent variable = Mean 2 Independent variable = Dose 3 The power is not restricted 4 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) * rho) 5 6 Total number of dose groups = 4 7 Total number of records with missing values = 0 8 Maximum number of iterations = 250 9 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 10 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 11 12 13 14 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 15 lalpha = 0.90039 16 rho = 0 17 control = 0.741372 18 slope = 0.93685 19 power = 0.224904 20 21 22 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 23 24 lalpha rho control slope power 25 CO Lf) o CO Lf) o 26 l a l p h a 1 -0.62 -0.036 0.024 27 28 rho -0.62 1 0.43 -0.2 -0.16 29 30 c o n t r o l 0.43 1 -0.28 0.086 31 32 slope -0.036 -0.2 -0.28 1 -0.77 33 34 p o w e r 0 . 0 2 4 - 0 . 1 6 0 . 0 8 6 - 0 . 7 7 1 35 36 37 38 P a r a m e t e r E s t i m a t e s 39 40 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 41 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 42 lalpha -1.78125 0.617807 -2.99213 -0.570373 43 rho 2.64332 0.744946 1.18325 4.10338 44 control 0.75678 0.139979 0.482426 1.03113 45 slope 0.845767 0.324854 0.209065 1.48247 46 power 0.304211 0.135053 0.0395119 0.568909 47 48 49 50 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 51 52 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 53 54 55 04 0.741 0.757 0.348 0.284 -0.109 56 1 . 8 4 7 4 1.81 1.78 0.834 0.877 0.0705 57 8 . 8 3 9 4 2.73 2.4 1.51 1.3 0.515 58 5 0 . 0 5 4 3 3.54 2.6 2.18 -0.493 59 60 61 62 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 63 64 65 M o d e l A1: Yij = Mu(i) + e(ij) 66 V a r { e i ;i j ) } = S i g m a ^ 2 67 68 M o d e l A2: Yij = Mu(i) + e(ij) 69 V a r { e i :i j ) } = S i g m a ( i ) ^ 2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-48 DRAFT--DO NOT CITE OR QUOTE 1 Model A3: Yij = Mu(i) + e(ij) 2 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 3 Model A3 uses any fixed variance parameters that 4 were specified by the user 5 6 M o d e l R: Yi = M u + e(i) 7 Var{e(i)} = Sigma^2 8 9 10 L i k e l i h o o d s o f I n t e r e s t 11 12 Model Log(likelihood) # Param's AIC 13 A1 -12.901663 5 35.803325 14 A2 -6.203643 8 28.407287 15 A3 -6.487204 6 24.974409 16 fitted -6.617347 5 23.234694 17 R -15.737135 2 35.474269 18 19 20 E x p l a n a t i o n o f T e s t s 21 22 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 23 (A2 vs. R) 24 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 25 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 26 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 27 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 28 29 T e s t s o f I n t e r e s t 30 31 Test -2*log(Likelihood Ratio) Test df p-value 32 33 T e s t 1 19.067 6 0.004052 34 T e s t 2 13.396 3 0.003854 35 T e s t 3 0.567122 2 0.7531 36 T e s t 4 0.260285 1 0.6099 37 38 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 39 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 40 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 41 42 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 43 m o d e l a p p e a r s t o b e a p p r o p r i a t e 44 45 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 46 t o b e a p p r o p r i a t e h e r e 47 48 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 49 t o a d e q u a t e l y d e s c r i b e t h e d a t a 50 51 52 B e n c h m a r k D o s e C o m p u t a t i o n 53 54 S p e c i f i e d e f f e c t = 1 55 56 R i s k T y p e = Estimated standard deviations from the control mean 57 58 C o n f i d e n c e l e v e l = 0.95 59 60 B M D = 0 . 0 2 7 6 5 9 9 61 62 63 B M D L = 2 . 0 3 1 4 3 e - 0 0 5 64 65 This document is a draftfor review purposes only and does not constitute Agency policy. E-49 DRAFT--DO NOT CITE OR QUOTE 1 E.2.6.5. F igure f o r A d d itio n a l M o d el Presented: Power, U nrestricted Power Model with 0.95 Confidence Level Mean Response 2 10:46 02/08 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-50 DRAFT--DO NOT CITE OR QUOTE 1 E.2.7. C antoni et al., 1981: U rinary Porphyrins 2 E .2 .7 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) b 2 <0.001 55.465 3.760E+00 2.762E+00 Notes exponential (M 3) 2 < 0 .001 55.465 3 .7 6 0 E + 0 0 2 .7 6 2 E + 0 0 p o w er hit boun d (d = 1) exponential (M 4) 1 <0.0001 59.187 2.484E-01 1.448E-01 exponential (M 5) 0 N /A 61.084 2.878E-01 1.461E-01 Hill 0 N /A 62.199 6.233E+00 3.341E+00 linear polynom ial, 3degree power 2 <0.001 57.187 2.484E-01 1.448E-01 1 <0.0001 10.000 error error 1 <0.0001 59.084 2.878E-01 1.461E-01 a N on-constant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2.7.2. Outputfor Selected Model: Exponential (M2) 6 Cantoni et al., 1981: Urinary Porphyrins 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 7 _ C a n t o n i _ 1 9 8 1 _ U r i P o r _ E x p _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 M o n F e b 08 1 0 : 4 7 : 2 4 2 0 1 0 14 15 16 T a b l e 1, d o s e c o n v e r t e d t o n g p e r k g p e r d a y 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 30 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 31 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 32 33 34 Dependent variable = Mean This document is a draftfor review purposes only and does not constitute Agency policy. E-51 DRAFT--DO NOT CITE OR QUOTE 1 Independent variable = Dose 2 Data are assumed to be distributed: normally 3 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 4 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) rho) 5 6 Total number of dose groups = 4 7 Total number of records with missing values = 0 8 Maximum number of iterations = 250 9 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 10 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 11 12 M L E s o l u t i o n p r o v i d e d : E x a c t 13 14 15 I n i t i a l P a r a m e t e r V a l u e s 16 17 Variable Model 2 18 19 lnalpha -3.57509 20 rho 2.23456 21 a 3 . 3 6 4 5 3 22 b 0 . 0 8 1 9 8 0 1 23 c 0 24 d 1 25 26 27 28 P a r a m e t e r E s t i m a t e s 29 30 Variable Model 2 31 32 lnalpha -1.85879 33 rho 1.82273 34 a 3 . 5 7 8 9 6 35 b 0 . 0 8 0 3 3 4 7 36 c 0 37 d 1 38 39 40 T a b l e o f S t a t s F r o m I n p u t D a t a 41 42 Dose N Obs Mean Obs Std Dev 43 44 04 2.27 0.49 45 1.847 4 5.55 0.85 46 8.839 3 7.62 1.79 47 50.05 3 196.9 63.14 48 49 50 E s t i m a t e d V a l u e s o f I n t e r e s t 51 52 Dose Est Mean Est Std Scaled Residual 53 54 0 3.579 1.262 -2.074 55 1 . 8 4 7 4.152 1.445 1.936 56 8 . 8 3 9 7.28 2.41 0.2441 57 5 0 . 0 5 199.5 49.25 -0.09069 58 59 60 61 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d : 62 63 M o d e l A1: Yij = M u ( i ) + e( ij) 64 V a r { e (ij)} = S i g m a " 2 65 66 M o d e l A2: Yij = M u ( i ) + e( ij) 67 V a r { e (ij)} = S i g m a ( i ) 1"2 68 69 M o d e l A3: Yij = M u ( i ) + e( ij) 70 V a r { e (ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) This document is a draftfor review purposes only and does not constitute Agency policy. E-52 DRAFT--DO NOT CITE OR QUOTE 1 2 M o d e l R: Yij = M u + e(i) 3 Var{e(ij)} = Sigma^2 4 5 6 Likelihoods of Interest 7 8 Model Log(likelihood) DF AIC 9 10 A1 -51.42175 5 112.8435 11 A2 -15.31211 8 46.62422 12 A3 -15.66963 6 43.33925 13 R -68.75058 2 141.5012 14 2 -23.73254 4 55.46509 15 16 17 Additive constant for all log-likelihoods = -12.87. This constant added to the 18 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 19 d e p e n d o n t h e m o d e l p a r a m e t e r s . 20 21 22 E x p l a n a t i o n o f T e s t s 23 24 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 25 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 26 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 27 T e s t 4: D o e s M o d e l 2 f i t t h e d a t a ? (A3 vs. 2) 28 29 30 T e s t s o f I n t e r e s t 31 32 T e s t -2*log(Likelihood Ratio ) D. F. p- value 33 34 T e s t 1 106.9 6 < 0.0001 35 T e s t 2 72.22 3 < 0.0001 36 T e s t 3 0.715 2 0.6994 37 T e s t 4 16.13 2 0 .000315 38 39 40 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 41 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 42 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 43 44 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 45 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e . 46 47 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 48 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 49 50 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. M o d e l 2 m a y n o t a d e q u a t e l y 51 d e s c r i b e t h e d a t a ; y o u m a y w a n t t o c o n s i d e r a n o t h e r m o d e l . 52 53 54 B e n c h m a r k D o s e C o m p u t a t i o n s : 55 56 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 57 58 R i s k T y p e = E s t i m a t e d s t a n d a r d d e v i a t i o n s f r o m c o n t r o l 59 60 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 61 62 BMD 3.75968 63 64 BMDL 2.76247 This document is a draftfor review purposes only and does not constitute Agency policy. E-53 DRAFT--DO NOT CITE OR QUOTE 1 E.2.7.3. F igure f o r Selected M odel: E x p o n en tia l (M2) Exponential Model 2 with 0.95 Confidence Level Mean Response 2 10:47 02/08 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-54 DRAFT--DO NOT CITE OR QUOTE 1 E .2.8. C rofton et al., 2005: Serum , T4 2 E .2 .8 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 8 <0.0001 516.356 1.144E+02 6.239E+01 Notes exponential (M 3) exponential (M4) b exponential (M 5) 8 < 0 .0 0 0 1 5 1 6 .3 5 6 1 .144E + 02 6 .2 3 9 E + 0 1 p o w er hit b oun d (d = 1) 7 0.942 476.449 5.190E+00 3.029E+00 6 0.912 478.234 5.757E+00 3.094E+00 H ill 6 0.972 477.450 5.724E+00 3.024E+00 linear 8 <0.0001 522.460 2.406E+02 1.761E+02 polynom ial, 8degree 8 <0.0001 522.460 2.406E+02 1.761E+02 power 8 < 0 .0 0 0 1 5 2 2 .4 6 0 2 .4 0 6 E + 0 2 1 .761E + 02 p o w er b ound hit (p ow er = 1) power, unrestricted 7 0.018 491.101 2.449E +00 3.307E-01 unrestricted (power = 0.243) a C onstant variance m odel selected (p = 0 .7 6 4 7 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2.8.2. Outputfor Selected Model: Exponential (M4) 6 Crofton et al., 2005: Serum, T4 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 8 _ C r o f t o n _ 2 0 0 5 _ T 4 _ E x p C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 M o n F e b 08 1 0 : 4 8 : 0 4 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 30 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 31 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. This document is a draftfor review purposes only and does not constitute Agency policy. E-55 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Dependent variable = Mean 4 Independent variable = Dose 5 Data are assumed to be distributed: normally 6 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 7 r h o is s e t t o 0. 8 A constant v a r i a n c e model is fit. 9 10 T o t a l n u m b e r o f d o s e g r o u p s = 10 11 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 12 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 13 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 16 M L E s o l u t i o n p r o v i d e d : E x a c t 17 18 19 I n i t i a l P a r a m e t e r V a l u e s 20 21 Variable Model 4 22 23 lnalpha 5.47437 24 rho(S) 0 25 a 1 0 4 . 9 9 9 26 b 0 . 0 0 6 4 1 8 9 5 27 c 0 . 4 4 5 7 6 4 28 d 1 29 30 (S) = S p e c i f i e d 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 Variable Model 4 37 38 lnalpha 5.50623 39 rho 0 40 a 1 0 0 . 3 3 2 41 b 0 . 0 7 6 6 7 8 42 c 0 . 5 2 3 6 2 6 43 d 1 44 45 46 T a b l e o f S t a t s F r o m I n p u t D a t a 47 48 Dose N Obs Mean Obs Std 49 50 0 14 100 15.44 51 0.0202 6 96.27 14.98 52 0.4882 12 98.57 18.11 53 1.384 6 99.76 19.04 54 3.455 6 93.32 12.11 55 9.257 6 70.94 12.74 56 23.07 6 62.52 14.75 57 65.65 6 52.68 22.73 58 180.9 6 54.66 19.71 59 583.5 4 49.15 11.15 60 61 62 E s t i m a t e d V a l u e s o f I n t e r e s t 63 64 Dose Est Mean Est Std Scaled Residual 65 66 0 100.3 15.69 -0.07952 67 0 . 0 2 0 2 100.3 15.69 -0.6231 68 0 . 4 8 8 2 98.58 15.69 -0.000744 69 1 . 3 8 4 95.52 15.69 0.6614 70 3 . 4 5 5 89.21 15.69 0.6422 This document is a draftfor review purposes only and does not constitute Agency policy. E-56 DRAFT--DO NOT CITE OR QUOTE 1 9.257 76.04 15.69 -0.7962 2 23.07 60.69 15.69 0.2854 3 65.65 52.85 15.69 -0.02621 4 180.9 52.54 15.69 0.3319 5 583.5 52.54 15.69 -0.4323 6 7 8 9 Other models for which likelihoods are calculated: 10 11 M o d e l A1: Yij = M u ( i ) + e(i j ) 12 V a r { e (ij)} = S i g m a ^ 2 13 14 M o d e l A2: Yij = M u ( i ) + e(i j ) 15 V a r { e (ij)} = S i g m a ( i ) ^ 2 16 17 M o d e l A3: Yij = M u ( i ) + e(i j ) 18 V a r { e (ij)} = e x p ( l a l p h a + l o g (mean(i)) rho) 19 20 M o d e l R: Yij = M u + e(i) 21 V a r { e (ij)} = S i g m a ^ 2 22 23 24 L i k e l i h o o d s o f I n t e r e s t 25 26 Model Log(likelihood) DF AIC 27 28 A1 -233.0774 11 48 8 . 1 5 4 9 29 A2 -230.2028 20 500.4056 30 A3 -233.0774 11 48 8 . 1 5 4 9 31 R -268.4038 2 540.8076 32 4 -234.2243 4 476.4486 33 34 35 Additive constant for all log-likelihoods = -66.16. This constant added to the 36 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 37 d e p e n d o n t h e m o d e l p a r a m e t e r s . 38 39 40 E x p l a n a t i o n o f T e s t s 41 42 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 43 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 44 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 45 46 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 47 48 49 T e s t s o f I n t e r e s t 50 51 T e s t -2*log(Likelihood Ratio) D. F. p-value 52 53 T e s t 1 76.4 18 < 0.0001 54 T e s t 2 5.749 9 0.7647 55 T e s t 3 5.749 9 0.7647 56 T e s t 6a 2.294 7 0.9418 57 58 59 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 60 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 61 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 62 63 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 64 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e . 65 66 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 67 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 68 69 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 70 t o a d e q u a t e l y d e s c r i b e t h e d a t a . This document is a draftfor review purposes only and does not constitute Agency policy. E-57 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Benchmark Dose Computations: 4 5 Specified Effect = 1.000000 6 7 Risk Type = Estimated standard deviations from control 8 9 Confidence Level = 0.950000 10 11 BMD = 5.18983 12 13 BMDL = 3.02894 14 15 16 E .2 .8 .3 . Figurefor Selected Model: Exponential (M4) Exponential Model 4 with 0.95 Confidence Level Mean Response 17 18 This document is a draftfor review purposes only and does not constitute Agency policy. E-58 DRAFT--DO NOT CITE OR QUOTE 1 E .2.9. F ranc et al., 2001: S-D R ats, R elative L iver W eight 2 E .2 .9 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x 2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes exponential (M 2) 2 0.968 234.369 7.800E+00 6.040E+00 exponential (M 3) 1 0.880 236.327 9.201E+00 6.051E+00 exponential (M 4) 1 0.580 236.610 6.365E+00 4.512E+00 exponential (M 5) 0 N /A 238.346 9.474E+00 4.425E+00 Hill 0 N /A 238.346 9.479E+00 3.004E+00 linear polynom ial, 3degree power b 2 1 1 0.858 234.610 6.365E+00 4.512E+00 0.935 236.311 8.946E+00 4.598E+00 0.839 236.346 9.474E+00 4.587E+00 a C onstant variance m odel selected (p = 0.1 0 7 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2 .9 .2 . Outputfor Selected Model: Power 6 Franc et al., 2001: S-D Rats, R elative L iver W eight 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 8 8 _ F r a n c _ 2 0 0 1 _ S D _ R e l L i v W t _ P o w e r C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 8 8 _ F r a n c _ 2 0 0 1 _ S D _ R e l L i v W t _ P o w e r C V _ 1 . p l t 13 T h u A p r 15 1 1 : 4 6 : 3 2 2 0 1 0 14 15 16 F i g u r e 5, S D r a t s , r e l a t i v e l i v e r w e i g h t 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 T h e p o w e r is r e s t r i c t e d t o b e g r e a t e r t h a n o r e q u a l t o 1 28 A c o n s t a n t v a r i a n c e m o d e l is f i t 29 30 T o t a l n u m b e r o f d o s e g r o u p s = 4 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 33 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 34 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 This document is a draftfor review purposes only and does not constitute Agency policy. E-59 DRAFT--DO NOT CITE OR QUOTE 1 2 3 4 Default Initial Parameter Values 5 alpha = 527.447 6 rho = 0S p e c i f i e d 7 control = 100 8 slope = 0.947018 9 power = 1.13144 10 11 12 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 13 14 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 15 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 16 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 17 18 alpha control slope power 19 20 a l p h a 1 -6.3e-009 5.4e-009 -4.7e-009 21 22 control -6.3e-009 1 -0.74 0.71 23 24 slope 5.4e-009 -0.74 1 -1 25 26 power -4.7e-009 0.71 -1 1 27 28 29 30 P a r a m e t e r E s t i m a t e s 31 32 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 33 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 34 alpha 462.113 115.528 235.682 688.544 35 control 100.494 7.31114 86.1645 114.824 36 slope 0.593276 1.31535 -1.98476 3.17131 37 power 1.25841 0.597816 0.086712 2.43011 38 39 40 41 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 42 43 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 44 45 46 0 8 1 0 0 100 14 21.5 - 0 . 0 6 5 47 6 . 5 8 7 8 108 107 16.9 21.5 0.158 48 1 4 . 4 8 8 117 118 25.9 21.5 -0.109 49 3 6 . 4 3 8 155 155 30.9 21.5 0.0157 50 51 52 53 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 54 55 56 M o d e l A1: Yij = Mu(i) + e(ij) 57 V a r { e i ;i j ) } = S i g m a ^ 2 58 59 M o d e l A2: Yij = Mu(i) + e(ij) 60 V a r { e i :i j ) } = S i g m a ( i ) ^2 61 62 M o d e l A3: Yij = Mu(i) + e(ij) 63 V a r { e i :i j ) } = S i g m a ^ 2 64 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 65 w e r e s p e c i f i e d b y t h e u s e r 66 67 M o d e l R: Yi = M u + e(i) 68 V a r { e ( i ) } = S i g m a ^ 2 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-60 DRAFT--DO NOT CITE OR QUOTE 1 Likelihoods of Interest 2 3 Model Log(likelihood) # Param's AIC 4 A1 -114.152281 5 238.304562 5 A2 -111.103649 8 238.207299 6 A3 -114.152281 5 238.304562 7 fitted -114.172940 4 236.345880 8 R -125.052064 2 254.104127 9 10 11 E x p l a n a t i o n o f T e s t s 12 13 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 14 (A2 vs. R) 15 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 16 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 17 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 18 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 19 20 T e s t s o f I n t e r e s t 21 22 Test -2*log(Likelihood Ratio) Test df p-val 23 24 T e s t 1 27.8968 6 <.0001 25 T e s t 2 6.09726 3 0.107 26 T e s t 3 6.09726 3 0.107 27 T e s t 4 0.0413179 1 0.8389 28 29 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 30 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 31 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 32 33 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 34 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 35 36 37 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 38 t o b e a p p r o p r i a t e h e r e 39 40 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 41 t o a d e q u a t e l y d e s c r i b e t h e d a t a 42 43 44 B e n c h m a r k D o s e C o m p u t a t i o n 45 46 S p e c i f i e d e f f e c t = 0.1 47 48 R i s k T y p e = Relative risk 49 50 C o n f i d e n c e l e v e l = 0.95 51 52 B M D = 9 . 4 7 4 0 8 53 54 55 B M D L = 4 . 5 8 7 3 56 57 This document is a draftfor review purposes only and does not constitute Agency policy. E-61 DRAFT--DO NOT CITE OR QUOTE 1 E.2.9.3. F igure f o r Selected M odel: Pow er Power Model with 0.95 Confidence Level Mean Response 2 11:46 04/15 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-62 DRAFT--DO NOT CITE OR QUOTE 1 E .2.10. F ranc et al., 2001: L-E R ats, R elative L iver W eight 2 E .2 .1 0 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x 2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes exponential (M 2) 2 0.441 208.974 1.708E+01 1.098E+01 exponential (M 3) 2 0.441 2 0 8 .9 7 4 1.708E +01 1.098E +01 p ow er hit b ound (d = 1) exponential (M 4) 1 0.785 209.408 7.997E+00 2.601E+00 exponential (M 5) 1 Hill b 1 0.785 0.829 2 0 9 .4 0 8 7 .9 9 7 E + 0 0 2 .6 0 1 E + 0 0 p ow er hit b ound (d = 1) 2 0 9 .3 8 1 7 .7 2 5 E + 0 0 1 .225E + 00 n lo w er b oun d hit (n = 1) linear 2 0.499 208.725 1.570E+01 9.619E+00 polynom ial, 3degree 1 <0.0001 10.000 8.604E +00 error power 2 0.499 2 0 8 .7 2 5 1.570E +01 9 .6 1 9 E + 0 0 p ow er b oun d hit (p ow er = 1) H ill, unrestricted c0 N /A 211.337 7.217E +00 1.147E+00 unrestricted (n = 0.545) power, unrestricted 1 0.965 209.336 7.193E +00 error a N on-constant variance m odel selected (p = 0 .0632) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .1 0 .2 . Outputfor Selected Model: Hill unrestricted (power = 0.524) 6 Franc et al., 2001: L-E Rats, R elative Liver W eight 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 8 9 _ F r a n c _ 2 0 0 1 _ L E _ R e l L i v W t _ H i l l _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 8 9 _ F r a n c _ 2 0 0 1 _ L E _ R e l L i v W t _ H i l l _ 1 . p l t 13 T h u A p r 15 1 1 : 4 8 : 4 4 2 0 1 0 14 15 16 F i g u r e 5, L - E r a t s , r e l a t i v e l i v e r w e i g h t 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 P o w e r p a r a m e t e r r e s t r i c t e d t o b e g r e a t e r t h a n 1 This document is a draftfor review purposes only and does not constitute Agency policy. E-63 DRAFT--DO NOT CITE OR QUOTE 1 The variance is to be modeled as Var(i) = exp(lalpha + rho ln(mean(i))) 2 3 Total number of dose groups = 4 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 12 lalpha 5.41581 13 rho 0 14 intercept 100 15 2 2 . 2 2 5 16 0 . 4 4 3 1 5 5 17 1 8 . 7 4 6 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( * * * T h e m o d e l p a r a m e t e r ( s ) -n 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 24 a n d d o n o t a p p e a r in t h e c o r r e l a t i o n m a t r i x ) 25 26 lalpha rho intercept v k 27 28 l a l p h a 1 -1 -0.21 0.33 0.18 29 CO <--1 o CO CO o CO CO CO <--1 oo 30 rh o -1 1 0.21 31 32 i n t e r c e p t -0.21 0.21 1 0.028 0.35 33 34 v 0 . 3 3 0.028 1 0.91 35 36 k 0 . 1 8 0.35 0.91 1 37 38 39 40 P a r a m e t e r E s t i m a t e s 41 42 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 43 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 44 lalpha -17.2754 17.3066 -51.1957 16.6449 45 rho 4.77884 3.67625 -2.42648 11.9842 46 i n t e r c e p t 99.5348 3.61286 92.4538 106.616 47 v 36.3963 24.1862 -11.0079 83.8004 48 n 1 N A 49 k 20.5223 28.2566 -34.8596 75.9042 50 51 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 52 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 53 h a s n o s t a n d a r d e r r o r . 54 55 56 57 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 58 59 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 60 61 62 0 8 1 0 0 99.5 10 10.5 0.125 63 6 . 5 8 4 8 106 108 17.9 12.9 -0.455 64 1 4 . 4 7 8 117 115 8.97 14.8 0.426 65 3 6 . 4 1 8 122 123 19.9 17.4 -0.0954 66 67 68 69 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-64 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A1: Yij Mu(i) + e(ij) 3 Var{e(ij)} Sigma^2 4 5 Model A2: Yij Mu(i) + e(ij) 6 Var{e(ij)} Sigma(i)^2 7 8 Model A3: Yij = Mu(i) + e(ij) 9 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 10 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 11 w e r e s p e c i f i e d b y t h e u s e r 12 13 M o d e l R: Yi = M u + e(i) 14 V a r { e ( i ) } = S i g m a ^ 2 15 16 17 L i k e l i h o o d s o f I n t e r e s t 18 19 Model Log(likelihood) # Param's AIC 20 A1 -100.516456 5 211.032912 21 A2 -96.870820 8 209.741641 22 A3 -99.666984 6 211.333969 23 fitted -99.690373 5 209.380746 24 R -105.717087 2 215.434174 25 26 27 E x p l a n a t i o n o f T e s t s 28 29 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 30 (A2 vs. R) 31 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 32 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 33 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 34 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 35 36 T e s t s o f I n t e r e s t 37 38 Test -2*log(Likelihood Ratio) Test df p-value 39 40 T e s t 1 17.6925 6 0.007048 41 T e s t 2 7.29127 3 0.06317 42 T e s t 3 5.59233 2 0.06104 43 T e s t 4 0.0467774 1 0.8288 44 45 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 46 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 47 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 48 49 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 50 m o d e l a p p e a r s t o b e a p p r o p r i a t e 51 52 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t t o c o n s i d e r a 53 d i f f e r e n t v a r i a n c e m o d e l 54 55 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 56 t o a d e q u a t e l y d e s c r i b e t h e d a t a 57 58 59 B e n c h m a r k D o s e C o m p u t a t i o n 60 61 S p e c i f i e d e f f e c t = 0.1 62 63 R i s k T y p e = Relative risk 64 65 C o n f i d e n c e l e v e l = 0.95 66 67 BMD = 7.72492 68 69 BMDL = 1.22451 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-65 DRAFT--DO NOT CITE OR QUOTE 1 E.2.10.3. F igure f o r Selected M odel: H ill Hill Model with 0.95 Confidence Level 2 3 4 5 E .2.10.4. Outputfor Additional Model Presented: Hill, Unrestricted 6 Franc et al., 2001: L-E Rats, R elative Liver W eight 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 8 9 _ F r a n c _ 2 0 0 1 _ L E _ R e l L i v W t _ H i l l _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 8 9 _ F r a n c _ 2 0 0 1 _ L E _ R e l L i v W t _ H i l l _ U _ 1 . p l t 13 T h u A p r 15 1 1 : 4 8 : 5 0 2 0 1 0 14 15 16 F i g u r e 5, L - E r a t s , r e l a t i v e l i v e r w e i g h t 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 P o w e r p a r a m e t e r is n o t r e s t r i c t e d 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + r h o * l n ( m e a n ( i ) ) ) 28 29 T o t a l n u m b e r o f d o s e g r o u p s = 4 This document is a draftfor review purposes only and does not constitute Agency policy. E -66 D RAFT-- DO N O T CITE OR QUOTE 1 Total number of records with missing values = 0 2 Maximum number of iterations = 250 3 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 4 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 5 6 7 8 Default Initial Parameter Values 9 lalpha 5.41581 10 rho 0 11 intercept 100 12 2 2 . 2 2 5 13 0 . 4 4 3 1 5 5 14 1 8 . 7 4 6 15 16 17 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 18 19 lalpha rho intercept v n k 20 21 l a l p h a 1 -1 -0.22 -0.14 0.24 -0.15 22 23 rh o -1 1 0.22 0.14 -0.24 0.15 24 25 i n t e r c e p t -0.22 0.22 1 0.022 0.11 0.013 26 27 v -0.14 0.14 0.022 1 -0.9 1 28 29 n 0.24 -0.24 0.11 -0.9 1 -0.92 30 31 k -0.15 0.15 0.013 1 -0.92 1 32 33 34 35 P a r a m e t e r E s t i m a t e s 36 37 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 38 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 39 lalpha -19.2405 18.21 -54.9315 16.4505 40 rho 5.19575 3.86861 -2.38657 12.7781 41 i n t e r c e p t 99.5348 3.51796 92.6398 106.43 42 v 440.285 13708.5 -26427.9 27308.5 43 n 0.544741 0.730981 -0.887956 1.97744 44 k 7266.27 485402 -944104 958637 45 46 47 48 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 49 50 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 51 52 53 0 8 1 0 0 99.5 10 10.3 0.128 54 6 . 5 8 4 8 106 109 17.9 13 -0.589 55 1 4 . 4 7 8 117 114 8.97 14.6 0.558 56 3 6 . 4 1 8 122 123 19.9 17.8 -0.0957 57 58 D e g r e e s o f f r e e d o m f o r T e s t A 3 v s f i t t e d < = 0 59 60 61 62 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 63 64 65 M o d e l A1: Yij = Mu(i) + e(ij) 66 V a r { e i ;i j ) } = S i g m a ^ 2 67 68 M o d e l A2: Yij = Mu(i) + e(ij) 69 V a r { e i :i j ) } = S i g m a ( i ) ^ 2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-67 DRAFT--DO NOT CITE OR QUOTE 1 Model A3: Yij = Mu(i) + e(ij) 2 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 3 Model A3 uses any fixed variance parameters that 4 were specified by the user 5 6 M o d e l R: Yi = M u + e(i) 7 Var{e(i)} = Sigma^2 8 9 10 L i k e l i h o o d s o f I n t e r e s t 11 12 Model Log(likelihood) # Param's AIC 13 A1 -100.516456 5 211.032912 14 A2 -96.870820 8 209.741641 15 A3 -99.666984 6 211.333969 16 fitted -99.668321 6 211.336641 17 R -105.717087 2 215.434174 18 19 20 E x p l a n a t i o n o f T e s t s 21 22 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 23 (A2 vs. R) 24 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 25 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 26 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 27 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 28 29 T e s t s o f I n t e r e s t 30 31 Test -2*log(Likelihood Ratio) Test df p-value 32 33 T e s t 1 17.6925 6 0.007048 34 T e s t 2 7.29127 3 0.06317 35 T e s t 3 5.59233 2 0.06104 36 T e s t 4 0.00267242 0 NA 37 38 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 39 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 40 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 41 42 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 43 m o d e l a p p e a r s t o b e a p p r o p r i a t e 44 45 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t t o c o n s i d e r a 46 d i f f e r e n t v a r i a n c e m o d e l 47 48 N A D e g r e e s o f f r e e d o m f o r T e s t 4 a r e l e s s t h a n o r e q u a l t o 0. T h e C h i - S q u a r e 49 t e s t f o r f i t is n o t v a l i d 50 51 52 B e n c h m a r k D o s e C o m p u t a t i o n 53 54 S p e c i f i e d e f f e c t = 0.1 55 56 R i s k T y p e = Relative risk 57 58 C o n f i d e n c e l e v e l 0.95 59 60 BMD 7.21718 61 62 BMDL 1.14742 63 64 This document is a draftfor review purposes only and does not constitute Agency policy. E-68 DRAFT--DO NOT CITE OR QUOTE 1 E.2.10.5. F igure f o r A d d itio n a l M o d el Presented: H ill, Unrestricted Hill Model with 0.95 Confidence Level Mean Response 2 11:48 04/15 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-69 DRAFT--DO NOT CITE OR QUOTE 1 E.2.11. Franc et al., 2001: S-D Rats, Relative Thymus Weight 2 E.2.11.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x 2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes exponential (M 2) 2 0.814 285.107 2.478E+00 1.535E+00 exponential (M 3) 1 exponential (M4) b 1 exponential (M 5) 0 0.016 292.452 3.173E+01 1.007E+00 0.720 286.825 1.878E+00 9.221E-01 N /A 288.696 3.296E+00 9.365E-01 H ill 0 N /A 288.696 3.625E+00 6.199E-01 linear 2 0.404 286.508 4.783E+00 3.893E+00 polynom ial, 3degree c 2 0.404 286.508 4.783E+00 3.893E+00 power 2 0 .4 0 4 2 8 6 .5 0 8 4 .7 8 3 E + 0 0 3 .8 9 3 E + 0 0 p o w er b oun d hit (p ow er = 1) power, unrestricted 1 0.483 287.189 6.795E-01 3.271E-03 unrestricted (pow er = 0.515) a N on-constant variance m odel selected (p = 0 .0320) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .1 1 .2 . Outputfor Selected Model: Exponential (M4) 6 Franc et al., 2001: S-D Rats, R elative Thym us W eight 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 9 1 _ F r a n c _ 2 0 0 1 _ S D _ R e l T h y W t _ E x p _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 T h u A p r 15 1 1 : 5 1 : 1 9 2 0 1 0 14 15 16 F i g u r e 5, S D r a t s , r e l a t i v e t h y m u s w e i g h t 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. This document is a draftfor review purposes only and does not constitute Agency policy. E-70 DRAFT--DO NOT CITE OR QUOTE 1 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 2 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 3 4 5 Dependent variable = Mean 6 Independent variable = Dose 7 Data are assumed to be distributed: normally 8 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 9 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) * rho) 10 11 T o t a l n u m b e r o f d o s e g r o u p s = 4 12 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 13 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 14 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 16 17 M L E s o l u t i o n p r o v i d e d : E x a c t 18 19 20 I n i t i a l P a r a m e t e r V a l u e s 21 22 Variable Model 4 23 24 lnalpha 3.35464 25 rho 1.08199 26 a 1 0 5 27 b 0 . 0 5 6 9 9 7 9 28 c 0 . 1 0 8 5 3 1 29 d 1 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 Variable Model 4 36 37 lnalpha 2.4312 38 rho 1.28672 39 a 1 1 0 . 9 5 9 40 b 0 . 0 6 6 3 4 9 8 41 c 0 . 1 4 6 4 8 6 42 d 1 43 44 45 T a b l e o f S t a t s F r o m I n p u t D a t a 46 47 Dose N Obs Mean Obs Std Dev 48 49 08 100 83.2 50 6.587 8 91.17 47.97 51 14.48 8 51.41 43.48 52 36.43 8 22.79 29.98 53 54 55 E s t i m a t e d V a l u e s o f I n t e r e s t 56 57 Dose Est Mean Est Std Scaled Residual 58 59 0 111 69.78 -0.4442 60 6 . 5 8 7 77.43 55.36 0.7019 61 1 4 . 4 8 52.49 43.11 -0.0709 62 3 6 . 4 3 24.7 26.54 -0.2031 63 64 65 66 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d 67 68 M o d e l A1: Yij = M u ( i ) + e(i j ) 69 V a r { e ( i j ) } = S i g m a ^ 2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-71 DRAFT--DO NOT CITE OR QUOTE 1 Model A2: Yij Mu(i) + e(ij) 2 Var{e(ij)} Sigma(i)^2 3 4 Model A3: Yij Mu(i) + e(ij) 5 Var{e(ij)} e x p ( l a l p h a + log(mean(i)) * rho) 6 7 M o d e l R: Yij M u + e(i) 8 Var{e(ij)} Sigma^2 9 10 11 L i k e l i h o o d s o f I n t e r e s t 12 13 Model Log(likelihood) DF AIC 14 15 A1 -141.9834 5 293.9669 16 A2 -137.5818 8 291.1637 17 A3 -138.3482 6 288.6964 18 R -146.9973 2 297.9946 19 4 -138.4123 5 286.8245 20 21 22 Additive constant for all log-likelihoods = -29.41. This constant added to the 23 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 24 d e p e n d o n t h e m o d e l p a r a m e t e r s . 25 26 27 E x p l a n a t i o n o f T e s t s 28 29 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 30 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 31 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 32 33 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 34 35 36 T e s t s o f I n t e r e s t 37 38 T e s t -2*log(Likelihood Ratio) D. F. p-value 39 40 T e s t 1 18.83 6 0.004459 41 T e s t 2 8.803 3 0.03203 42 T e s t 3 1.533 2 0.4647 43 T e s t 6a 0.1282 1 0.7203 44 45 46 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 47 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 48 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 49 50 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 51 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e . 52 53 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 54 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 55 56 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 57 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 58 59 60 B e n c h m a r k D o s e C o m p u t a t i o n s : 61 62 S p e c i f i e d E f f e c t = 0 . 1 0 0 0 0 0 63 64 R i s k T y p e = R e l a t i v e d e v i a t i o n 65 66 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 67 68 BMD = 1.87814 69 70 BMDL = 0.922136 This document is a draftfor review purposes only and does not constitute Agency policy. E-72 DRAFT--DO NOT CITE OR QUOTE 1 E .2 .1 1 .3 . Figurefor Selected Model: Exponential (M4) Exponential_beta Model 4 with 0.95 Confidence Level 2 3 4 5 E .2.11.4. Outputfor Additional Model Presented: Polynomial, 3-degree 6 Franc et al., 2001: S-D Rats, R elative Thym us W eight 7 8 9 10 P o l y n o m i a l M o d e l . ( V e r s i o n : 2 . 1 3 ; D a t e : 0 4 / 0 8 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 9 1 _ F r a n c _ 2 0 0 1 _ S D _ R e l T h y W t _ P o l y _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 9 1 _ F r a n c _ 2 0 0 1 _ S D _ R e l T h y W t _ P o l y _ 1 . p l t 13 T h u A p r 15 1 1 : 5 1 : 2 0 2 0 1 0 14 15 16 F i g u r e 5, S D r a t s , r e l a t i v e t h y m u s w e i g h t 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 T h e p o l y n o m i a l c o e f f i c i e n t s a r e r e s t r i c t e d t o b e n e g a t i v e 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 28 29 T o t a l n u m b e r o f d o s e g r o u p s = 4 This document is a draftfor review purposes only and does not constitute Agency policy. E-73 DRAFT--DO NOT CITE OR QUOTE 1 Total number of records with missing values = 0 2 Maximum number of iterations = 250 3 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 4 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 5 6 7 8 Default Initial Parameter Values 9 lalpha = 8.0075 10 rho = 0 11 beta_0 = 100 12 beta_1 = 0 13 beta_2 = -0.475283 14 beta 3 = 0 15 16 17 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 18 19 ( *** The model parameter(s) -beta 2 -beta 3 20 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 21 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 22 23 lalpha rho beta_0 beta_1 24 25 l a l p h a 1 -0.99 0.018 0.0095 26 27 rho -0.99 1 -0.022 -0.0024 28 rco o rco o 29 b e t a _ 0 0.018 -0.022 1 30 31 beta_1 0.0095 -0.0024 1 32 33 34 35 P a r a m e t e r E s t i m a t e s 36 37 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 38 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 39 lalpha 2.8315 1.71297 -0.525852 6.18885 40 rho 1.19884 0.416889 0.381756 2.01593 41 beta_0 94.5944 14.6685 65.8446 123.344 42 beta_1 -1.97776 0.509904 -2.97715 -0.978362 43 b e t a _ 2 0 NA 44 b e t a _ 3 0 NA 45 46 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 47 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 48 h a s n o s t a n d a r d e r r o r . 49 50 51 52 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 53 54 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 55 56 57 0 8 1 0 0 94.6 83.2 63 0.243 58 6 . 5 8 7 8 91.2 81.6 48 57.6 0.471 59 1 4 . 4 8 8 51.4 66 43.5 50.7 -0.811 60 3 6 . 4 3 8 22.8 22.5 30 26.7 0.0269 61 62 63 64 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 65 66 67 M o d e l A1: Yij = Mu(i) + e(ij) 68 V a r { e ( i j ) } = S i g m a ^ 2 69 70 M o d e l A2: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-74 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma(i)^2 2 3 Model A3: Yij = Mu(i) + e(ij) 4 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 5 Model A3 uses any fixed variance parameters that 6 were specified by the user 7 8 M o d e l R: Yi = M u + e(i) 9 Var{e(i)} = Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) # Param's AIC 15 A1 -141.983433 5 293.966865 16 A2 -137.581833 8 291.163667 17 A3 -138.348184 6 288.696368 18 fitted -139.254163 4 286.508326 19 R -146.997301 2 297.994602 20 21 22 E x p l a n a t i o n o f T e s t s 23 24 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 25 (A2 vs. R) 26 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 27 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 28 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 29 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 30 31 T e s t s o f I n t e r e s t 32 33 Test -2*log(Likelihood Ratio) Test df p-value 34 35 T e s t 1 18.8309 6 0.004459 36 T e s t 2 8.8032 3 0.03203 37 T e s t 3 1.5327 2 0.4647 38 T e s t 4 1.81196 2 0.4041 39 40 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 41 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 42 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 43 44 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 45 m o d e l a p p e a r s t o b e a p p r o p r i a t e 46 47 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 48 t o b e a p p r o p r i a t e h e r e 49 50 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 51 t o a d e q u a t e l y d e s c r i b e t h e d a t a 52 53 54 B e n c h m a r k D o s e C o m p u t a t i o n 55 56 S p e c i f i e d e f f e c t = 0.1 57 58 R i s k T y p e = Relative risk 59 60 C o n f i d e n c e l e v e l = 0.95 61 62 BMD = 4.78292 63 64 65 BMDL = 3.8932 66 67 This document is a draftfor review purposes only and does not constitute Agency policy. E-75 DRAFT--DO NOT CITE OR QUOTE 1 E.2.11.5. F igure f o r A d d itio n a l M o d el Presented: Polynom ial, 3-degree Polynomial Model with 0.95 Confidence Level Mean Response 2 11:51 04/15 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-76 DRAFT--DO NOT CITE OR QUOTE 1 E.2.12. Franc et al., 2001: L-E Rats, Relative Thymus Weight 2 E.2.12.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x 2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes exponential (M 2) 2 0.440 301.449 2.726E+00 1.212E+00 exponential (M 3) 2 exponential (M4) b 1 exponential (M 5) 0 0 .4 4 0 3 0 1 .4 4 9 2 .7 2 6 E + 0 0 1 .212E + 00 p o w er hit bound (d = 1) 0.227 303.266 2.084E+00 5.926E-01 N /A 303.805 7.859E+00 9.801E-01 H ill 0 N /A 303.805 7.480E+00 7.512E-01 linear 2 0.304 302.186 5.045E+00 3.349E+00 polynom ial, 3degree 2 0.304 302.186 5.045E+00 3.349E+00 power 2 0 .3 0 4 3 0 2 .1 8 6 5 .0 4 5 E + 0 0 3 .3 4 9 E + 0 0 p o w er b oun d hit (p ow er = 1) power, unrestricted 1 0.168 303.710 1.374E+00 9.032E-09 unrestricted (pow er = 0.601) a C onstant variance m odel selected (p = 0 .5 0 6 3 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 E .2 .1 2 .2 . Outputfor Selected Model: Exponential (M4) 5 Franc et al., 2001: L-E Rats, R elative Thym us W eight 6 7 8 9 Exponential Model. (Version: 1.61; Date: 7/24/2009) 10 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 9 2 _ F r a n c _ 2 0 0 1 _ L E _ R e l T h y W t _ E x p C V _ 1 . ( d ) 11 G n u p l o t P l o t t i n g Fil e : 12 T h u A p r 15 1 1 : 5 3 : 3 7 2 0 1 0 13 14 15 F i g u r e 5, L - E r a t s , r e l a t i v e t h y m u s w e i g h t 16 17 18 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 19 M o d e l 2: Y dose] = a * exp{sign * b * dose} 20 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 21 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 22 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 23 24 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 25 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 26 s i g n = -1 f o r d e c r e a s i n g t r e n d . 27 28 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 29 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 30 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 31 This document is a draftfor review purposes only and does not constitute Agency policy. E-77 DRAFT--DO NOT CITE OR QUOTE 1 2 Dependent variable = Mean 3 Independent variable = Dose 4 Data are assumed to be distributed: normally 5 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 6 r h o is s e t t o 0. 7 A constant v a r i a n c e model is fit. 8 9 Total number of dose groups = 4 10 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 11 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 12 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 13 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 15 M L E s o l u t i o n p r o v i d e d : E x a c t 16 17 18 I n i t i a l P a r a m e t e r V a l u e s 19 20 Variable Model 4 21 22 lnalpha 8.1814 23 rho(S) 0 24 a 1 0 5 25 b 0 . 0 5 0 6 1 6 8 26 c 0 . 1 6 6 5 8 2 27 d 1 28 29 (S) = S p e c i f i e d 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 Variable Model 4 36 37 lnalpha 8.22706 38 rho 0 39 a 1 0 5 . 9 7 7 40 b 0 . 0 6 6 0 0 4 2 41 c 0 . 2 2 1 7 8 6 42 d 1 43 44 45 T a b l e o f S t a t s F r o m I n p u t D a t a 46 47 Dose N Obs Mean Obs Std Dev 48 49 08 100 54.72 50 6.584 8 95.41 70.46 51 14.47 8 38.69 47.97 52 36.41 8 34.98 77.96 53 54 55 E s t i m a t e d V a l u e s o f I n t e r e s t 56 57 Dose Est Mean Est Std Scaled Residual 58 59 0 106 61.16 -0.2764 60 6 . 5 8 4 76.91 61.16 0.8555 61 1 4 . 4 7 55.24 61.16 -0.765 62 3 6 . 4 1 30.96 61.16 0.186 63 64 65 66 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d 67 68 M o d e l A1: Yij = M u ( i ) + e(i j ) 69 V a r { e ( i j ) } = S i g m a ^ 2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-78 DRAFT--DO NOT CITE OR QUOTE 1 Model A2: Yij Mu(i) + e(ij) 2 Var{e(ij)} Sigma(i)^2 3 4 Model A3: Yij Mu(i) + e(ij) 5 Var{e(ij)} e x p ( l a l p h a + log(mean(i)) * rho) 6 7 M o d e l R: Yij M u + e(i) 8 Var{e(ij)} Sigma^2 9 10 11 L i k e l i h o o d s o f I n t e r e s t 12 13 Model Log(likelihood) DF AIC 14 15 A1 -146.9024 5 303.8049 16 A2 -145.7361 8 307.4723 17 A3 -146.9024 5 303.8049 18 R -150.6049 2 305.2098 19 4 -147.6329 4 303.2658 20 21 22 Additive constant for all log-likelihoods = -29.41. This constant added to the 23 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 24 d e p e n d o n t h e m o d e l p a r a m e t e r s . 25 26 27 E x p l a n a t i o n o f T e s t s 28 29 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 30 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 31 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 32 33 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 34 35 36 T e s t s o f I n t e r e s t 37 38 T e s t -2*log(Likelihood Ratio) D. F. p-value 39 40 T e s t 1 9.738 6 0.1362 41 T e s t 2 2.333 3 0.5063 42 T e s t 3 2.333 3 0.5063 43 T e s t 6a 1.461 1 0.2268 44 45 46 T h e p - v a l u e f o r T e s t 1 is g r e a t e r t h a n .05. T h e r e m a y n o t b e a 47 d i f f e n c e b e t w e e n r e s p o n s e s a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 48 M o d e l l i n g t h e d a t a w i t h a d o s e / r e s p o n s e c u r v e m a y n o t b e a p p r o p r i a t e . 49 50 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 51 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e . 52 53 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 54 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 55 56 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 57 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 58 59 60 B e n c h m a r k D o s e C o m p u t a t i o n s : 61 62 S p e c i f i e d E f f e c t = 0 . 1 0 0 0 0 0 63 64 R i s k T y p e = R e l a t i v e d e v i a t i o n 65 66 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 67 68 BMD = 2.08379 69 70 BMDL = 0.592601 This document is a draftfor review purposes only and does not constitute Agency policy. E-79 DRAFT--DO NOT CITE OR QUOTE 1 E.2.12.3. F igure f o r Selected M odel: E x p o n en tia l (M4) Exponential_beta Model 4 with 0.95 Confidence Level Mean Response 2 11:53 04/15 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-80 DRAFT--DO NOT CITE OR QUOTE 1 E.2.13. Franc et al., 2001: H/W Rats, Relative Thymus Weight 2 E.2.13.1. S u m m a ry Table o f B M D S M odeling R esults Model a exponential (M2) b Degrees of Freedom x 2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes 2 0.698 261.646 5.094E+00 3.132E+00 exponential (M 3) 1 0.407 263.616 5.944E+00 3.140E+00 exponential (M 4) 1 0.396 263.646 5.063E+00 1.864E+00 exponential (M 5) 0 N /A 264.927 9.945E+00 2.127E+00 H ill 0 N /A 264.927 9.638E+00 1.853E+00 linear 2 0.645 261.804 6.874E+00 5.006E+00 polynom ial, 3degree 2 0.645 261.804 6.874E+00 5.006E+00 power 2 0.6 4 5 2 6 1 .8 0 4 6 .8 7 4 E + 0 0 5 .0 0 6 E + 0 0 p o w er b oun d hit (p ow er = 1) power, unrestricted 1 0.363 263.755 5.487E+00 2.573E-01 unrestricted (pow er = 0.881) a C onstant variance m odel selected (p = 0 .4 3 3 1 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 E .2 .1 3 .2 . Outputfor Selected Model: Exponential (M2) 5 Franc et al., 2001: H /W Rats, R elative Thym us W eight 6 7 8 9 Exponential Model. (Version: 1.61; Date: 7/24/2009) 10 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 9 3 _ F r a n c _ 2 0 0 1 _ H W _ R e l T h y W t _ E x p C V _ 1 . ( d ) 11 G n u p l o t P l o t t i n g Fil e : 12 T h u A p r 15 1 1 : 5 5 : 5 5 2 0 1 0 13 14 15 F i g u r e 5, H / W r a t s , r e l a t i v e t h y m u s w e i g h t 16 17 18 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 19 M o d e l 2: Y dose] = a * exp{sign * b * dose} 20 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 21 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 22 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 23 24 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 25 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 26 s i g n = -1 f o r d e c r e a s i n g t r e n d . 27 28 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 29 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 30 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 31 This document is a draftfor review purposes only and does not constitute Agency policy. E-81 DRAFT--DO NOT CITE OR QUOTE 1 2 Dependent variable = Mean 3 Independent variable = Dose 4 Data are assumed to be distributed: normally 5 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 6 r h o is s e t t o 0. 7 A constant v a r i a n c e model is fit. 8 9 Total number of dose groups = 4 10 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 11 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 12 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 13 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 15 M L E s o l u t i o n p r o v i d e d : E x a c t 16 17 18 I n i t i a l P a r a m e t e r V a l u e s 19 20 Variable Model 2 21 22 lnalpha 6.96647 23 rho(S) 0 24 a 5 6 . 9 4 3 3 25 b 0 . 0 2 0 4 8 0 6 26 c 0 27 d 1 28 29 (S) = S p e c i f i e d 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 Variable Model 2 36 37 lnalpha 6.98895 38 rho 0 39 a 1 0 3 . 0 4 7 40 b 0 . 0 2 0 6 8 2 8 41 c 0 42 d 1 43 44 45 T a b l e o f S t a t s F r o m I n p u t D a t a 46 47 Dose N Obs Mean Obs Std Dev 48 49 08 100 35.98 50 6.588 8 97.53 32.98 51 14.48 8 71.02 23.99 52 36.44 8 49.29 43.48 53 54 55 E s t i m a t e d V a l u e s o f I n t e r e s t 56 57 Dose Est Mean Est Std Scaled Residual 58 59 0 103 32.93 -0.2617 60 6 . 5 8 8 89.92 32.93 0.6532 61 1 4 . 4 8 76.38 32.93 -0.4596 62 3 6 . 4 4 48.49 32.93 0.06871 63 64 65 66 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d 67 68 M o d e l A1: Yij = M u ( i ) + e(i j ) 69 V a r { e ( i j ) } = S i g m a ^ 2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-82 DRAFT--DO NOT CITE OR QUOTE 1 Model A2: Yij = Mu(i) + e(ij) 2 V a r { e i ;ij)} = S i g m a ( i ) ^ 2 3 4 Model A3: Yij = Mu(i) + e(ij) 5 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 6 7 M o d e l R: Yij = M u + e(i) 8 V a r { e i :ij)} = S i g m a ^ 2 9 10 11 L i k e l i h o o d s o f I n t e r e s t 12 13 Model Log(likelihood) DF AIC 14 15 A1 -127.4636 5 264.9271 16 A2 -126.0925 8 268.185 17 A3 -127.4636 5 264.9271 18 R -132.935 2 269.87 19 2 -127.8231 3 261.6463 20 21 22 Additive constant for all log-likelihoods = -29.41. This constant added to the 23 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 24 d e p e n d o n t h e m o d e l p a r a m e t e r s . 25 26 27 E x p l a n a t i o n o f T e s t s 28 29 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 30 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 31 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 32 T e s t 4: D o e s M o d e l 2 f i t t h e d a t a ? (A3 vs. 2) 33 34 35 T e s t s o f I n t e r e s t 36 37 T e s t -2*log(Likelihood Ratio) D. F. p-value 38 39 T e s t 1 13.69 6 0.03336 40 T e s t 2 2.742 3 0.4331 41 T e s t 3 2.742 3 0.4331 42 T e s t 4 0.7192 2 0.698 43 44 45 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 46 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 47 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 48 49 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 50 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e . 51 52 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 53 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 54 55 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. M o d e l 2 s e e m s 56 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 57 58 59 B e n c h m a r k D o s e C o m p u t a t i o n s : 60 61 S p e c i f i e d E f f e c t = 0 . 1 0 0 0 0 0 62 63 R i s k T y p e = R e l a t i v e d e v i a t i o n 64 65 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 66 67 BMD = 5.09411 68 69 BMDL = 3.13214 This document is a draftfor review purposes only and does not constitute Agency policy. E-83 DRAFT--DO NOT CITE OR QUOTE 1 E.2.13.3. F igure f o r Selected M odel: E x p o n en tia l (M2) Exponential_beta Model 2 with 0.95 Confidence Level Mean Response 3 4 5 This document is a draftfor review purposes only and does not constitute Agency policy. E-84 DRAFT--DO NOT CITE OR QUOTE 1 E .2.14. H ojo et al., 2002: D R L R einforce P er M inute 2 E .2 .1 4 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes H ill 1 0.101 4.465 1.667E+00 6.209E -08 n upper bound hit (n = 18) linear polynom ial, 3degree power power, unrestricted exponential (M 2) 2 0.009 9.124 1.352E+01 6.020E+00 2 0.009 9.124 1.352E+01 6.020E+00 2 0.009 9.124 1.352E +01 6 .0 2 0 E + 0 0 p ow er b oun d hit (p ow er = 1) 1 0.025 6.780 2.428E-01 1.070E-14 unrestricted (power = 0.103) 2 0.007 9.612 1.623E+01 8.673E+00 exponential (M 3) exponential (M4) b exponential (M 5) 2 0.007 9.612 1.623E +01 8 .6 7 3 E + 0 0 p ow er hit bound (d = 1) 1 0.054 5.488 1.316E+00 2.367E-03 0 N /A 6.465 1.728E+00 9.452E-03 a C onstant variance m odel selected (p = 0 .4 3 2 1 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2 .1 4 .2 . Outputfor Selected Model: Exponential (M4) 6 H ojo et al., 2002: D R L R einforce Per M inute 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 2 1 _ H o j o _ 2 0 0 2 _ D R L r e i n _ E x p C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 M o n F e b 08 1 0 : 4 9 : 0 8 2 0 1 0 14 15 16 T a b l e 5, v a l u e s a d j u s t e d b y a c o n s t a n t t o a l l o w e x p o n e n t i a l m o d e l 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 30 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 31 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. This document is a draftfor review purposes only and does not constitute Agency policy. E-85 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Dependent variable = Mean 4 Independent variable = Dose 5 Data are assumed to be distributed: normally 6 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 7 r h o is s e t t o 0. 8 A constant v a r i a n c e model is fit. 9 10 T o t a l n u m b e r o f d o s e g r o u p s = 4 11 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 12 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 13 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 16 M L E s o l u t i o n p r o v i d e d : E x a c t 17 18 19 I n i t i a l P a r a m e t e r V a l u e s 20 21 Variable Model 4 22 23 lnalpha -1.29672 24 rho(S) 0 25 a 0 . 0 8 1 7 26 b 0 . 1 5 6 4 2 27 c 1 6 . 3 7 3 3 28 d 1 29 30 (S) = S p e c i f i e d 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 Variable Model 4 37 38 lnalpha -1.11961 39 rho 0 40 a 0 . 0 5 4 7 4 5 2 41 b 0 . 7 0 8 1 5 4 42 c 1 8 . 2 1 4 43 d 1 44 45 46 T a b l e o f S t a t s F r o m I n p u t D a t a 47 48 Dose N Obs Mean Obs Std Dev 49 50 05 0.086 0.448 51 1.625 5 0.536 0.821 52 4.169 6 1.274 0.54 53 10.7 5 0.737 0.443 54 55 56 E s t i m a t e d V a l u e s o f I n t e r e s t 57 58 Dose Est Mean Est Std Scaled Residual 59 60 0 0.05475 0.5713 0.1223 61 1.625 0.6989 0.5713 -0.6375 62 4.169 0.9479 0.5713 1.398 63 10.7 0.9966 0.5713 -1.016 64 65 66 67 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d : 68 69 Model A1: Yij = Mu(i) + e(ij) 70 V a r { e ( i j ) } = S i g m a ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-86 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A2: Yij Mu(i) + e(ij) 3 Var{e(ij)} Sigma(i)^2 4 5 Model A3: Yij Mu(i) + e(ij) 6 Var{e(ij)} e x p ( l a l p h a + log(mean(i)) * rho) 7 8 M o d e l R: Yij M u + e(i) 9 Var{e(ij)} Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) DF AIC 15 16 A1 3.11555 5 3.7689 17 A2 4.489557 8 7.020886 18 A3 3.11555 5 3.7689 19 R -2.435087 2 8.870174 20 4 1.255891 4 5.488219 21 22 23 Additive constant for all log-likelihoods = -19.3. This constant added to the 24 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 25 d e p e n d o n t h e m o d e l p a r a m e t e r s . 26 27 28 E x p l a n a t i o n o f T e s t s 29 30 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 31 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 32 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 33 34 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 35 36 37 T e s t s o f I n t e r e s t 38 39 T e s t -2*log(Likelihood Ratio) D. F. p-value 40 41 T e s t 1 13.85 6 0.03137 42 T e s t 2 2.748 3 0.4321 43 T e s t 3 2.748 3 0.4321 44 T e s t 6a 3.719 1 0.05379 45 46 47 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 48 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 49 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 50 51 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 52 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e . 53 54 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 55 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 56 57 T h e p - v a l u e f o r T e s t 6a is l e s s t h a n .1. M o d e l 4 m a y n o t a d e q u a t e l y 58 d e s c r i b e t h e d a t a ; y o u m a y w a n t t o c o n s i d e r a n o t h e r m o d e l . 59 60 61 B e n c h m a r k D o s e C o m p u t a t i o n s : 62 63 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 64 65 R i s k T y p e = E s t i m a t e d s t a n d a r d d e v i a t i o n s f r o m c o n t r o l 66 67 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 68 69 BMD = 1.31616 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-87 DRAFT--DO NOT CITE OR QUOTE 1 BMDL 0.00236664 2 3 4 E .2 .1 4 .3 . Figurefor Selected Model: Exponential (M4) Exponential Model 4 with 0.95 Confidence Level Mean Response 5 10:49 02/08 2010 6 This document is a draftfor review purposes only and does not constitute Agency policy. E-88 DRAFT--DO NOT CITE OR QUOTE 1 E.2.15. H ojo et al., 2002: D R L R esponse Per M inute 2 E .2 .1 5 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) H ill 0 N /A 126.353 1.373E+00 1.070E-14 Notes linear polynom ial, 3degree power power, unrestricted exponential (M 2) 2 0.006 132.243 1.064E+01 5.340E +00 2 0.006 132.243 1.064E+01 5.340E +00 2 0.006 132.243 1.064E +01 5 .3 4 0 E + 0 0 p o w er b oun d hit (p ow er = 1) 2 0.741 122.455 1.070E+03 error unrestricted (power = 0) 2 0.570 122.980 5.027E-01 error exponential (M 3) exponential (M4) b exponential (M 5) 2 0.570 122.980 5.027E-01 error p o w er hit bound (d = 1) 1 0.477 124.360 3.813E-01 1.553E-02 0 N /A 126.353 8.430E-01 2.221E-02 a C onstant variance m odel selected (p = 0 .3 0 0 4 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2 .1 5 .2 . Outputfor Selected Model: Exponential (M4) 6 H ojo et al., 2002: D R L R esponse Per M inute 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 2 3 _ H o j o _ 2 0 0 2 _ D R L r e s p _ E x p C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 M o n F e b 08 1 0 : 5 0 : 1 0 2 0 1 0 14 15 16 T a b l e 5, v a l u e s a d j u s t e d b y a c o n s t a n t t o a l l o w e x p o n e n t i a l m o d e l 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 30 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 31 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. This document is a draftfor review purposes only and does not constitute Agency policy. E-89 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Dependent variable = Mean 4 Independent variable = Dose 5 Data are assumed to be distributed: normally 6 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 7 r h o is s e t t o 0. 8 A constant v a r i a n c e model is fit. 9 10 T o t a l n u m b e r o f d o s e g r o u p s = 4 11 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 12 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 13 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 16 M L E s o l u t i o n p r o v i d e d : E x a c t 17 18 19 I n i t i a l P a r a m e t e r V a l u e s 20 21 Variable Model 4 22 23 lnalpha 4.51689 24 rho(S) 0 25 a 2 4 . 6 3 6 2 26 b 0 . 3 7 9 3 2 7 27 c 0 . 0 1 8 4 7 8 5 28 d 1 29 30 (S) = S p e c i f i e d 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 Variable Model 4 37 38 lnalpha 4.54096 39 rho 0 40 a 2 3 . 4 6 7 4 41 b 1 . 6 1 1 8 5 42 c 0 . 1 0 1 3 1 7 43 d 1 44 45 46 T a b l e o f S t a t s F r o m I n p u t D a t a 47 48 Dose N Obs Mean Obs Std Dev 49 50 05 23.46 7.986 51 1.625 5 4.013 10.96 52 4.169 6 0.478 7.194 53 10.7 5 4.594 15.23 54 55 56 E s t i m a t e d V a l u e s o f I n t e r e s t 57 58 Dose Est Mean Est Std Scaled Residual 59 60 0 23.47 9.684 -0.001008 61 1 . 6 2 5 3.915 9.684 0.02265 62 4 . 1 6 9 2.403 9.684 -0.4869 63 10.7 2.378 9.684 0.5118 64 65 66 67 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d 68 69 M o d e l A1: Yij = M u ( i ) + e(i j ) 70 V a r { e ( i j ) } = S i g m a ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-90 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A2: Yij Mu(i) + e(ij) 3 Var{e(ij)} Sigma(i)^2 4 5 Model A3: Yij Mu(i) + e(ij) 6 Var{e(ij)} e x p ( l a l p h a + log(mean(i)) * rho) 7 8 M o d e l R: Yij M u + e(i) 9 Var{e(ij)} Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) DF AIC 15 16 A1 -57.92733 5 125.8547 17 A2 -56.09669 8 128.1934 18 A3 -57.92733 5 125.8547 19 R -64.49611 2 132.9922 20 4 -58.1801 4 124.3602 21 22 23 Additive constant for all log-likelihoods = -19.3. This constant added to the 24 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 25 d e p e n d o n t h e m o d e l p a r a m e t e r s . 26 27 28 E x p l a n a t i o n o f T e s t s 29 30 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 31 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 32 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 33 34 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 35 36 37 T e s t s o f I n t e r e s t 38 39 T e s t -2*log(Likelihood Ratio) D. F. p-value 40 41 T e s t 1 16.8 6 0.01005 42 T e s t 2 3.661 3 0.3004 43 T e s t 3 3.661 3 0.3004 44 T e s t 6a 0.5056 1 0.4771 45 46 47 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 48 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 49 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 50 51 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 52 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e . 53 54 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 55 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 56 57 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 58 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 59 60 61 B e n c h m a r k D o s e C o m p u t a t i o n s : 62 63 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 64 65 R i s k T y p e = E s t i m a t e d s t a n d a r d d e v i a t i o n s f r o m c o n t r o l 66 67 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 68 69 BMD = 0.381347 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-91 DRAFT--DO NOT CITE OR QUOTE 1 BMDL 0.0155267 2 3 4 E .2 .1 5 .3 . Figurefor Selected Model: Exponential (M4) Exponential Model 4 with 0.95 Confidence Level Mean Response 5 10:50 02/08 2010 6 This document is a draftfor review purposes only and does not constitute Agency policy. E-92 DRAFT--DO NOT CITE OR QUOTE 1 E .2.16. K attainen et al., 2001: 3rd M olar E ruption, Fem ale 2 E .2 .1 6 .1 . Summary Table o f BMDS Modeling Results Model logistic Degrees of Freedom x2p Value AIC 3 0.360 88.508 BMD (ng/kg) BMDL (ng/kg) Notes negative intercept (intercept = 9.223E+00 6.671E+00 -1.586) log-logistic a 3 0.982 85.227 2.399E+00 1.328E+00 slope bound hit (slope = 1) log-probit 3 0 .5 2 2 8 7 .4 2 4 7 .3 4 6 E + 0 0 4 .5 6 1 E + 0 0 slop e bound hit (slop e = 1) probit multistage, 4degree log-logistic, unrestricted b log-probit, unrestricted negative intercept (intercept = 3 0.379 88.352 8.802E+00 6.549E+00 -0.975) 3 0.781 86.155 4.042E+00 2.626E+00 final h = 0 2 0.949 87.162 1.931E+00 1.840E-01 unrestricted (slope = 0.91) 2 0.941 87.181 2.075E+00 2.395E-01 unrestricted (slope = 0.549) aBest-fitting m odel, BM D S output presented in this appendix b Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .1 6 .2 . Outputfor Selected Model: Log-Logistic 6 Kattainen et al., 2001: 3rd M olar Eruption, Fem ale 7 8 9 10 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 2 4 _ K a t t _ 2 0 0 1 _ E r u p _ L o g L o g i s t i c _ B M R 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 2 4 _ K a t t _ 2 0 0 1 _ E r u p _ L o g L o g i s t i c _ B M R 1 . p l t 13 M o n F e b 08 1 0 : 5 0 : 3 9 2 0 1 0 14 15 16 F i g u r e 2 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 S l o p e p a r a m e t e r is r e s t r i c t e d as s l o p e > = 1 27 28 T o t a l n u m b e r o f o b s e r v a t i o n s = 5 29 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 30 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 31 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 32 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 33 34 35 36 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 37 This document is a draftfor review purposes only and does not constitute Agency policy. E-93 DRAFT--DO NOT CITE OR QUOTE 1 2 Default Initial Parameter Values 3 background = 0.0625 4 intercept = -3.07535 5 slope = 1 6 7 8 Asymptotic Correlation Matrix of Parameter Estimates 9 10 ( * * * T h e m o d e l p a r a m e t e r ( s ) - s l o p e 11 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 12 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 13 14 background intercept 15 16 b a c k g r o u n d 1 -0.53 17 18 i n t e r c e p t -0.53 1 19 20 21 22 P a r a m e t e r E s t i m a t e s 23 24 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 25 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 26 b a c k g r o u n d 0 . 0 6 9 9 3 3 9 27 intercept -3.07219 28 s l o p e 1 29 30 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 31 32 33 34 A n a l y s i s o f D e v i a n c e T a b l e 35 36 Model Log(likelihood) # Param's Deviance Test d.f. P-value 37 Full model -40.5286 5 38 F i t t e d m o d e l -40.6137 2 0.170195 3 0.9823 39 R e d u c e d m o d e l -50.7341 1 20.411 4 0.0004142 40 41 AIC: 85.2274 42 43 44 G o o d n e s s o f F i t 45 S c a l e d 46 Dose Est . Prob Expected Observed Size Residual 47 48 0 .0000 0. 0 6 9 9 1. 1 1 9 1. 0 0 0 16 -0 117 49 2 .2297 0. 1 5 7 0 2. 6 6 9 3. 0 0 0 17 0. 2 2 1 50 6 .2523 0. 2 7 8 8 4. 1 8 2 4. 0 0 0 15 -0 105 51 16 .0824 0. 4 6 7 0 5. 604 6. 0 0 0 12 0. 2 2 9 52 46 .8576 0. 7 0 6 6 13 426 13. 000 19 -0 215 53 54 C h i ^ 2 = 0 .17 d.f. = 3 P- v a l u e = 0.9820 55 56 57 Be n c h m a r k D o s e C o m p u t a t i o n 58 59 S p e c i f i e d e f f e c t = 0. 1 60 61 R i s k T y p e = Extra risk 62 63 C o n f i d e n c e l e v e l 0.95 64 65 BMD = 2.39879 66 67 BMDL = 1.32815 68 69 This document is a draftfor review purposes only and does not constitute Agency policy. E-94 DRAFT--DO NOT CITE OR QUOTE 1 E.2.16.3. F igure f o r Selected M odel: Log-Logistic Log-Logistic Model with 0.95 Confidence Level 2 3 4 5 E .2.16.4. Outputfor Additional Model Presented: Log-Logistic, Unrestricted 6 Kattainen et al., 2001: 3rd M olar Eruption, Fem ale 7 8 9 10 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 2 4 _ K a t t _ 2 0 0 1 _ E r u p _ L o g L o g i s t i c _ U _ B M R 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 2 4 _ K a t t _ 2 0 0 1 _ E r u p _ L o g L o g i s t i c _ U _ B M R 1 . p l t 13 M o n F e b 08 1 0 : 5 0 : 4 0 2 0 1 0 14 15 16 F i g u r e 2 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 S l o p e p a r a m e t e r is n o t r e s t r i c t e d 27 28 T o t a l n u m b e r o f o b s e r v a t i o n s = 5 This document is a draftfor review purposes only and does not constitute Agency policy. E-95 DRAFT--DO NOT CITE OR QUOTE 1 Total number of records with missing values = 0 2 Maximum number of iterations = 250 3 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 4 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 5 6 7 8 User has chosen the log transformed model 9 10 11 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 12 background = 0.0625 13 intercept = -2.7659 14 slope = 0.901885 15 16 17 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 18 19 background intercept slope 20 21 b a c k g r o u n d 1 -0.52 0.38 22 23 i n t e r c e p t -0.52 1 -0.94 24 25 s l o p e 0.38 -0.94 1 26 27 28 29 P a r a m e t e r E s t i m a t e s 30 31 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 32 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf 33 b a c k g r o u n d 0 . 0 6 3 0 0 4 5 * * * 34 intercept -2.79616 * * * 35 slope 0.910333 36 37 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 38 39 40 41 A n a l y s i s o f D e v i a n c e T a b l e 42 43 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 44 Full model -40.5286 5 45 F i t t e d m o d e l -40.5811 3 0.105049 2 0.9488 46 R e d u c e d m o d e l -50.7341 1 20.411 4 0.0004142 47 48 AIC: 87.1622 49 50 51 G o o d n e s s o f F i t 52 53 Dose Est. Prob. Expected Observed Size Residual 54 55 0.0000 0.0630 1.008 1. 0 0 0 16 -0. 008 56 2.2297 0.1683 2.862 3. 0 0 0 17 0. 0 9 0 57 6.2523 0.2922 4.383 4. 0 0 0 15 -0. 217 58 16.0824 0.4692 5.631 6. 0 0 0 12 0. 2 14 59 46.8576 0.6903 13.116 13. 000 19 -0. 058 60 61 C h i ^ 2 = 0 . 1 0 d.f. = 2 P-value = 0.9491 62 63 64 B e n c h m a r k D o s e C o m p u t a t i o n 65 66 S p e c i f i e d e f f e c t = 0.1 67 68 R i s k T y p e = Extra risk 69 70 C o n f i d e n c e l e v e l = 0.95 This document is a draftfor review purposes only and does not constitute Agency policy. E-96 DRAFT--DO NOT CITE OR QUOTE 1 2 BMD 1.93079 3 4 BMDL = 0.18403 5 6 7 E .2.16.5. Figurefor Additional Model Presented: Log-Logistic, Unrestricted Log-Logistic Model with 0.95 Confidence Level Fraction Affected 8 10:50 02/08 2010 9 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-97 DRAFT--DO NOT CITE OR QUOTE 1 E .2.17. K attainen et al., 2001: 3rd M olar L ength, F em ale 2 E .2 .1 7 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes exponential (M 2) 3 <0.0001 1.669E+01 9.933E+00 124.866 exponential (M 3) 3 <0.0001 1.669E +01 9 .9 3 3 E + 0 0 p o w er hit b ound (d = 1) 124.866 exponential (M 4) 2 0.002 4.237E-01 2.530E-01 147.120 exponential (M 5) Hill b 2 0.002 4.2 3 7 E -0 1 2.5 3 0 E -0 1 p o w er hit b ound (d = 1) 147.120 2 0.022 152.239 3.132E-01 1.679E-01 n lower bound hit (n = 1) linear polynom ial, 4degree 3 <0.0001 1.982E+01 1.277E+01 124.024 3 <0.0001 1.982E+01 1.277E+01 124.024 power 3 <0.0001 1.982E +01 1.277E +01 p o w er b oun d hit (p ow er = 1) 124.024 H ill, unrestricted c 1 <0.0001 1.215E-02 error 130.856 unrestricted (n = 13.042) power, unrestricted 2 0.263 1.964E-03 8.002E-06 unrestricted (power = 0.195) 157.201 a N on-con stant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .1 7 .2 . Outputfor Selected Model: Hill 6 Kattainen et al., 2001: 3rd M olar Length, Fem ale 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 2 5 _ K a t t _ 2 0 0 1 _ L e n g t h _ H i l l _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 2 5 _ K a t t _ 2 0 0 1 _ L e n g t h _ H i l l _ 1 . p l t 13 M o n F e b 08 1 0 : 5 1 : 0 9 2 0 1 0 14 15 16 F i g u r e 3 f e m a l e o n l y 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 P o w e r p a r a m e t e r r e s t r i c t e d t o b e g r e a t e r t h a n 1 This document is a draftfor review purposes only and does not constitute Agency policy. E-98 DRAFT--DO NOT CITE OR QUOTE 1 The v ariance is to be mode l e d as Var(i) = exp(lalpha + rho * ln(mean(i))) 2 3 Total number of dose groups = 5 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 12 lalpha -2.37155 13 rho 0 14 intercept 1.85591 15 - 0 . 5 0 7 8 7 4 16 0 . 8 4 5 9 3 2 17 2 . 0 3 1 2 9 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( * * * T h e m o d e l p a r a m e t e r ( s ) -n 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 24 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 25 26 lalpha rho intercept v k 27 28 l a l p h a 1 -0.98 -0.16 0.84 -0.38 29 30 rho -0.98 1 0.2 -0.79 0.4 31 32 i n t e r c e p t -0.16 0.2 1 -0.3 -0.11 33 34 v 0.84 -0.79 -0.3 1 -0.52 35 36 k CO CO o 0.4 -0.11 -0.52 1 37 38 39 40 P a r a m e t e r E s t i m a t e s 41 42 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 43 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 44 lalpha 3.31084 1.404 0.559057 6.06262 45 rho -14.2657 2.62739 -19.4153 -9.11612 46 ntercept 1.85483 0.0159477 1.82357 1.88609 47 v -0.453667 0.0620227 -0.575229 -0.332105 48 n 1 N A 49 k 1.91219 0.624785 0.687636 3.13675 50 51 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 52 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 53 h a s n o s t a n d a r d e r r o r . 54 55 56 57 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 58 59 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 60 61 62 0 16 1.86 1.85 0.0661 0.0639 0.0674 63 2.23 17 1.58 1.61 0.185 0.175 -0.789 64 6 . 2 5 2 15 1.6 1.51 0.265 0.28 1.22 65 1 6 . 0 8 12 1.5 1.45 0.221 0.371 0.51 66 4 6 . 8 6 19 1.35 1.42 0.515 0.431 -0.716 67 68 69 70 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d This document is a draftfor review purposes only and does not constitute Agency policy. E-99 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Model A1: Yij Mu(i) + e(ij) 4 Var{e(ij)} Sigma^2 5 6 Model A2: Yij Mu(i) + e(ij) 7 Var{e(ij)} Sigma(i)^2 8 9 Model A3: Yij = Mu(i) + e(ij) 10 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 11 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 12 w e r e s p e c i f i e d b y t h e u s e r 13 14 M o d e l R: Yi = M u + e(i) 15 V a r { e ( i ) } = S i g m a ^ 2 16 17 18 L i k e l i h o o d s o f I n t e r e s t 19 20 Model Log(likelihood) # Param's AIC 21 A1 56.758717 6 -101.517434 22 A 2 8 5 . 8 5 6 4 5 0 10 - 1 5 1 . 7 1 2 9 0 1 23 A3 84.934314 7 -155.868628 24 fitted 81.119648 5 -152.239295 25 R 45.373551 2 -86.747101 26 27 28 E x p l a n a t i o n o f T e s t s 29 30 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 31 (A2 vs. R) 32 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 33 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 34 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 35 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 36 37 T e s t s o f I n t e r e s t 38 39 Test -2*log(Likelihood Ratio) Test df p-value 40 41 T e s t 1 80.9658 8 <.0001 42 T e s t 2 58.1955 4 <.0001 43 T e s t 3 1.84427 3 0.6053 44 T e s t 4 7.62933 2 0.02205 45 46 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 47 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 48 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 49 50 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 51 m o d e l a p p e a r s t o b e a p p r o p r i a t e 52 53 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 54 t o b e a p p r o p r i a t e h e r e 55 56 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 57 m o d e l 58 59 60 B e n c h m a r k D o s e C o m p u t a t i o n 61 62 S p e c i f i e d e f f e c t = 1 63 64 R i s k T y p e = Estimated standard deviations from the control mean 65 66 C o n f i d e n c e l e v e l = 0.95 67 68 BMD = 0.313211 69 70 BMDL = 0.167922 This document is a draftfor review purposes only and does not constitute Agency policy. E-100 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E.2.17.3. F igure f o r Selected M odel: H ill Hill Model with 0.95 Confidence Level 2 3 4 5 E .2.17.4. Outputfor Additional Model Presented: Hill, Unrestricted 6 Kattainen et al., 2001: 3rd M olar Length, Fem ale 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 2 5 _ K a t t _ 2 0 0 1 _ L e n g t h _ H i l l _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 2 5 _ K a t t _ 2 0 0 1 _ L e n g t h _ H i l l _ U _ 1 . p l t 13 M o n F e b 08 1 0 : 5 1 : 0 9 2 0 1 0 14 15 16 F i g u r e 3 f e m a l e o n l y 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 P o w e r p a r a m e t e r is n o t r e s t r i c t e d 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + r h o * l n ( m e a n ( i ) ) ) 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-101 DRAFT--DO NOT CITE OR QUOTE 1 Total number of dose groups = 5 2 Total number of records with missing values = 0 3 Maximum number of iterations = 250 4 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 5 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 6 7 8 9 Default Initial Parameter Values 10 lalpha -2.37155 11 rho 0 12 intercept 1.85591 13 - 0 . 5 0 7 8 7 4 14 0 . 8 4 5 9 3 2 15 2 . 0 3 1 2 9 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 lalpha rho intercept v n k 21 22 l a l p h a 1 -0.98 -0.16 0.84 1.4e-016 3.3e-017 23 24 rho -0.98 1 0.22 -0.77 -2.2e-016 -5.1e-017 25 Lf) CO o Lf) CO o 26 i n t e r c e p t -0.16 0.22 1 6e-017 1.4e-017 27 28 v 0.84 -0.77 1 -2.6e-016 -6.2e-017 29 30 n 1.4e-016 -2.2e-016 6e-017 -2.6e-016 11 31 32 k 3.3e-017 -5.1e-017 1.4e-017 -6.2e-017 11 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 39 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 40 lalpha 4.25154 1.5913 1.13265 7.37044 41 rho -15.7639 2.90127 -21.4503 -10.0776 42 i n t e r c e p t 1.85591 0.0160104 1.82453 1.88729 43 v -0.357293 0.0463784 -0.448193 -0.266393 44 n 13.0417 4.64308e+013 -9.10027e+013 9.10027e+013 45 k 0.0136512 2.57737e+011 -5.05155e+011 5.05155e+011 46 47 48 49 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 50 51 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 52 53 54 0 16 1.86 1.86 0.0661 0.064 2.09e-009 55 2.23 17 1.58 1.5 0.185 0.345 0.937 56 6 . 2 5 2 15 1.6 1.5 0.265 0.345 1.09 57 1 6 . 0 8 12 1.5 1.5 0.221 0.345 0.0534 58 4 6 . 8 6 19 1.35 1.5 0.515 0.345 -1.9 59 60 61 62 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 63 64 65 M o d e l A1: Yij = Mu(i) + e(ij) 66 V a r { e i ;i j ) } = S i g m a ^ 2 67 68 M o d e l A2: Yij = Mu(i) + e(ij) 69 V a r { e i :i j ) } = S i g m a ( i ) ^ 2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-102 DRAFT--DO NOT CITE OR QUOTE 1 Model A3: Yij = Mu(i) + e(ij) 2 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 3 Model A3 uses any fixed variance parameters that 4 were specified by the user 5 6 M o d e l R: Yi = M u + e(i) 7 Var{e(i)} = Sigma^2 8 9 10 L i k e l i h o o d s o f I n t e r e s t 11 12 Model Log(likelihood) # Param's AIC 13 A1 56.758717 6 -101.517434 14 A 2 8 5 . 8 5 6 4 5 0 10 - 1 5 1 . 7 1 2 9 0 1 15 A3 84.934314 7 -155.868628 16 fitted 71.427978 6 -130.855955 17 R 45.373551 2 -86.747101 18 19 20 E x p l a n a t i o n o f T e s t s 21 22 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 23 (A2 vs. R) 24 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 25 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 26 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 27 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 28 29 T e s t s o f I n t e r e s t 30 31 Test -2*log(Likelihood Ratio) Test df p-value 32 33 T e s t 1 80.9658 8 <.0001 34 T e s t 2 58.1955 4 <.0001 35 T e s t 3 1.84427 3 0.6053 36 T e s t 4 27.0127 1 <.0001 37 38 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 39 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 40 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 41 42 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 43 m o d e l a p p e a r s t o b e a p p r o p r i a t e 44 45 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 46 t o b e a p p r o p r i a t e h e r e 47 48 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 49 m o d e l 50 51 52 B e n c h m a r k D o s e C o m p u t a t i o n 53 54 S p e c i f i e d e f f e c t = 1 55 56 R i s k T y p e = Estimated standard deviations from the control mean 57 58 C o n f i d e n c e l e v e l = 0.95 59 60 BMD = 0.012148 61 62 63 B M D L c o m p u t a t i o n f a i l e d . 64 65 This document is a draftfor review purposes only and does not constitute Agency policy. E-103 DRAFT--DO NOT CITE OR QUOTE 1 E.2.17.5. F igure f o r A d d itio n a l M o d el Presented: H ill, U nrestricted Hill Model t-- 1-- 1-- 1-- 1-- 1-- 1-- 1-- 1-- 1-- i-- 1-- 1-- 1-- 1-- 1-- 1-- 1-- '-- '-- i-- '-- '-- '-- '-- '-- '-- '-- '-- '-- i-- '-- '-- '-- '-- '-- '-- '-- '-- '-- r Hill -------------1.9 1.8 1.7 Mean Response 1.6 1.5 1.4 1.3 1.2 1.1 BMD...................................................................................................................... 0 2 10:51 02/08 2010 3 10 20 30 40 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-104 DRAFT--DO NOT CITE OR QUOTE 1 E.2.18. Keller et al., 2007: Missing Mandibular Molars, CBA J 2 E.2.18.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2pValue AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 1 0.105 52.510 3.342E+00 8.986E-01 logistic negative intercept (intercept = 2 0.335 49.984 3.069E+00 2.212E+00 -3.414) log-logistic 1 0.105 52.524 4.009E+00 2.411E+00 log-probit multistage, 1degree a multistage, 2degree multistage, 3degree probit W eibull 1 0.105 52.524 3.845E+00 2.421E+00 3 0.255 50.425 1.091E+00 7.624E-01 1 0.122 51.391 1.916E+00 9.654E-01 1 0.150 50.853 1.713E+00 9.584E-01 negative intercept (intercept = 2 0.342 49.904 2.927E+00 2.053E+00 -1.873) 1 0.108 52.219 2.744E+00 9.350E-01 aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2.18.2. Outputfor Selected Model: Multistage, 1-Degree 6 K eller et al., 2007: M issing M andibular M olars, C B A J 7 8 9 10 M u l t i s t a g e M o d e l . ( V e r s i o n : 3.0; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 2 6 _ K e l l e r _ 2 0 0 7 _ M o l a r s _ M u l t i 1 _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 2 6 _ K e l l e r _ 2 0 0 7 _ M o l a r s _ M u l t i 1 _ 1 . p l t 13 M o n F e b 08 1 0 : 5 1 : 4 7 2 0 1 0 14 15 16 T a b l e 1 u s i n g m a n d i b u l a r m o l a r s o n l y 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) * [ 1 - E X P ( 22 - b e t a i * d o s e ^ i ) ] 23 24 T h e p a r a m e t e r b e t a s a r e r e s t r i c t e d t o b e p o s i t i v e 25 26 27 D e p e n d e n t v a r i a b l e = D i c h E f f 28 I n d e p e n d e n t v a r i a b l e = D o s e 29 30 T o t a l n u m b e r o f o b s e r v a t i o n s = 4 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 T o t a l n u m b e r o f p a r a m e t e r s i n m o d e l = 2 33 T o t a l n u m b e r o f s p e c i f i e d p a r a m e t e r s = 0 This document is a draftfor review purposes only and does not constitute Agency policy. E-105 DRAFT--DO NOT CITE OR QUOTE 1 Degree of polynomial = 1 2 3 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 Background = 0 12 B e t a ( 1 ) = 3 . 0 3 9 8 8 e + 0 1 8 13 14 15 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 16 17 ( * * * T h e m o d e l p a r a m e t e r ( s ) - B a c k g r o u n d 18 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 19 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 20 21 B e t a ( 1 ) 22 23 B e t a ( 1 ) 1 24 25 26 27 P a r a m e t e r E s t i m a t e s 28 29 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 30 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 31 B a c k g r o u n d 0 32 Beta(1) 0.096571 33 34 I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 35 36 37 38 A n a l y s i s o f D e v i a n c e T a b l e 39 40 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 41 Full model -21.5798 4 42 F i t t e d m o d e l -24.2126 1 5.26564 3 0.1533 43 R e d u c e d m o d e l -71.326 1 99.4926 3 <.0001 44 45 AIC: 50.4251 46 47 48 G o o d n e s s o f F i t 49 S c a l e d 50 Dose Est . Prob. Expected Observed Size Residu; 51 52 0.0000 0. 0 0 0 0 0 .000 0. 0 0 0 29 0. 0 0 0 53 0.5374 0. 0 5 0 6 1 .163 2. 0 0 0 23 0..796 54 4.2881 0. 3 3 9 1 9 .833 6. 0 0 0 29 -1. 504 55 34.0560 0. 9627 28 .881 30. 000 30 1..0 78 56 57 ii^2 = 4 .06 d . f . = 3 P- v a l u e = 0.2554 58 59 60 B e n c h m a rk D o s e C o m p u t a t i o n 61 62 r e i f i e d e f f e c t = 0 .1 63 64 k T y p e = Extra r isk 65 66 f i d e n c e l e v e l = 0. 95 67 68 BMD = 1.09102 69 70 BMDL = 0.762404 This document is a draftfor review purposes only and does not constitute Agency policy. E-106 DRAFT--DO NOT CITE OR QUOTE 1 2 BMDU 1.56496 3 4 T a k e n t ogether, (0.762404, 1.56496) is a 90 5 interval for the BMD 6 7 % two-sided confidence 8 E .2.18.3. Figurefor Selected Model: Multistage, 1-Degree Multistage Model with 0.95 Confidence Level Fraction Affected 9 10:51 02/08 2010 10 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-107 DRAFT--DO NOT CITE OR QUOTE 1 E.2.19. Kociba et al., 1978: Urinary Coproporphyrin, Females 2 E.2.19.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 2 <0.0001 82.975 2.378E+01 1.340E+01 Notes exponential (M 3) exponential (M4) b exponential (M 5) 2 < 0 .0 0 0 1 82 .9 7 5 2 .3 7 8 E + 0 1 1.340E +01 p ow er hit b ound (d = 1) 1 0.006 73.823 1.566E+00 7.180E-01 0 N /A 69.047 6.225E+00 1.586E+00 H ill 0 N /A 69.047 5.473E +00 error linear 2 <0.001 82.233 1.790E+01 3.862E+00 polynom ial, 3degree 2 <0.001 82.233 1.790E+01 3.862E+00 power 2 < 0 .0 0 1 82.233 1.790E +01 3 .8 6 2 E + 0 0 p ow er b oun d hit (p ow er = 1) power, unrestricted 1 <0.001 78.691 1.148E+00 8.984E-09 unrestricted (power = 0.416) a N on-constant variance m odel selected (p = 0 .0298) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2 .1 9 .2 . Outputfor Selected Model: Exponential (M4) 6 K ociba et al., 1978: Urinary Coproporphyrin, Fem ales 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 2 9 _ K o c i b a _ 1 9 7 8 _ C o p r o _ E x p _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 M o n F e b 08 1 0 : 5 2 : 4 7 2 0 1 0 14 15 16 T a b l e 2 - U r i n a r y C o p r o p o r p h y r i n 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 30 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 31 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. This document is a draftfor review purposes only and does not constitute Agency policy. E-108 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Dependent variable = Mean 4 Independent variable = Dose 5 Data are assumed to be distributed: normally 6 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 7 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) * rho) 8 9 Total number of dose groups = 4 10 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 11 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 12 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 13 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 15 M L E s o l u t i o n p r o v i d e d : E x a c t 16 17 18 I n i t i a l P a r a m e t e r V a l u e s 19 20 Variable Model 4 21 22 lnalpha -5.58269 23 rho 2.98472 24 a 8 .17 25 b 0 . 0 6 9 2 4 7 8 26 c 2 . 2 3 6 2 3 27 d 1 28 29 30 31 P a r a m e t e r E s t i m a t e s 32 33 Variable Model 4 34 35 lnalpha -4.90852 36 rho 2.80743 37 a 8 . 9 1 0 7 1 38 b 0 . 1 5 3 0 4 39 c 1 . 9 7 5 2 6 40 d 1 41 42 43 T a b l e o f S t a t s F r o m I n p u t D a t a 44 45 Dose N Obs Mean Obs Std Dev 46 CO CO Gj CO 47 0 5 1.3 48 1.547 5 2 49 7.155 5 16.4 4.7 50 38.56 5 17.4 4 51 52 53 E s t i m a t e d V a l u e s o f I n t e r e s t 54 55 Dose Est Mean Est Std Scaled Residual 56 57 0 8.911 1.852 1.074 58 1 . 5 4 7 10.74 2.407 -1.991 59 7 . 1 5 5 14.69 3.736 1.021 60 3 8 . 5 6 17.58 4.805 -0.08246 61 62 63 64 O t h e r m o d e l s; f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d 65 66 M o d e l A1: Yij = M u ( i ) + e( ij) 67 V a r { e ( i j ) } = S i g m a ^ 2 68 69 M o d e l A2: Yij = M u ( i ) + e( ij) 70 V a r { e ( i j ) } = S i g m a ( i ) ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-109 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A3: Yij = Mu(i) + e(ij) 3 V ar{e(ij)} = exp(l a l p h a + log(mean(i)) * rho) 4 5 M o d e l R: Yij = M u + e(i) 6 Var{e(ij)} = Sigma^2 7 8 9 Likelihoods of Interest 10 11 Model Log(likelihood) DF AIC 12 13 A1 -31.69739 5 73.39478 14 A2 -27.21541 8 70.43081 15 A3 -28.16434 6 68.32868 16 R -41.73188 2 87.46376 17 4 -31.91136 5 73.82272 18 19 20 Additive constant for all log-likelihoods = -18.38. This constant added to the 21 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 22 d e p e n d o n t h e m o d e l p a r a m e t e r s . 23 24 25 E x p l a n a t i o n o f T e s t s 26 27 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 28 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 29 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 30 31 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 32 33 34 T e s t s o f I n t e r e s t 35 36 T e s t -2*log(Likelihood Ratio) D. F. p-value 37 38 T e s t 1 29.03 6 < 0.0001 39 T e s t 2 8.964 3 0.02977 40 T e s t 3 1.898 2 0.3872 41 T e s t 6a 7.494 1 0.00619 42 43 44 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 45 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 46 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 47 48 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 49 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e . 50 51 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 52 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 53 54 T h e p - v a l u e f o r T e s t 6a is l e s s t h a n .1. M o d e l 4 m a y n o t a d e q u a t e l y 55 d e s c r i b e t h e d a t a ; y o u m a y w a n t t o c o n s i d e r a n o t h e r m o d e l . 56 57 58 B e n c h m a r k D o s e C o m p u t a t i o n s : 59 60 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 61 62 R i s k T y p e = E s t i m a t e d s t a n d a r d d e v i a t i o n s f r o m c o n t r o l 63 64 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 65 66 BMD = 1.56562 67 68 BMDL = 0.718033 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-110 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E .2 .1 9 .3 . Figurefor Selected Model: Exponential (M4) Exponential Model 4 with 0.95 Confidence Level 2 10:52 02/08 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-111 DRAFT--DO NOT CITE OR QUOTE 1 E.2.20. Kociba et al., 1978: Uroporphyrin per Creatinine, Female 2 E.2.20.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2pValue AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 2 0.755 -93.828 1.641E+01 1.259E+01 Notes exponential (M 3) 2 0.7 5 5 -9 3 .8 2 8 1.641E +01 1.259E +01 p o w er hit b ound (d = 1) exponential (M 4) 1 0.499 -91.935 1.216E+01 3.958E +00 exponential (M 5) 0 N /A -90.190 7.542E+00 4.128E+00 Hill 0 N /A -90.190 7.607E+00 3.966E+00 linear b polynom ial, 3degree power 2 0.793 -93.928 1.306E+01 9.287E+00 2 0.793 -93.928 1.306E+01 9.287E +00 1 0.497 -91.928 1.326E+01 9.287E +00 a C onstant variance m odel selected (p = 0 .4 9 1 9 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2 .2 0 .2 . Outputfor Selected Model: Linear 6 K ociba et al., 1978: Uroporphyrin per Creatinine, Fem ale 7 8 9 10 P o l y n o m i a l M o d e l . ( V e r s i o n : 2 . 1 3 ; D a t e : 0 4 / 0 8 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 2 8 _ K o c i b a _ 1 9 7 8 _ U r o p o r _ L i n e a r C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 2 8 _ K o c i b a _ 1 9 7 8 _ U r o p o r _ L i n e a r C V _ 1 . p l t 13 M o n F e b 08 1 0 : 5 2 : 1 7 2 0 1 0 14 15 16 T a b l e 2 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 S i g n s o f t h e p o l y n o m i a l c o e f f i c i e n t s a r e n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t 29 30 T o t a l n u m b e r o f d o s e g r o u p s = 4 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 33 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 34 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 This document is a draftfor review purposes only and does not constitute Agency policy. E-112 DRAFT--DO NOT CITE OR QUOTE 1 2 3 4 Default Initial Parameter Values 5 alpha = 0.0030385 6 rho = 0 Specified 7 beta_0 = 0.149139 8 beta 1 = 0.00381789 9 10 11 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 12 13 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 14 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 15 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 16 17 alpha beta_0 beta_1 18 19 a l p h a 1 1.9e-009 -2.6e-009 20 21 beta_0 1.9e-009 1 -0.6 22 23 beta 1 2.6e-009 -0.6 1 24 25 26 27 P a r a m e t e r E s t i m a t e s 28 29 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 30 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 31 alpha 0.00248773 0.000786688 0.000945846 0.00402961 32 beta_0 0.149139 0.0139684 0.121761 0.176517 33 beta 1 0.00381789 0.000711776 0.00242284 0.00521295 34 35 36 37 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 38 39 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 40 41 42 05 0.157 0.149 0.05 0.0499 0.352 Lf) O 43 1 . 5 4 7 5 0.143 0.155 0.037 0.0499 44 7 . 1 5 5 5 0.181 0.176 0.053 0.0499 0.204 45 3 8 . 5 6 5 0.296 0.296 0.074 0.0499 -0.0161 46 47 48 49 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 50 51 52 M o d e l A1: Yij = Mu(i) + e(ij) 53 V a r { e i ;i j ) } = S i g m a ^ 2 54 55 M o d e l A2: Yij = Mu(i) + e(ij) 56 V a r { e i :i j ) } = S i g m a ( i ) ^2 57 58 M o d e l A3: Yij = Mu(i) + e(ij) 59 V a r { e i i j ) } = S i g m a ^ 2 60 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 61 w e r e s p e c i f i e d b y t h e u s e r 62 63 M o d e l R: Yi = M u + e(i) 64 V a r { e ( i ) } = S i g m a ^ 2 65 66 67 L i k e l i h o o d s o f I n t e r e s t 68 69 Model Log(likelihood) # P a r a m 's AIC 70 A1 50.195349 5 -90.390697 This document is a draftfor review purposes only and does not constitute Agency policy. E-113 DRAFT--DO NOT CITE OR QUOTE 1 A2 51. 4 0 0 0 5 1 8 -86. 800103 2 A 3 50.. 195349 5 -90. 390697 3 fitted 49..9 6 3 8 6 3 3 -93. 927727 4 R 41..0 4 9 7 5 5 2 -78. 099510 5 6 7 Explanation of Tests 8 9 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 10 (A2 vs. R) 11 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 12 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 13 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 14 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 15 16 T e s t s o f I n t e r e s t 17 18 Test -2*log(Likelihood Ratio) Test df p-vali 19 20 T e s t 1 20.7006 6 0.002076 21 T e s t 2 2.40941 3 0.4919 22 T e s t 3 2.40941 3 0.4919 23 T e s t 4 0.46297 2 0.7934 24 25 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 26 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 27 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 28 29 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 30 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 31 32 33 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 34 t o b e a p p r o p r i a t e h e r e 35 36 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 37 t o a d e q u a t e l y d e s c r i b e t h e d a t a 38 39 40 B e n c h m a r k D o s e C o m p u t a t i o n 41 42 S p e c i f i e d e f f e c t = 1 43 44 R i s k T y p e = Estimated standard deviations from the control mean 45 46 C o n f i d e n c e l e v e l = 0.95 47 48 BMD = 13.064 49 50 51 BMDL = 9.28715 52 This document is a draftfor review purposes only and does not constitute Agency policy. E-114 DRAFT--DO NOT CITE OR QUOTE 1 E .2 .2 0 .3 . Figurefor Selected Model: Linear Linear Model with 0.95 Confidence Level Mean Response 2 10:52 02/08 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-115 DRAFT--DO NOT CITE OR QUOTE 1 E.2.21. Latchoumycandane and Mathur, 2002: Sperm Production 2 E.2.21.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 2 <0.0001 93.831 1.739E+01 9.432E +00 Notes exponential (M 3) 2 < 0 .0 0 0 1 93.831 1.739E +01 9 .4 3 2 E + 0 0 p o w er hit b ound (d = 1) exponential (M 4) 1 0.700 75.261 1.912E-01 7.976E -02 exponential (M 5) Hill b 0 N /A 77.263 2.925E-01 7.970E-02 1 0.962 75.115 1.171E-01 1.324E-02 n lower bound hit (n = 1) linear polynom ial, 3degree power 2 <0.0001 94.250 1.995E+01 1.212E+01 2 <0.0001 94.250 1.995E+01 1.212E+01 2 < 0 .0 0 0 1 9 4 .2 5 0 1.995E +01 1.212E +01 p o w er b oun d hit (p ow er = 1) H ill, unrestricted c 0 N /A 77.113 9.955E-02 1.228E-09 unrestricted (n = 0.916) power, unrestricted 1 0.501 75.566 6.921E -06 6.921E -06 unrestricted (pow er = 0.087) a C onstant variance m odel selected (p = 0 .8 5 0 6 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .2 1 .2 . Outputfor Selected Model: Hill 6 Latchoum ycandane and Mathur, 2002: Sperm Production 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 3 0 _ L a t c h _ 2 0 0 2 _ S p e r m _ H i l l C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 3 0 _ L a t c h _ 2 0 0 2 _ S p e r m _ H i l l C V _ 1 . p l t 13 M o n F e b 08 1 0 : 5 3 : 2 6 2 0 1 0 14 15 16 ( x 1 0 A 6) T a b l e 1 w i t h o u t V i t a m i n E 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 This document is a draftfor review purposes only and does not constitute Agency policy. E-116 DRAFT--DO NOT CITE OR QUOTE 1 Power parameter restricted to be greater than 1 2 A constant v ariance model is fit 3 4 Total number of dose groups = 4 5 Total number of records with missing values = 0 6 Maximum number of iterations = 250 7 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 8 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 9 10 11 12 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 13 alpha 7.23328 14 rho 0 Specified 15 intercept 22.19 16 - 9 . 0 9 17 1 . 9 3 0 5 9 18 0 . 5 4 6 8 6 4 19 20 21 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 22 23 ( *** The model parameter(s) -rho -n 24 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 25 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 26 27 alpha intercept v k 28 29 a l p h a 1 -2.2e-009 -3.7e-008 -5.9e-009 30 31 i n t e r c e p t 2.2e-009 1 -0.76 -0.23 32 33 v 3.7e-008 -0.76 1 -0.24 34 35 k 5.9e-009 -0.23 -0.24 1 36 37 38 39 P a r a m e t e r E s t i m a t e s 40 41 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 42 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Lim 43 alpha 6.0283 1.74022 2.61753 9.43907 44 i n t e r c e p t 22.1894 1.00236 20.2248 24.154 45 v -9.16715 1.30966 -11.734 -6.60026 46 n 1 N A 47 k 0.320198 0.220443 -0.111862 0.752259 48 49 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 50 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 51 h a s n o s t a n d a r d e r r o r . 52 53 54 55 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 56 57 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 58 59 60 06 22.2 22.2 2.67 2.46 0.000631 61 0 . 7 8 4 5 6 15.7 15.7 2.65 2.46 -0.00931 62 4 . 6 5 1 6 13.7 13.6 2.19 2.46 0.0372 63 2 7 . 2 7 6 13.1 13.1 3.16 2.46 -0.0285 64 65 66 67 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 68 69 70 M o d e l A1: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-117 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma^2 2 3 Model A2: Yij = Mu(i) + e(ij) 4 Var{e(ij)} = Sigma(i)^2 5 6 Model A3: Yij = Mu(i) + e(ij) 7 Var{e(ij)} = Sigma^2 8 Model A3 uses any fixed variance parameters that 9 were specified by the user 10 11 M o d e l R: Yi = M u + e(i) 12 V a r { e ( i ) } = S i g m a ^ 2 13 14 15 L i k e l i h o o d s o f I n t e r e s t 16 17 Model Log(likelihood) # Param's AIC 18 A1 -33.556444 5 77 .112888 19 A2 -33.158811 8 82 . 317623 20 A3 -33.556444 5 77 .112888 21 fitted -33.557588 4 75 .115176 22 R -47.392394 2 98 . 7 8 4 7 8 8 23 24 25 E x p l a n a t i o n o f T e s t s 26 27 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 28 (A2 vs. R) 29 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 30 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 31 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 32 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 33 34 T e s t s o f I n t e r e s t 35 36 Test -2*log(Likelihood Ratio) Test df p-val 37 38 T e s t 1 28.4672 6 <.0001 39 T e s t 2 0.795266 3 0.8506 40 T e s t 3 0.795266 3 0.8506 41 T e s t 4 0.00228746 1 0.9619 42 43 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 44 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 45 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 46 47 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 48 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 49 50 51 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 52 t o b e a p p r o p r i a t e h e r e 53 54 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 55 t o a d e q u a t e l y d e s c r i b e t h e d a t a 56 57 58 B e n c h m a r k D o s e C o m p u t a t i o n 59 60 S p e c i f i e d e f f e c t = 1 61 62 R i s k T y p e = Estimated standard deviations from the control mean 63 64 C o n f i d e n c e l e v e l = 0.95 65 66 BMD = 0.117131 67 68 BMDL = 0.0132353 69 This document is a draftfor review purposes only and does not constitute Agency policy. E-118 DRAFT--DO NOT CITE OR QUOTE 1 E .2 .2 1 .3 . Figurefor Selected Model: Hill Hill Model with 0.95 Confidence Level Mean Response 2 10:53 02/08 2010 3 dose 4 5 E .2.21.4. Outputfor Additional Model Presented: Hill, Unrestricted 6 Latchoum ycandane and Mathur, 2002: Sperm Production 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 3 0 _ L a t c h _ 2 0 0 2 _ S p e r m _ H i l l C V _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 3 0 _ L a t c h _ 2 0 0 2 _ S p e r m _ H i l l C V _ U _ 1 . p l t 13 M o n F e b 08 1 0 : 5 3 : 2 6 2 0 1 0 14 15 16 ( x 1 0 A 6) T a b l e 1 w i t h o u t V i t a m i n E 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 P o w e r p a r a m e t e r is n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t This document is a draftfor review purposes only and does not constitute Agency policy. E-119 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 4 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha = 7.23328 12 rho = 0 Specified 13 intercept = 22.19 14 - 9 . 0 9 15 1 . 9 3 0 5 9 16 0 . 5 4 6 8 6 4 17 18 19 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 20 21 ( J T h e m o d e l p a r a m e t e r ( s ) - r h o 22 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r 23 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 24 25 alpha intercept v n k 26 27 a l p h a 1 -9.8e-009 1.6e-007 1.6e-007 1.2e-007 28 29 intercept -9.8e-009 1 -0.5 -0.015 -0.13 30 31 v 1.6e-007 -0.5 1 0.76 0.56 32 33 n 1.6e-007 -0.015 0.76 1 0.86 34 35 k 1.2e-007 -0.13 0.56 0.86 1 36 37 38 39 P a r a m e t e r E s t i m a t e s 40 41 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 42 Variable Estimate Std. Err. Lower Conf . Limit Upper Conf. Limit 43 alpha 6.02773 1.74006 2. 6 1 7 2 8 9.43818 44 i n t e r c e p t 22.19 1.00231 20 .2255 24.1545 45 v -9.23667 2.03204 -13 .2194 -5.25394 46 n 0.916265 1.66287 -2. 34291 4.17544 47 k 0.301742 0.440535 -0.561692 1.16518 48 49 50 51 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 52 53 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res. 54 55 56 06 22.2 22.2 2.67 2.46 3.4e-008 57 . 7 8 4 5 6 15.7 15.7 2.65 2.46 -1.51e-007 58 1.651 6 13.7 13.6 2.19 2.46 2.62e-007 59 7 . 2 7 6 13.1 13.1 3.16 2.46 -5.45e-007 60 61 D e g r e e s o f f r e e d o m f o r T e s t A 3 v s f i t t e d 0 62 63 64 65 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 66 67 68 M o d e l A1: Yij = Mu(i) + e(ij) 69 V a r { e ( i j ) } = S i g m a ^ 2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-120 DRAFT--DO NOT CITE OR QUOTE 1 Model A2: Yij = Mu(i) + e(ij) 2 Var{e(ij)} = Sigma(i)^2 3 4 Model A3: Yij = Mu(i) + e(ij) 5 Var{e(ij)} = Sigma^2 6 Model A3 uses any fixed variance parameters that 7 were specified by the user 8 9 M o d e l R: Yi = M u + e(i) 10 V a r { e ( i ) } = S i g m a ^ 2 11 12 13 L i k e l i h o o d s o f I n t e r e s t 14 15 Model Log(likelihood) # Param's AIC 16 A1 -33.556444 5 77 .112888 17 A2 -33.158811 8 82 . 317623 18 A3 -33.556444 5 77 .112888 19 fitted -33.556444 5 77 .112888 20 R -47.392394 2 98 . 7 8 4 7 8 8 21 22 23 E x p l a n a t i o n o f T e s t s 24 25 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 26 (A2 vs. R) 27 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 28 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 29 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 30 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 31 32 T e s t s o f I n t e r e s t 33 34 Test -2*log(Likelihood Ratio) Test df p-vali 35 36 T e s t 1 28.4672 6 <.0001 37 T e s t 2 0.795266 3 0.8506 38 T e s t 3 0.795266 3 0.8506 39 T e s t 4 6.96332e-013 0 NA 40 41 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 42 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 43 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 44 45 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 46 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 47 48 49 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 50 t o b e a p p r o p r i a t e h e r e 51 52 N A - D e g r e e s o f f r e e d o m f o r T e s t 4 a r e l e s s t h a n o r e q u a l t o 0. T h e C h i - S q u a r e 53 t e s t f o r f i t is n o t v a l i d 54 55 56 B e n c h m a r k D o s e C o m p u t a t i o n 57 58 S p e c i f i e d e f f e c t = 1 59 60 R i s k T y p e = Estimated standard deviations from the control mean 61 62 C o n f i d e n c e l e v e l = 0.95 63 64 BMD = 0.0995543 65 66 B M D L = 1 . 2 2 8 1 8 e - 0 0 9 67 68 This document is a draftfor review purposes only and does not constitute Agency policy. E-121 DRAFT--DO NOT CITE OR QUOTE 1 E.2.21.5. F igure f o r A d d itio n a l M o d el Presented: H ill, U nrestricted Hill Model with 0.95 Confidence Level Mean Response 2 10:53 02/08 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-122 DRAFT--DO NOT CITE OR QUOTE 1 E.2.22. Li et al., 1997: FSH 2 E.2.22.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 8 <0.0001 1095.292 5.222E+02 4.121E+02 Notes exponential (M 3) 8 < 0 .0 0 0 1 1 0 95.292 5 .2 2 2 E + 0 2 4 .1 2 1 E + 0 2 p o w er hit bound (d = 1) exponential (M 4) 7 <0.0001 1059.480 3.432E+01 9.930E+00 exponential (M 5) 6 <0.0001 1066.195 1.019E+02 8.583E-01 H ill 7 <0.0001 1056.459 5.423E +00 error n lo w er b ound hit (n = 1) linear 8 <0.0001 1077.695 2.003E+02 1.357E+02 polynom ial, 8degree power b H ill, unrestricted 9 <0.0001 1155.670 error 1 .9 1 6 E + 0 2 8 <0.0001 1077.695 2.003E+02 1.357E+02 power bound hit (power = 1) 6 0.001 1039.481 2.204E -01 error unrestricted (n = 0.32) power, unrestricted c 7 0.002 1037.474 1.963E-01 2.484E -02 unrestricted (pow er = 0.305) a N on-con stant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .2 2 .2 . Outputfor Selected Model: Power 6 Li et al., 1997: FSH 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 7 2 _ L i _ 1 9 9 7 _ F S H _ P w r _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 7 2 _ L i _ 1 9 9 7 _ F S H _ P w r _ 1 . p l t 13 M o n F e b 0 8 ~ 1 3 : 3 6 : 3 5 2 0 1 0 14 15 16 F i g u r e 3: F S H i n f e m a l e S - D r a t s 2 4 h r a f t e r d o s i n g , 22 d a y o l d r a t s 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 T h e p o w e r is r e s t r i c t e d t o b e g r e a t e r t h a n o r e q u a l t o 1 This document is a draftfor review purposes only and does not constitute Agency policy. E-123 DRAFT--DO NOT CITE OR QUOTE 1 The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i)) rho) 2 3 Total number of dose groups = 10 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 12 lalpha = 9.8191 13 rho = 0 14 control = 22.1591 15 slope = 52.284 16 power = 0.294106 17 18 19 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 20 21 ( * * * T h e m o d e l p a r a m e t e r ( s ) - p o w e r 22 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 23 a n d d o n o t a p p e a r in t h e c o r r e l a t i o n m a t r i x ) 24 25 lalpha rho control slope 26 27 l a l p h a 1 -0.99 -0.29 -0.033 28 29 rho -0.99 1 0.2 0.033 30 CO o CO o 31 c o n t r o l -0.29 0.2 1 32 33 slope -0.033 0.033 1 34 35 36 37 P a r a m e t e r E s t i m a t e s 38 39 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 40 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 41 lalpha 3.50054 1.225 1.09958 5.9015 42 rho 1.27087 0.241869 0.796814 1.74492 43 control 87.4348 12.9347 62.0833 112.786 44 slope 0.492306 0.0919718 0.312044 0.672567 45 p o w e r 1 NA 46 47 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 48 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 49 h a s n o s t a n d a r d e r r o r . 50 51 52 53 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 54 55 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 56 57 58 0 10 23.9 87.4 29.6 98.6 -2.04 59 0 . 2 6 6 10 22.2 87.6 48.5 98.7 -2.1 60 0 . 7 9 8 8 10 85.2 87.8 94.3 98.9 -0.0832 61 2 . 0 9 7 10 73.3 88.5 48.5 99.4 -0.483 62 5 . 8 6 7 10 126 90.3 159 101 1.12 63 15 10 132 94.8 116 104 1.14 64 4 3 . 3 3 10 117 109 51.2 113 0.223 65 1 1 9 . 9 10 304 146 154 137 3.65 66 386 10 347 277 151 205 1.07 67 1172 10 455 664 286 358 -1.85 68 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-124 DRAFT--DO NOT CITE OR QUOTE 1 Model Descriptions for likelihoods calculated 2 3 4 Model A1: Yij Mu(i) + e(ij) 5 Var{e(ij)} Sigma^2 6 7 Model A2: Yij Mu(i) + e(ij) 8 Var{e(ij)} Sigma(i)^2 9 10 M o d e l A3: Yij = Mu(i) + e(ij) 11 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 12 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 13 w e r e s p e c i f i e d b y t h e u s e r 14 15 M o d e l R: Yi = M u + e(i) 16 V a r { e ( i ) } = S i g m a ^ 2 17 18 19 L i k e l i h o o d s o f I n t e r e s t 20 21 Model Log(likelihood) # Param's AIC 22 A1 -535.687163 11 1 0 9 3 . 3 7 4 3 2 7 23 A2 -496.367061 20 1032.734122 24 A3 -502.709623 12 1 0 2 9 . 4 1 9 2 4 6 25 fitted -534.847518 4 1077.695035 26 R -574.835246 2 1153.670492 27 28 29 E x p l a n a t i o n o f T e s t s 30 31 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 32 (A2 vs. R) 33 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 34 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 35 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 36 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 37 38 T e s t s o f I n t e r e s t 39 40 Test -2*log(Likelihood Ratio) Test df p-val 41 42 T e s t 1 156.936 18 <.0001 43 T e s t 2 78.6402 9 <.0001 44 T e s t 3 12.6851 8 0.1232 45 T e s t 4 64.2758 8 <.0001 46 47 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 48 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 49 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 50 51 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 52 m o d e l a p p e a r s t o b e a p p r o p r i a t e 53 54 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 55 t o b e a p p r o p r i a t e h e r e 56 57 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 58 m o d e l 59 60 61 B e n c h m a r k D o s e C o m p u t a t i o n 62 63 S p e c i f i e d e f f e c t = 1 64 65 R i s k T y p e = Estimated standard deviations from the control mean 66 67 C o n f i d e n c e l e v e l = 0.95 68 69 B M D = 2 0 0 . 3 1 4 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-125 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 2 BMDL 135.673 3 4 5 E .2 .2 2 .3 . Figurefor Selected Model: Power Power Model with 0.95 Confidence Level 6 7 8 9 E .2.22.4. Outputfor Additional Model Presented: Power, Unrestricted 10 L i et al., 1997: F S H 11 12 13 14 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 15 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 7 2 _ L i _ 1 9 9 7 _ F S H _ P w r _ U _ 1 . ( d ) 16 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 7 2 _ L i _ 1 9 9 7 _ F S H _ P w r _ U _ 1 . p l t 17 M o n F e b 08 1 3 : 3 6 : 4 6 2 0 1 0 18 19 20 F i g u r e 3: F S H i n f e m a l e S - D r a t s 2 4 h r a f t e r d o s i n g , 22 d a y o l d r a t s 21 22 23 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 24 25 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 26 27 28 D e p e n d e n t v a r i a b l e = M e a n This document is a draftfor review purposes only and does not constitute Agency policy. E-126 DRAFT--DO NOT CITE OR QUOTE 1 Independent variable = Dose 2 The power is not restricted 3 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) rho) 4 5 Total number of dose groups = 10 6 Total number of records with missing values = 0 7 Maximum number of iterations = 250 8 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 9 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 10 11 12 13 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 14 lalpha = 9.8191 15 rho = 0 16 control = 22.1591 17 slope = 52.284 18 power = 0.294106 19 20 21 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 22 23 lalpha rho control slope power 24 Lf) CO o Lf) o CO oo o o 25 l a l p h a 1 -0.99 -0.69 0.26 26 27 rho -0.99 1 0.65 0.0089 -0.23 28 29 c o n t r o l -0.69 0.65 1 -0.23 0.029 30 31 s l o p e 0.0089 -0.23 1 32 33 p o w e r 0.26 -0.23 0.029 1 34 35 36 37 P a r a m e t e r E s t i m a t e s 38 39 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 40 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 41 lalpha 3.67487 1.12134 1.47708 5.87265 42 rho 1.17882 0.221526 0.744632 1.613 43 control 15.8201 6.87715 2.34113 29.299 44 slope 52.528 9.46821 33.9706 71.0853 45 power 0.304867 0.0336805 0.238855 0.37088 46 47 48 49 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 50 51 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 52 53 54 0 10 23.9 15.8 29.6 32 0.795 CO \-- 1 55 0 . 2 6 6 10 22.2 50.9 48.5 63.7 56 0 . 7 9 8 8 10 85.2 64.9 94.3 73.5 0.876 57 2 . 0 9 7 10 73.3 81.7 48.5 84.1 -0.314 58 5 . 8 6 7 10 126 106 159 98.1 0.652 59 15 10 132 136 116 114 -0.102 60 4 3 . 3 3 10 117 182 51.2 135 -1.52 61 1 1 9 . 9 10 304 242 154 160 1.24 62 386 10 347 339 151 195 0.134 63 1172 10 455 469 286 236 -0.182 64 65 66 67 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 68 69 70 M o d e l A1: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-127 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma^2 2 3 Model A2: Yij = Mu(i) + e(ij) 4 Var{e(ij)} = Sigma(i)^2 5 6 Model A3: Yij = Mu(i) + e(ij) 7 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 8 Model A3 uses any fixed variance parameters that 9 were specified by the user 10 11 M o d e l R: Yi = M u + e(i) 12 V a r { e ( i ) } = S i g m a ^ 2 13 14 15 L i k e l i h o o d s o f I n t e r e s t 16 17 Model Log(likelihood) # Param's AIC 18 A1 -535.687163 11 1 0 9 3 . 3 7 4 3 2 7 19 A2 -496.367061 20 1032.734122 20 A3 -502.709623 12 1 0 2 9 . 4 1 9 2 4 6 21 fitted -513.737215 5 1037.474431 22 R -574.835246 2 1153.670492 23 24 25 E x p l a n a t i o n o f T e s t s 26 27 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 28 (A2 vs. R) 29 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 30 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 31 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 32 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 33 34 T e s t s o f I n t e r e s t 35 36 Test -2*log(Likelihood Ratio) Test df p-value 37 38 T e s t 1 156.936 18 <.0001 39 T e s t 2 78.6402 9 <.0001 40 T e s t 3 12.6851 8 0.1232 41 T e s t 4 22.0552 7 0.002485 42 43 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 44 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 45 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 46 47 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 48 m o d e l a p p e a r s t o b e a p p r o p r i a t e 49 50 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 51 t o b e a p p r o p r i a t e h e r e 52 53 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 54 m o d e l 55 56 57 B e n c h m a r k D o s e C o m p u t a t i o n 58 59 S p e c i f i e d e f f e c t = 1 60 61 R i s k T y p e = Estimated standard deviations from the control mean 62 63 C o n f i d e n c e l e v e l = 0.95 64 65 B M D = 0 . 1 9 6 2 7 8 66 67 68 B M D L = 0 . 0 2 4 8 3 6 4 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-128 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E.2.22.5. F igure f o r A d d itio n a l M o d el Presented: Power, Unrestricted Power Model with 0.95 Confidence Level 2 13:36 02/08 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-129 DRAFT--DO NOT CITE OR QUOTE 1 E.2.23. Li et al., 2006: E stradiol, 3-D ay 2 E .2 .2 3 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 2 0.156 269.027 1.416E+01 5.544E+00 Notes exponential (M 3) 2 0 .1 5 6 2 6 9 .0 2 7 1.416E +01 5 .5 4 4 E + 0 0 p ow er hit b ound (d = 1) exponential (M 4) 1 0.341 268.212 error error exponential (M 5) 0 N /A 270.212 error error H ill 0 N /A 270.212 error error linear b polynom ial, 3degree power 2 0.162 268.952 1.606E+01 5.379E+00 2 0.162 268.952 1.606E+01 5.379E+00 2 0 .1 6 2 2 6 8 .9 5 2 1.606E +01 5 .3 7 9 E + 0 0 p ow er b oun d hit (p ow er = 1) H ill, unrestricted 0 N /A 270.265 9.273E+12 9.273E+12 unrestricted (n = 0.03) power, unrestricted 1 0.328 268.265 9.455E +10 error a C onstant variance m odel selected (p = 0 .4 3 7 2 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2 .2 3 .2 . Outputfor Selected Model: Linear unrestricted (power = 0.015) 6 Li et al., 2006: Estradiol, 3-D ay 7 8 9 10 P o l y n o m i a l M o d e l . ( V e r s i o n : 2 . 1 3 ; D a t e : 0 4 / 0 8 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 3 1 _ L i _ 2 0 0 6 _ E s t r a _ L i n e a r C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 3 1 _ L i _ 2 0 0 6 _ E s t r a _ L i n e a r C V _ 1 . p l t 13 M o n F e b 08 1 0 : 5 4 : 0 0 2 0 1 0 14 15 16 F i g u r e 3, 3 - d a y e s t r a d i o l 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 S i g n s o f t h e p o l y n o m i a l c o e f f i c i e n t s a r e n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t This document is a draftfor review purposes only and does not constitute Agency policy. E-130 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 4 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha = 267.211 12 rho = 0S p e c i f i e d 13 beta_0 = 16.1705 14 beta 1 = 1.0106 15 16 17 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 18 19 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 20 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 21 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 22 23 alpha beta_0 beta_1 24 25 a l p h a 1 2.1e-012 5e-014 26 27 beta 0 2.1e-012 1 -0.69 28 29 beta 1 5e-014 -0.69 1 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 36 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 37 alpha 263.435 58.9057 147.981 378.888 38 beta_0 16.1705 3.55949 9.19407 23.147 39 beta 1 1.0106 1.2148 -1.37037 3.39156 40 41 42 43 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 44 45 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 46 47 48 0 10 10.2 16.2 12.2 16.2 -1.17 49 0 . 1 5 8 8 10 19.9 16.3 20 16.2 0.697 50 2 . 8 3 9 10 24.7 19 14.6 16.2 1.11 51 5 . 1 2 4 10 18.1 21.3 17.6 16.2 -0.635 52 53 54 55 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 56 57 58 M o d e l A1: Yij = Mu(i) + e(ij) 59 V a r { e i ;i j ) } = S i g m a ^ 2 60 61 M o d e l A2: Yij = Mu(i) + e(ij) 62 V a r { e i :i j ) } = S i g m a ( i ) ^2 63 64 M o d e l A3: Yij = Mu(i) + e(ij) 65 V a r { e i :i j ) } = S i g m a ^ 2 66 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 67 w e r e s p e c i f i e d b y t h e u s e r 68 69 M o d e l R: Yi = M u + e(i) 70 V a r { e ( i ) } = S i g m a ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-131 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Likelihoods of Interest 4 5 Model Log(likelihood) # Param's AIC 6 A1 -129.653527 5 269.307054 7 A2 -128.294657 8 272.589314 8 A3 -129.653527 5 269.307054 9 fitted -131.476097 3 268.952193 10 R -131.819169 2 267.638338 11 12 13 E x p l a n a t i o n o f T e s t s 14 15 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 16 (A2 vs. R) 17 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 18 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 19 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 20 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 21 22 T e s t s o f I n t e r e s t 23 24 Test -2*log(Likelihood Ratio ) Test df p-value 25 26 T e s t 1 7.04902 6 0.3163 27 T e s t 2 2.71774 3 0.4372 28 T e s t 3 2.71774 3 0.4372 29 T e s t 4 3.64514 2 0.1616 30 31 p - v a l u e f o r T e s t 1 is g r e a t e r t h a n .05. T h e r e m a y n o t b e 32 d i f f e n c e b e t w e e n r e s p o n s e s a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 33 M o d e l l i n g t h e d a t a w i t h a d o s e / r e s p o n s e c u r v e m a y n o t b e a p p r o p r i a t e 34 35 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 36 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 37 38 39 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 40 t o b e a p p r o p r i a t e h e r e 41 42 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 43 t o a d e q u a t e l y d e s c r i b e t h e d a t a 44 45 46 B e n c h m a r k D o s e C o m p u t a t i o n 47 48 S p e c i f i e d e f f e c t = 1 49 50 R i s k T y p e = Estimated standard deviations from the control mean 51 52 C o n f i d e n c e l e v e l = 0.95 53 54 BMD = 16.0605 55 56 57 BMDL = 5.37895 58 This document is a draftfor review purposes only and does not constitute Agency policy. E-132 DRAFT--DO NOT CITE OR QUOTE 1 E.2.23.3. F igure f o r Selected M odel: L in ea r Linear Model with 0.95 Confidence Level Mean Response 2 10:54 02/08 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-133 DRAFT--DO NOT CITE OR QUOTE 1 E.2.24. Li et al., 2006: Progesterone, 3-D ay 2 E .2 .2 4 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 2 <0.001 329.928 2.619E +00 error Notes exponential (M 3) 2 0.001 328.101 1.340E-01 error p o w er hit boun d (d = 1) exponential (M 4) 1 0.384 315.734 1.074E-02 6.633E-03 exponential (M 5) Hill b 0 N /A 317.734 4.301E-02 4.272E-03 1 0.386 315.728 9.461E-04 8.006E-11 n lower bound hit (n = 1) linear 2 <0.001 330.729 3.891E+00 2.626E+00 polynom ial, 3degree 2 <0.001 330.729 3.891E+00 2.626E+00 power 2 < 0 .0 0 1 3 3 0 .7 2 9 3 .8 9 1 E + 0 0 2 .6 2 6 E + 0 0 p o w er bound hit (p ow er = 1) power, unrestricted 1 0.404 315.673 2.812E -59 2.812E-59 unrestricted (power = 0.01) a N on-con stant variance m odel selected (p = 0 .0 0 1 3 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2 .2 4 .2 . Outputfor Selected Model: Hill 6 Li et al., 2006: Progesterone, 3-D ay 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 3 2 _ L i _ 2 0 0 6 _ P r o g e s t _ H i l l _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 3 2 _ L i _ 2 0 0 6 _ P r o g e s t _ H i l l _ 1 . p l t 13 W e d F e b 1 0 ~ 1 0 : 5 7 : 1 4 2 0 1 0 14 15 16 F i g u r e 4, 3 - d a y p r o g e s t e r o n e 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 P o w e r p a r a m e t e r r e s t r i c t e d t o b e g r e a t e r t h a n 1 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + r h o * l n ( m e a n ( i ) ) ) 28 29 T o t a l n u m b e r o f d o s e g r o u p s = 4 30 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 31 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 This document is a draftfor review purposes only and does not constitute Agency policy. E-134 DRAFT--DO NOT CITE OR QUOTE 1 Relative Function h a s b e e n s e t to: 1e-008 2 Parameter has b e e n set to: 1e-008 3 4 5 6 Default Initial Parameter Values 7 lalpha 7.08699 8 rho 0 9 intercept 61.7404 10 - 5 0 . 3 8 3 5 11 1 . 4 7 2 8 6 12 0 . 1 2 8 3 0 2 13 14 15 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 16 17 ( * * * T h e m o d e l p a r a m e t e r ( s ) -n 18 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 19 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 20 21 lalpha rho intercept v k 22 23 l a l p h a 1 -0.99 -0.093 0.82 0.22 24 25 rho -0.99 1 0.12 -0.79 -0.2 26 CO o CO o 27 i n t e r c e p t -0.093 0.12 1 0.014 28 29 v 0.82 -0.79 1 0.035 30 31 k 0.22 -0.2 0.014 0.035 1 32 33 34 35 P a r a m e t e r E s t i m a t e s 36 37 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 38 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 39 lalpha 14.0902 3.36095 7.50284 20.6775 40 rho -2.27438 0.861553 -3.963 -0.585772 41 ntercept 61.7488 3.3373 55.2078 68.2898 42 v -42.1007 7.70852 -57.2091 -26.9922 43 n 1 N A 44 k 0.00282851 0.020619 -0.037584 0.0432411 45 46 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 47 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 48 h a s n o s t a n d a r d e r r o r . 49 50 51 52 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 53 54 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 55 56 57 0 10 61.7 61.7 11.1 10.6 -0.00251 58 0 . 1 5 8 8 10 30.6 20.4 40.5 37.2 0.865 59 2 . 8 3 9 10 16.9 19.7 33.3 38.7 -0.225 60 5 . 1 2 4 10 11.4 19.7 43.7 38.8 -0.678 61 62 63 64 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 65 66 67 M o d e l A1: Yij = Mu(i) + e(ij) 68 V a r { e ( i j ) } = S i g m a ^ 2 69 70 M o d e l A2: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-135 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma(i)^2 2 3 Model A3: Yij = Mu(i) + e(ij) 4 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 5 Model A3 uses any fixed variance parameters that 6 were specified by the user 7 8 M o d e l R: Yi = M u + e(i) 9 Var{e(i)} = Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) # Param's AIC 15 A1 -159.632675 5 329.265349 16 A2 -151.812765 8 319.625529 17 A3 -152.488175 6 316.976349 18 fitted -152.863841 5 315.727683 19 R -165.698875 2 335.397750 20 21 22 E x p l a n a t i o n o f T e s t s 23 24 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 25 (A2 vs. R) 26 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 27 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 28 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 29 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 30 31 T e s t s o f I n t e r e s t 32 33 Test -2*log(Likelihood Ratio) Test df p-value 34 35 T e s t 1 27.7722 6 0.0001037 36 T e s t 2 15.6398 3 0.001344 37 T e s t 3 1.35082 2 0.5089 38 T e s t 4 0.751333 1 0.3861 39 40 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 41 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 42 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 43 44 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 45 m o d e l a p p e a r s t o b e a p p r o p r i a t e 46 47 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 48 t o b e a p p r o p r i a t e h e r e 49 50 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 51 t o a d e q u a t e l y d e s c r i b e t h e d a t a 52 53 54 B e n c h m a r k D o s e C o m p u t a t i o n 55 56 S p e c i f i e d e f f e c t : 1 57 58 R i s k T y p e Estimated standard deviations from the control mean 59 60 C o n f i d e n c e l e v e l : 0.95 61 62 BMD ; 0.000946102 63 64 B M D L ; 8 . 0 0 6 3 9 e - 0 1 1 65 This document is a draftfor review purposes only and does not constitute Agency policy. E-136 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E .2 .2 4 .3 . Figurefor Selected Model: Hill Hill Model with 0.95 Confidence Level 2 10:57 02/10 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-137 DRAFT--DO NOT CITE OR QUOTE 1 E.2.25. M arkow ski et al., 2001: FR 10 R un O pportunities 2 E .2 .2 5 .1 . Summary Table o f BMDS Modeling Results Model a exponential (M2) b Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) 2 0.304 117.150 8.570E+00 2.887E+00 Notes exponential (M 3) 2 0.304 1 1 7.150 8 .5 7 0 E + 0 0 2 .8 8 7 E + 0 0 p o w er hit bound (d = 1) exponential (M 4) 1 0.371 117.570 3.452E+00 1.299E-02 exponential (M 5) 0 N /A 118.918 2.315E+00 1.391E-02 Hill 0 N /A 118.918 1.801E+00 1.274E-09 linear 2 0.226 117.744 1.106E+01 5.741E+00 polynom ial, 3degree 2 0.226 117.744 1.106E+01 5.741E+00 power 2 0 .2 2 6 1 1 7.744 1.106E +01 5 .7 4 1 E + 0 0 p o w er b oun d hit (p ow er = 1) power, unrestricted 1 0.239 118.158 5.768E +00 1.032E-14 unrestricted (pow er = 0.276) a C onstant variance m odel selected (p = 0 .1 7 1 9 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2 .2 5 .2 . Outputfor Selected Model: Exponential (M2) 6 M arkowski et al., 2001: FR 10 Run Opportunities 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 3 3 _ M a r k _ 2 0 0 1 _ F R 1 0 o p p _ E x p C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 M o n F e b 08 1 0 : 5 5 : 1 3 2 0 1 0 14 15 16 T a b l e 3 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 30 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 31 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. This document is a draftfor review purposes only and does not constitute Agency policy. E-138 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Dependent variable = Mean 4 Independent variable = Dose 5 Data are assumed to be distributed: normally 6 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 7 r h o is s e t t o 0. 8 A constant v a r i a n c e model is fit. 9 10 T o t a l n u m b e r o f d o s e g r o u p s = 4 11 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 12 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 13 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 16 M L E s o l u t i o n p r o v i d e d : E x a c t 17 18 19 I n i t i a l P a r a m e t e r V a l u e s 20 21 Variable Model 2 22 23 lnalpha 3.5321 24 rho(S) 0 25 a 6 . 7 7 9 7 5 26 b 0 . 0 5 8 1 9 3 7 27 c 0 28 d 1 29 30 (S) = S p e c i f i e d 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 Variable Model 2 37 38 lnalpha 3.63127 39 rho 0 40 a 1 2 . 2 9 0 1 41 b 0 . 0 8 0 8 8 3 2 42 c 0 43 d 1 44 45 46 T a b l e o f S t a t s F r o m I n p u t D a t a 47 48 Dose N Obs Mean Obs Std Dev 49 50 07 13.29 8.65 51 1.557 4 11.25 5.56 52 4.03 6 5.75 3.53 53 10.32 7 7 6.01 54 55 56 E s t i m a t e d V a l u e s o f I n t e r e s t 57 58 Dose Est Mean Est Std Scaled Residual 59 60 0 12.29 6.145 0.4305 61 1 . 5 5 7 10.84 6.145 0.1347 62 4.03 8.871 6.145 -1.244 63 1 0 . 3 2 5.335 6.145 0.717 64 65 66 67 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d 68 69 M o d e l A1: Yij = M u ( i ) + e(i j ) 70 V a r { e ( i j ) } = S i g m a ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-139 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A2: Yij = Mu(i) + e(ij) 3 V a r { e i ;ij)} = S i g m a ( i ) ^ 2 4 5 Model A3: Yij = Mu(i) + e(ij) 6 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 7 8 M o d e l R: Yij = M u + e(i) 9 V a r { e i :ij)} = S i g m a ^ 2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) DF AIC 15 16 A1 -54.38526 5 118.7705 17 A2 -51.88568 8 119.7714 18 A3 -54.38526 5 118.7705 19 R -57.45429 2 118.9086 20 2 -55.57522 3 117.1504 21 22 23 Additive constant for all log-likelihoods = -22.05. This constant added to the 24 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 25 d e p e n d o n t h e m o d e l p a r a m e t e r s . 26 27 28 E x p l a n a t i o n o f T e s t s 29 30 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 31 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 32 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 33 T e s t 4: D o e s M o d e l 2 f i t t h e d a t a ? (A3 vs. 2) 34 35 36 T e s t s o f I n t e r e s t 37 38 T e s t -2*log(Likelihood Ratio) D. F. p-value 39 40 T e s t 1 11.14 6 0.08423 41 T e s t 2 4.999 3 0.1719 42 T e s t 3 4.999 3 0.1719 43 T e s t 4 2.38 2 0.3042 44 45 46 T h e p - v a l u e f o r T e s t 1 is g r e a t e r t h a n .05. T h e r e m a y n o t b e a 47 d i f f e n c e b e t w e e n r e s p o n s e s a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 48 M o d e l l i n g t h e d a t a w i t h a d o s e / r e s p o n s e c u r v e m a y n o t b e a p p r o p r i a t e . 49 50 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 51 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e . 52 53 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 54 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 55 56 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. M o d e l 2 s e e m s 57 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 58 59 60 B e n c h m a r k D o s e C o m p u t a t i o n s : 61 62 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 63 64 R i s k T y p e = E s t i m a t e d s t a n d a r d d e v i a t i o n s f r o m c o n t r o l 65 66 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 67 68 BMD = 8.56961 69 70 BMDL = 2.88708 This document is a draftfor review purposes only and does not constitute Agency policy. E-140 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E .2 .2 5 .3 . Figurefor Selected Model: Exponential (M2) Exponential Model 2 with 0.95 Confidence Level 2 10:55 02/08 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-141 DRAFT--DO NOT CITE OR QUOTE 1 E.2.26. M arkow ski et al., 2001: FR 2 R evolutions 2 E .2 .2 6 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 2 0.236 217.219 8.486E+00 3.232E+00 Notes exponential (M 3) 2 0 .2 3 6 2 1 7 .2 1 9 8 .4 8 6 E + 0 0 3 .2 3 2 E + 0 0 p o w er hit b ound (d = 1) exponential (M 4) 1 0.263 217.583 3.413E+00 1.766E-02 exponential (M 5) Hill b 0 N /A 218.532 2.415E+00 9.313E-01 1 0.654 216.532 1.840E+00 5.992E-01 n upper bound hit (n = 18) linear 2 0.180 217.764 1.058E+01 5.602E+00 polynom ial, 3degree 2 0.180 217.764 1.058E+01 5.602E+00 power 2 0 .1 8 0 2 1 7 .7 6 4 1.058E +01 5 .6 0 2 E + 0 0 p o w er b oun d hit (p ow er = 1) power, unrestricted c 1 0.161 218.294 5.739E +00 1.032E-14 unrestricted (power = 0.318) a C onstant variance m odel selected (p = 0 .1 0 9 2 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .2 6 .2 . Outputfor Selected Model: Hill 6 M arkowski et al., 2001: FR2 R evolutions 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 3 4 _ M a r k _ 2 0 0 1 _ F R 2 r e v _ H i l l C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 3 4 _ M a r k _ 2 0 0 1 _ F R 2 r e v _ H i l l C V _ 1 . p l t 13 M o n F e b 08 1 0 : 5 5 : 4 7 2 0 1 0 14 15 16 T a b l e 3 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 P o w e r p a r a m e t e r r e s t r i c t e d t o b e g r e a t e r t h a n 1 28 A c o n s t a n t v a r i a n c e m o d e l is f i t 29 This document is a draftfor review purposes only and does not constitute Agency policy. E-142 DRAFT--DO NOT CITE OR QUOTE 1 Total number of dose groups = 4 2 Total number of records with missing values = 0 3 Maximum number of iterations = 250 4 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 5 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 6 7 8 9 Default Initial Parameter Values 10 alpha = 2598.74 11 rho = 0S p e c i f i e d 12 intercept = 119.29 13 v= -62.79 14 n= 2.13752 15 k= 2.53662 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 ( *** The model parameter(s) -rho -n 21 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 22 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 23 24 alpha intercept v k 25 26 a l p h a 1 1.2e-008 1e-009 3.5e-008 27 28 i n t e r c e p t 1.2e-008 1 -0.81 -0.52 29 30 v 1e-009 -0.81 1 0.37 31 32 k 3.5e-008 -0.52 0.37 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 39 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 40 alpha 2183.85 630.425 948.245 3419.46 41 i n t e r c e p t 119.29 17.6629 84.6713 153.909 42 v -56.5223 21.9082 -99.4615 -13.5831 43 n 18 N A 44 k 1.68653 0.295154 1.10804 2.26502 45 46 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 47 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 48 h a s n o s t a n d a r d e r r o r . 49 50 51 52 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 53 54 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled R 55 56 57 0 7 119 119 69.9 46.7 -2.41e-007 58 1 . 5 5 7 4 109 108 61 46.7 2.29e-007 59 4 . 0 3 6 56.5 62.8 31.2 46.7 -0.329 60 1 0 . 3 2 7 68.1 62.8 33.2 46.7 0.304 61 62 63 64 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 65 66 67 M o d e l A1: Yij = Mu(i) + e(ij) 68 V a r { e (ij)} = S i g m a ^ 2 69 70 M o d e l A2: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-143 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma(i)^2 2 3 Model A3: Yij = Mu(i) + e(ij) 4 Var{e(ij)} = Sigma^2 5 Model A3 uses any fixed variance parameters that 6 were specified by the user 7 8 M o d e l R: Yi = M u + e(i) 9 Var{e(i)} = Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) # Param's AIC 15 A1 -104.165520 5 218.331040 16 A2 -101.140174 8 218.280349 17 A3 -104.165520 5 218.331040 18 fitted -104.266162 4 216.532324 19 R -107.599268 2 219.198536 20 21 22 E x p l a n a t i o n o f T e s t s 23 24 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 25 (A2 vs. R) 26 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 27 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 28 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 29 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 30 31 T e s t s o f I n t e r e s t 32 33 Test -2*log(Likelihood Ratio) Test df p-val 34 35 T e s t 1 12.9182 6 0.04435 36 T e s t 2 6.05069 3 0.1092 37 T e s t 3 6.05069 3 0.1092 38 T e s t 4 0.201284 1 0.6537 39 40 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 41 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 42 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 43 44 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 45 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 46 47 48 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 49 t o b e a p p r o p r i a t e h e r e 50 51 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 52 t o a d e q u a t e l y d e s c r i b e t h e d a t a 53 54 55 B e n c h m a r k D o s e C o m p u t a t i o n 56 57 S p e c i f i e d e f f e c t = 1 58 59 R i s k T y p e = Estimated standard deviations from the control mean 60 61 C o n f i d e n c e l e v e l = 0.95 62 63 BMD = 1.83952 64 65 BMDL = 0.599228 66 This document is a draftfor review purposes only and does not constitute Agency policy. E-144 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E .2 .2 6 .3 . Figurefor Selected Model: Hill Hill Model with 0.95 Confidence Level 2 3 4 5 E .2.26.4. Outputfor Additional Model Presented: Power, Unrestricted 6 M arkowski et al., 2001: FR2 R evolutions 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 3 4 _ M a r k _ 2 0 0 1 _ F R 2 r e v _ P o w e r C V _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 3 4 _ M a r k _ 2 0 0 1 _ F R 2 r e v _ P o w e r C V _ U _ 1 . p l t 13 M o n F e b 08 1 0 : 5 5 : 4 9 2 0 1 0 14 15 16 T a b l e 3 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 T h e p o w e r is n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t This document is a draftfor review purposes only and does not constitute Agency policy. E-145 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 4 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha = 2598.74 12 rho = 0 Specified 13 control = 119.29 14 slope = -10.3599 15 power = 0.824761 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 21 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 22 a n d d o n o t a p p e a r in t h e c o r r e l a t i o n m a t r i x ) 23 24 alpha control slope power 25 26 a l p h a 1 -3e-010 6.9e-010 9.9e-010 27 o CK>O o CK>O CCOO o Cc oO o 28 control -3e-010 1 29 30 slope 6.9e-010 1 0.87 31 32 power 9.9e-010 0.87 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 39 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 40 alpha 2350.22 678.449 1020.48 3679.95 41 control 120.082 18.0782 84.6491 155.514 42 slope -27.8164 24.2447 -75.3352 19.7023 43 power 0.317923 0.350841 -0.369713 1.00556 44 45 46 47 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 48 49 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 50 51 52 0 7 119 120 69.9 48.5 -0.0432 53 1 . 5 5 7 4 109 88.1 61 48.5 0.843 54 4 . 0 3 6 56.5 76.8 31.2 48.5 -1.02 55 1 0 . 3 2 7 68.1 61.7 33.2 48.5 0.353 56 57 58 1-- 1 CD XC H 1-- 1 59 M o d e l D e s c r i p t i o n s f o r ihood s calculated 60 61 62 M o d e l A1: Yij = M u ( i ) + e( ij) 63 V a r { e (ij)} = Sigma'"2 64 65 M o d e l A2: Yij = M u ( i ) + e( ij) 66 V a r { e (ij)} = S i g m a (i)1"2 67 68 M o d e l A3: Yij = M u ( i ) + e( ij) 69 V a r { e (ij)} = Sigma'"2 70 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t This document is a draftfor review purposes only and does not constitute Agency policy. E-146 DRAFT--DO NOT CITE OR QUOTE 1 were specified by the user 2 3 M o d e l R: Yi = M u + e(i) 4 Var{e(i)} = Sigma^2 5 6 7 Likelihoods of Interest 8 9 Model Log(likelihood) # Param's AIC 10 A1 -104.165520 5 218.331040 11 A2 -101.140174 8 218.280349 12 A3 -104.165520 5 218.331040 13 fitted -105.147159 4 218.294317 14 R -107.599268 2 219.198536 15 16 17 E x p l a n a t i o n o f T e s t s 18 19 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 20 (A2 vs. R) 21 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 22 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 23 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 24 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 25 26 T e s t s o f I n t e r e s t 27 28 Test -2*log(Likelihood Ratio) Test df p-val 29 30 T e s t 1 12 .9182 6 0. 0 4 4 3 5 31 T e s t 2 6. 0 5 0 6 9 3 0 . 1 0 9 2 32 T e s t 3 6. 0 5 0 6 9 3 0 . 1 0 9 2 33 T e s t 4 1. 9 6 3 2 8 1 0 . 1 6 1 2 34 35 p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t 36 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 37 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 38 39 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 40 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 41 42 43 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 44 t o b e a p p r o p r i a t e h e r e 45 46 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 47 t o a d e q u a t e l y d e s c r i b e t h e d a t a 48 49 50 B e n c h m a r k D o s e C o m p u t a t i o n 51 52 S p e c i f i e d e f f e c t = 1 53 54 R i s k T y p e = Estimated standard deviations from the control mean 55 56 C o n f i d e n c e l e v e l = 0.95 57 58 B M D = 5 . 7 3 9 0 6 59 60 61 B M D L = 1 . 0 3 1 8 1 e - 0 1 4 62 63 This document is a draftfor review purposes only and does not constitute Agency policy. E-147 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E.2.26.5. F igure f o r A d d itio n a l M o d el Presented: Power, U nrestricted Power Model with 0.95 Confidence Level 2 10:55 02/08 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-148 DRAFT--DO NOT CITE OR QUOTE 1 E.2.27. M arkow ski et al., 2001: FR 5 R un O pportunities 2 E .2 .2 7 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 2 0.205 133.193 5.078E+00 2.439E+00 Notes exponential (M 3) 2 0.205 133.193 5 .0 7 8 E + 0 0 2 .4 3 9 E + 0 0 p o w er hit b ound (d = 1) exponential (M 4) 1 0.254 133.328 2.160E+00 6.854E-01 exponential (M 5) Hill b 0 N /A 134.032 2.124E+00 9.667E-01 1 0.939 132.032 1.723E+00 9.085E-01 n upper bound hit (n = 18) linear 2 0.122 134.229 7.234E+00 4.430E+00 polynom ial, 3degree 2 0.122 134.229 7.234E+00 4.430E+00 power 2 0 .1 2 2 1 3 4.229 7 .2 3 4 E + 0 0 4 .4 3 0 E + 0 0 p o w er b oun d hit (p ow er = 1) power, unrestricted c 1 0.134 134.268 2.666E +00 1.032E-14 unrestricted (power = 0.392) a C onstant variance m odel selected (p = 0 .2 2 6 2 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .2 7 .2 . Outputfor Selected Model: Hill 6 M arkowski et al., 2001: FR5 Run Opportunities 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 3 5 _ M a r k _ 2 0 0 1 _ F R 5 o p p _ H i l l C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 3 5 _ M a r k _ 2 0 0 1 _ F R 5 o p p _ H i l l C V _ 1 . p l t 13 M o n F e b 08 1 0 : 5 6 : 2 4 2 0 1 0 14 15 16 T a b l e 3 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 P o w e r p a r a m e t e r r e s t r i c t e d t o b e g r e a t e r t h a n 1 28 A c o n s t a n t v a r i a n c e m o d e l is f i t 29 This document is a draftfor review purposes only and does not constitute Agency policy. E-149 DRAFT--DO NOT CITE OR QUOTE 1 Total number of dose groups = 4 2 Total number of records with missing values = 0 3 Maximum number of iterations = 250 4 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 5 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 6 7 8 9 Default Initial Parameter Values 10 alpha 77.4849 11 rho 0 Specified 12 intercept 26.14 13 - 1 3 . 3 4 14 2 . 7 7 2 5 7 15 2 . 4 8 8 1 1 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 ( *** The model parameter(s) -rho -n 21 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 22 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 23 24 alpha intercept v k 25 26 a l p h a 1 -3.2e-009 1.9e-008 6.2e-008 27 28 i n t e r c e p t - 3 . 2 e - 0 0 9 1 -0.81 -0.51 29 30 1.9e-008 -0.81 1 0.36 31 32 k 6.2e-008 -0.51 0.36 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 39 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 40 alpha 64.5863 18.6445 28.0438 101.129 41 i n t e r c e p t 26.14 3.03753 20.1865 32.0935 42 -13.1569 3.7676 -20.5413 -5.77257 43 18 N A 44 1.68073 0.208677 1.27173 2.08973 45 46 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 47 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 48 h a s n o s t a n d a r d e r r o r . 49 50 51 52 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 53 54 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled R 55 56 57 07 26.1 26.1 12.3 8.04 -1.9e-008 58 1 . 5 5 7 4 23.5 23.5 7.04 8.04 -1.94e-007 59 4 . 0 3 6 12.8 13 6.17 8.04 -0.0558 60 1 0 . 3 2 7 13.1 13 7.14 8.04 0.0517 61 62 63 64 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 65 66 67 M o d e l A1: Yij = Mu(i) + e(ij) 68 V a r { e ( i j ) } = S i g m a ^ 2 69 70 M o d e l A2: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-150 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma(i)^2 2 3 Model A3: Yij = Mu(i) + e(ij) 4 Var{e(ij)} = Sigma^2 5 Model A3 uses any fixed variance parameters that 6 were specified by the user 7 8 M o d e l R: Yi = M u + e(i) 9 Var{e(i)} = Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) # Param's AIC 15 A1 -62.013133 5 134.026266 16 A2 -59.839035 8 135.678070 17 A3 -62.013133 5 134.026266 18 fitted -62.016025 4 132.032049 19 R -67.530040 2 139.060081 20 21 22 E x p l a n a t i o n o f T e s t s 23 24 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 25 (A2 vs. R) 26 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 27 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 28 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 29 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 30 31 T e s t s o f I n t e r e s t 32 33 Test -2*log(Likelihood Ratio) Test df p-val 34 35 T e s t 1 15.382 6 0.01748 36 T e s t 2 4.3482 3 0.2262 37 T e s t 3 4.3482 3 0.2262 38 T e s t 4 0.00578335 1 0.9394 39 40 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 41 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 42 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 43 44 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 45 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 46 47 48 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 49 t o b e a p p r o p r i a t e h e r e 50 51 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 52 t o a d e q u a t e l y d e s c r i b e t h e d a t a 53 54 55 B e n c h m a r k D o s e C o m p u t a t i o n 56 57 S p e c i f i e d e f f e c t = 1 58 59 R i s k T y p e = Estimated standard deviations from the control mean 60 61 C o n f i d e n c e l e v e l = 0.95 62 63 BMD = 1.72335 64 65 BMDL = 0.908491 66 This document is a draftfor review purposes only and does not constitute Agency policy. E-151 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E .2 .2 7 .3 . Figurefor Selected Model: Hill Hill Model with 0.95 Confidence Level 2 3 4 5 E .2.27.4. Outputfor Additional Model Presented: Power, Unrestricted 6 M arkowski et al., 2001: FR5 Run Opportunities 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 3 5 _ M a r k _ 2 0 0 1 _ F R 5 o p p _ P w r C V _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 3 5 _ M a r k _ 2 0 0 1 _ F R 5 o p p _ P w r C V _ U _ 1 . p l t 13 M o n F e b 08 1 0 : 5 6 : 2 4 2 0 1 0 14 15 16 T a b l e 3 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 T h e p o w e r is n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t This document is a draftfor review purposes only and does not constitute Agency policy. E-152 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 4 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha = 77.4849 12 rho = 0 Specified 13 control = 26.14 14 slope = -2.3827 15 power = 0.844532 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 21 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 22 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 23 24 alpha control slope power 25 26 a l p h a 1 -9.3e-009 1.4e-008 9.3e-009 27 CO o CO o 28 control -9.3e-009 1 -0.64 29 30 slope 1.4e-008 -0.64 1 0.9 31 32 power 9.3e-009 0.9 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 39 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 40 alpha 70.8926 20.4649 30.7821 111.003 41 control 26.3582 3.12902 20.2254 32.4909 42 slope -5.73309 4.02937 -13.6305 2.16433 43 power 0.391903 0.281862 -0.160536 0.944342 44 45 46 47 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 48 49 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 50 51 52 07 26.1 26.4 12.3 8.42 -0.0686 53 1 . 5 5 7 4 23.5 19.5 7.04 8.42 0.941 54 4 . 0 3 6 12.8 16.5 6.17 8.42 -1.06 55 1 0 . 3 2 7 13.1 12 7.14 8.42 0.343 56 57 58 59 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 60 61 62 M o d e l A1: Yij = Mu(i) + e(ij) 63 V a r j e i ;i j ) } = S i g m a ^ 2 64 65 M o d e l A2: Yij = Mu(i) + e(ij) 66 V a r j e i :i j ) } = S i g m a ( i ) ^2 67 68 M o d e l A3: Yij = Mu(i) + e(ij) 69 V a r j e i :i j ) } = S i g m a ^ 2 70 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t This document is a draftfor review purposes only and does not constitute Agency policy. E-153 DRAFT--DO NOT CITE OR QUOTE 1 were specified by the user 2 3 M o d e l R: Yi = M u + e(i) 4 Var{e(i)} = Sigma^2 5 6 7 Likelihoods of Interest 8 9 Model Log(likelihood) # Param's AIC 10 A1 -62.013133 5 134.026266 11 A2 -59.839035 8 135.678070 12 A3 -62.013133 5 134.026266 13 fitted -63.134001 4 134.268002 14 R -67.530040 2 139.060081 15 16 17 E x p l a n a t i o n o f T e s t s 18 19 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 20 (A2 vs. R) 21 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 22 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 23 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 24 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 25 26 T e s t s o f I n t e r e s t 27 28 Test -2*log(Likelihood Ratio) Test df p-val 29 30 T e s t 1 15.382 6 0.01748 31 T e s t 2 4.3482 3 0.2262 32 T e s t 3 4.3482 3 0.2262 33 T e s t 4 2.24174 1 0.1343 34 35 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 36 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 37 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 38 39 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 40 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 41 42 43 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 44 t o b e a p p r o p r i a t e h e r e 45 46 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 47 t o a d e q u a t e l y d e s c r i b e t h e d a t a 48 49 50 B e n c h m a r k D o s e C o m p u t a t i o n 51 52 S p e c i f i e d e f f e c t = 1 53 54 R i s k T y p e = Estimated standard deviations from the control mean 55 56 C o n f i d e n c e l e v e l = 0.95 57 58 B M D = 2 . 6 6 6 2 5 59 60 61 B M D L = 1 . 0 3 1 8 1 e - 0 1 4 62 63 This document is a draftfor review purposes only and does not constitute Agency policy. E-154 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E.2.27.5. F igure f o r A d d itio n a l M o d el Presented: Power, U nrestricted Power Model with 0.95 Confidence Level 2 10:56 02/08 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-155 DRAFT--DO NOT CITE OR QUOTE 1 E.2.28. Miettinen et al., 2006: Cariogenic Lesions, Pups 2 E.2.28.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2pValue AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 3 0.410 1 6 2.280 3 .4 0 1 E + 0 0 1 .889E + 00 p o w er b oun d hit (p ow er = 1) logistic 3 0.371 162.518 4.108E+00 2.450E+00 log-logistic a 3 0.602 161.292 1.428E+00 5.175E-01 slope bound hit (slope = 1) log-probit multistage, 4degree probit 3 0.300 1 6 3.040 6 .3 2 1 E + 0 0 3 .1 2 7 E + 0 0 slop e bound hit (slop e = 1) 3 0.410 162.280 3.401E+00 1.889E+00 final B = 0 3 0.350 162.656 4.548E+00 2.889E+00 W eibull 3 0.410 1 6 2.280 3 .4 0 1 E + 0 0 1 .889E + 00 p o w er b oun d hit (p ow er = 1) gamma, unrestricted log-logistic, unrestricted b log-probit, unrestricted W eibull, unrestricted 8.8 8 4 E - 2 0.798 161.801 3.374E-03 unrestricted (pow er = 0.215) 242 2 0.728 161.983 4.942E -02 error unrestricted (slope = 0.465) 2 0.732 161.972 6.495E -02 error unrestricted (slope = 0.289) 2 0.766 161.884 1.792E-02 error unrestricted (pow er = 0.324) aBest-fitting m odel, BM D S output presented in this appendix b Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .2 8 .2 . Outputfor Selected Model: Log-Logistic 6 M iettinen et al., 2006: C ariogenic L esions, Pups 7 8 9 10 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 3 6 _ M i e t _ 2 0 0 6 _ C a r i o g e n i c _ L o g L o g i s t i c _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 3 6 _ M i e t _ 2 0 0 6 _ C a r i o g e n i c _ L o g L o g i s t i c _ 1 . p l t 13 M o n F e b 08 1 0 : 5 6 : 5 9 2 0 1 0 14 15 16 T a b l e 2 c o n v e r t i n g t h e p e r c e n t a g e i n t o t h e n u m b e r o f a n i m a l s , a n d c o n t r o l is C o n t r o l II f r o m t h e 17 s t u d y . D o s e is i n n g p e r k g a n d is f r o m T a b l e 1 18 19 20 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 21 22 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 23 24 25 D e p e n d e n t v a r i a b l e = D i c h E f f This document is a draftfor review purposes only and does not constitute Agency policy. E -156 D RAFT-- DO N O T CITE OR QUOTE 1 Independent variable = Dose 2 Slope pa r a m e t e r is r estricted as slope >= 1 3 4 Total number of observations = 5 5 Total number of records with missing values = 0 6 Maximum number of iterations = 250 7 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 8 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 9 10 11 12 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 13 14 15 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 16 background = 0.595238 17 intercept = -2.494 18 slope = 1 19 20 21 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 22 23 ( * * * T h e m o d e l p a r a m e t e r ( s ) - s l o p e 24 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 25 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 26 27 background intercept 28 29 b a c k g r o u n d 1 -0.66 30 31 i n t e r c e p t -0.66 1 32 33 34 35 P a r a m e t e r E s t i m a t e s 36 37 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 38 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 39 b a c k g r o u n d 0.644165 * 40 intercept -2.55354 * 41 s l o p e 1* 42 43 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 44 45 46 47 A n a l y s i s o f D e v i a n c e T a b l e 48 49 Model Log(likelihood) # Param's Deviance Test d.f. P-value 50 Full model -77.6769 5 51 F i t t e d m o d e l -78.646 2 1.93832 3 0.5853 52 R e d u c e d m o d e l -83.2067 1 11.0597 4 0.0259 53 54 AIC: 161.292 55 56 57 G o o d n e s s o f F i t 58 S c a l e d 59 Dose Est. Prob. Expected Observed Size Residual 60 61 0 .0000 0. 6 4 4 2 27 055 25.000 42 - 0 . 6 6 2 62 2 .2195 0. 6 9 6 6 20. 200 23.000 29 1.131 63 6 .2259 0. 7 6 0 3 19. 007 19.000 25 -0.003 64 1 6 .0142 0. 8 4 1 6 20. 198 20.000 24 - 0 . 1 1 1 65 46 .6355 0. 9 2 3 1 29. 540 29.000 32 -0.358 66 67 C h i - 2 = 1 86 d.f. = 3 P value = 0.6024 68 69 70 B e n c h m a r k D o s e C o m p u t a t i o n This document is a draftfor review purposes only and does not constitute Agency policy. E-157 DRAFT--DO NOT CITE OR QUOTE 1 2 Specified effect 3 4 Risk Type 5 6 Confidence level 7 8 BMD 9 10 B M D L 11 12 0.1 Extra risk 0.95 1.42805 0.517495 13 E .2 .2 8 .3 . Figurefor Selected Model: Log-Logistic Log-Logistic Model with 0.95 Confidence Level 14 15 16 17 E .2.28.4. Outputfor Additional Model Presented: Log-Logistic, Unrestricted 18 M iettinen et al., 2006: C ariogen ic L esio n s, Pups 19 20 21 22 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 23 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 3 6 _ M i e t _ 2 0 0 6 _ C a r i o g e n i c _ L o g L o g i s t i c _ U _ 1 . ( d ) 24 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 3 6 _ M i e t _ 2 0 0 6 _ C a r i o g e n i c _ L o g L o g i s t i c _ U _ 1 . p l t 25 M o n F e b 08 1 0 : 5 6 : 5 9 2 0 1 0 26 27 This document is a draftfor review purposes only and does not constitute Agency policy. E-158 DRAFT--DO NOT CITE OR QUOTE 1 Table 2 c onverting the percen t a g e into the number of animals, and control is Control II from the 2 study. Dose is in ng per kg and is from Table 1 3 4 5 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 6 7 P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))] 8 9 10 D e p e n d e n t v a r i a b l e = D i c h E f f 11 I n d e p e n d e n t v a r i a b l e = D o s e 12 S l o p e p a r a m e t e r is n o t r e s t r i c t e d 13 14 T o t a l n u m b e r o f o b s e r v a t i o n s = 5 15 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 16 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 17 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 18 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 19 20 21 22 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 23 24 25 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 26 background = 0.595238 27 intercept = -0.739403 28 slope = 0.442847 29 30 31 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 32 33 background intercept slope 34 35 b a c k g r o u n d 1 -0.51 0.24 36 03 CO o 03 CO O 37 i n t e r c e p t -0.51 1 38 39 s l o p e 0.24 1 40 41 42 43 P a r a m e t e r E s t i m a t e s 44 45 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 46 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 47 b a c k g r o u n d 0.597745 48 intercept -0.798024 49 slope 0.465259 50 51 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 52 53 54 55 A n a l y s i s o f D e v i a n c e T a b l e 56 57 Model Log(likelihood) # Param's Deviance Test d.f. P-value 58 Full model -77.6769 5 59 F i t t e d m o d e l -77.9915 3 0.629204 2 0.7301 60 R e d u c e d m o d e l -83.2067 1 11.0597 4 0.0259 61 62 AIC: 161.983 63 64 65 G o o d n e s s o f F i t 66 S c a l e d 67 Dose Est. Prob. Expected Observed Size Residual 68 69 0.0000 0.5977 25.105 25.000 42 - 0 .033 70 2.2195 0.7566 21.940 23.000 29 0.458 This document is a draftfor review purposes only and does not constitute Agency policy. E-159 DRAFT--DO NOT CITE OR QUOTE 1 6.2259 0.8042 20.105 19.000 25 -0.557 2 16.0142 0.8474 20.338 20.000 24 - 0 . 1 9 2 3 46.6355 0.8910 28.512 29.000 32 0.277 4 5 Chi^2 = 0.63 d.f. = 2 P-value = 0.7281 6 7 8 Benchmark Dose Computation 9 10 S p e c i f i e d e f f e c t : 0.1 11 12 R i s k T y p e : Extra risk 13 14 C o n f i d e n c e l e v e l : 0.95 15 16 BMD ^ 0.049422 17 18 B e n c h m a r k d o s e c o m p u t a t i o n f a i l e d . L o w e r l i m i t i n c l u d e s zer o . 19 20 21 E .2.28.5. Figurefor Additional Model Presented: Log-Logistic, Unrestricted Log-Logistic Model t-- 1-- 1-- 1-- '-- '-- 1-- 1-- 1-- 1-- i-- '-- 1-- 1-- 1-- 1-- 1-- '-- '-- 1-- i-- 1-- 1-- '-- '-- 1-- 1-- 1-- 1-- 1-- i-- 1-- 1-- 1-- 1-- 1-- '-- '-- 1-- '-- r 1 Log-Logistic --------------- 0.9 0.8 T3 CD O -I--1 o (0 LL 0.6 0.5 0.4 J ____ i i i i i i i i i ____ I____ i i i i i i i i i ____ I____ i i i i i i i i i ____ I____ i i i i i i i i i ____ L 0 10 20 30 40 22 10:57 02/08 2010 23 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-160 DRAFT--DO NOT CITE OR QUOTE 1 E.2.29. M urray et al., 1979: Fertility in F2 Generation 2 E.2.29.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 0 N /A 61.729 4.481E+00 1.590E+00 logistic negative intercept (intercept = 1 0.051 61.318 2.420E+00 1.722E+00 -2.567) log-logistic 0 N /A 61.729 4.971E+00 1.565E+00 m ultistage, 1degree multistage, 2degree a probit 1 0.031 63.154 1.598E+00 8.747E-01 1 0.079 60.464 2.733E+00 1.366E+00 negative intercept (intercept = 1 0.048 61.544 2.250E+00 1.590E+00 -1.459) W eibull 0 N /A 61.729 5.042E+00 1.604E+00 log-probit, unrestricted 0 N /A 61.729 4.244E+00 1.506E+00 unrestricted (slope = 3.182) aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2.29.2. Outputfor Selected Model: Multistage, 2-Degree 6 Murray et al., 1979: Fertility in F2 Generation 7 8 9 10 M u l t i s t a g e M o d e l . ( V e r s i o n : 3.0; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ M u r r a y _ 1 9 7 9 _ f e r t _ i n d e x _ f 2 _ M u l t i 2 _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ M u r r a y _ 1 9 7 9 _ f e r t _ i n d e x _ f 2 _ M u l t i 2 _ 1 . p l t 13 W e d _ F e b 10 1 6 : 0 6 : 2 8 2 0 1 0 ~ 14 15 16 T a b l e 1 b u t e x p r e s s e d as n u m b e r o f d a m s w h o d o n o t p r o d u c e o f f s p r i n g 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) * [ 1 - E X P ( 22 - b e t a i * d o s e ^ 1 - b e t a 2 * d o s e ^ 2 ) ] 23 24 T h e p a r a m e t e r b e t a s a r e r e s t r i c t e d t o b e p o s i t i v e 25 26 27 D e p e n d e n t v a r i a b l e = D i c h E f f 28 I n d e p e n d e n t v a r i a b l e = D o s e 29 30 T o t a l n u m b e r o f o b s e r v a t i o n s = 3 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 T o t a l n u m b e r o f p a r a m e t e r s i n m o d e l = 3 33 T o t a l n u m b e r o f s p e c i f i e d p a r a m e t e r s = 0 34 D e g r e e o f p o l y n o m i a l = 2 35 This document is a draftfor review purposes only and does not constitute Agency policy. E-161 DRAFT--DO NOT CITE OR QUOTE 1 2 Maximum number of iterations = 250 3 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 4 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 5 6 7 8 Default Initial Parameter Values 9 Background = 0.0567204 10 Beta(1) = 0 11 Beta(2) = 0.0155037 12 13 14 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 15 16 ( * * * T h e m o d e l p a r a m e t e r ( s ) - B e t a ( 1 ) 17 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 18 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 19 20 Background Beta(2) 21 22 B a c k g r o u n d 1 -0.45 23 24 B e t a ( 2 ) -0.45 1 25 26 27 28 P a r a m e t e r E s t i m a t e s 29 30 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 31 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf 32 B a c k g r o u n d 0 . 0 7 8 0 1 8 8 * * * 33 B e t a ( 1 ) 0* * * 34 Beta(2) 0.0141051 35 36 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 37 38 39 40 A n a l y s i s o f D e v i a n c e T a b l e 41 42 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 43 Full model -25.8194 3 44 F i t t e d m o d e l -28.2318 2 4.82474 1 0.02805 45 R e d u c e d m o d e l -34.0009 1 16.363 2 0.0002798 46 47 AIC: 60.4636 48 49 50 G o o d n e s s o f F i t 51 S c a l e d 52 Dose Est. Prob. Expected Observed Size Residual 53 54 0.0000 0.0780 2.497 4.000 32 0.991 55 1.1242 0.0943 1.886 0.000 20 -1.443 56 5.8831 0.4341 8.683 9.000 20 0.143 57 58 C h i ^ 2 = 3 . 0 8 d.f. = 1 P-value = 0.0790 59 60 61 B e n c h m a r k D o s e C o m p u t a t i o n 62 63 S p e c i f i e d e f f e c t = 0.1 64 65 R i s k T y p e = Extra risk 66 67 C o n f i d e n c e l e v e l = 0.95 68 69 BMD = 2.73307 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-162 DRAFT--DO NOT CITE OR QUOTE 1 BMDL 1.36619 2 3 BMDU = 4.10938 4 5 T a k e n t ogether, (1.36619, 4.10938) is a 90 6 interval for the BMD 7 8 % two-sided confidence 9 E .2.29.3. Figurefor Selected Model: Multistage, 2-Degree Multistage Model with 0.95 Confidence Level Fraction Affected 10 11 This document is a draftfor review purposes only and does not constitute Agency policy. E-163 DRAFT--DO NOT CITE OR QUOTE 1 E.2.30. National Toxicology Program, 1982: Toxic Hepatitis, Male Mice 2 E.2.30.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2pValue AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 1 0.027 113.103 3.823E+00 2.005E+00 logistic negative intercept (intercept = 2 0.092 110.352 3.108E+00 2.465E+00 -3.388) log-logistic 1 0.026 113.089 3.797E+00 2.141E+00 log-probit multistage, 3degree a probit W eibull 1 0.027 113.111 3.565E+00 2.294E+00 1 0.036 112.045 2.782E+00 1.343E+00 negative intercept (intercept = 2 0.082 110.512 2.763E+00 2.241E+00 -1.894) 1 0.025 113.044 3.967E+00 1.704E+00 aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2.30.2. Outputfor Selected Model: Multistage, 3-Degree 6 National T oxicology Program, 1982: T oxic Hepatitis, M ale M ice 7 8 9 10 M u l t i s t a g e M o d e l . ( V e r s i o n : 3.0; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 3 7 _ N T P _ 1 9 8 2 _ T o x H e p _ M u l t i 3 _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 3 7 _ N T P _ 1 9 8 2 _ T o x H e p _ M u l t i 3 _ 1 . p l t 13 M o n F e b 08 1 0 : 5 7 : 3 2 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) * [ 1 - E X P ( 22 - b e t a i * d o s e ^ 1 - b e t a 2 * d o s e ^ 2 - b e t a 3 * d o s e ^ 3 ) ] 23 24 T h e p a r a m e t e r b e t a s a r e r e s t r i c t e d t o b e p o s i t i v e 25 26 27 D e p e n d e n t v a r i a b l e = D i c h E f f 28 I n d e p e n d e n t v a r i a b l e = D o s e 29 30 T o t a l n u m b e r o f o b s e r v a t i o n s = 4 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 T o t a l n u m b e r o f p a r a m e t e r s i n m o d e l = 4 33 T o t a l n u m b e r o f s p e c i f i e d p a r a m e t e r s = 0 34 D e g r e e o f p o l y n o m i a l = 3 35 36 37 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 38 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 This document is a draftfor review purposes only and does not constitute Agency policy. E-164 DRAFT--DO NOT CITE OR QUOTE 1 Parameter Convergence has been set to: 1e-008 2 3 4 5 Default Initial Parameter Values 6 Background = 0.0471757 7 Beta(1) = 0.00749116 8 Beta(2) = 0 9 Beta(3) = 0.00139828 10 11 12 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 13 14 ( * * * T h e m o d e l p a r a m e t e r ( s ) - B e t a ( 2 ) 15 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 16 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 17 18 Background Beta(1) Beta(3) 19 20 B a c k g r o u n d 1 -0.77 0.69 21 22 B e t a ( 1 ) -0.77 1 -0.95 23 24 B e t a ( 3 ) 0.69 -0.95 1 25 26 27 28 P a r a m e t e r E s t i m a t e s 29 30 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 31 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 32 B a c k g r o u n d 0 . 0 2 6 7 9 3 3 * * * 33 Beta(1) 0.0283198 * * * 34 B e t a ( 2 ) 0* * * 35 Beta(3) 0.0012342 * * * 36 37 I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 38 39 40 41 A n a l y s i s o f D e v i a n c e T a b l e 42 43 Model Log(likelihood) # Param's Deviance Test d.f. P-value 44 Full model -51.0633 4 45 F i t t e d m o d e l -53.0224 3 3.91812 1 0.04777 46 R e d u c e d m o d e l -121.743 1 141.358 3 <.0001 47 48 AIC: 112.045 49 50 51 G o o d n e s s o f F i t 52 S c a l e d 53 Dose Est. Prob. Expected Observed Size Residual 54 55 0.0000 0.0268 1.956 1.000 73 -0.693 56 0.7665 0.0482 2.363 5.000 49 1.759 57 2.2711 0.1005 4.925 3.000 49 - 0 .915 58 11.2437 0.8775 43.877 44.000 50 0.053 59 60 C h i ^ 2 = 4 . 4 1 d.f. = 1 P-value = 0.0357 61 62 63 B e n c h m a r k D o s e C o m p u t a t i o n 64 65 S p e c i f i e d e f f e c t = 0.1 66 67 R i s k T y p e = Extra risk 68 69 C o n f i d e n c e l e v e l = 0.95 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-165 DRAFT--DO NOT CITE OR QUOTE 1 BMD 2.78201 2 3 BMDL = 1.34308 4 5 BMDU = 4.5214 6 7 T a k e n t ogether, (1.34308, 4.5214 ) is a 90 8 interval for the BMD 9 10 % two-sided confidence 11 E .2 .3 0 .3 . Figurefor Selected Model: Multistage, 3-Degree Multistage Model with 0.95 Confidence Level Fraction Affected 12 10:57 02/08 2010 13 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-166 DRAFT--DO NOT CITE OR QUOTE 1 E.2.31. National Toxicology Program, 2006: Alveolar Metaplasia 2 E.2.31.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2pValue AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 4 0 .0 1 0 3 20.093 9 .886E -01 8 .393E -01 p o w er b oun d hit (p o w er = 1) logistic negative intercept (intercept = 4 <0.001 343.283 2.389E+00 2.052E+00 -1.059) log-logistic a 3 0.723 312.558 6.497E-01 3.751E-01 log-probit 4 0 .0 2 4 3 1 8 .6 8 0 1 .566E + 00 1 .318E + 00 slop e b ound hit (slop e = 1) multistage, 5degree probit 4 0.010 320.093 9.886E-01 8.393E-01 final b = 0 negative intercept (intercept = 4 <0.001 347.071 2.542E+00 2.219E+00 -0.599) W eibull 4 0 .0 1 0 3 20.093 9 .886E -01 8 .393E -01 p o w er b oun d hit (p o w er = 1) gamma, unrestricted log-probit, unrestricted W eibull, unrestricted 3 0.426 314.011 1.642E-01 1.874E-02 unrestricted (power = 0.503) 3 0.696 312.677 6.818E-01 2.740E-01 unrestricted (slope = 0.677) 3 0.522 313.492 2.644E-01 6.947E -02 unrestricted (power = 0.661) aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2 .3 1 .2 . Outputfor Selected Model: Log-Logistic 6 National T oxicology Program, 2006: A lveolar M etaplasia 7 8 9 10 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 4 0 _ N T P _ 2 0 0 6 _ A l v M e t a _ L o g L o g i s t i c _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 4 0 _ N T P _ 2 0 0 6 _ A l v M e t a _ L o g L o g i s t i c _ 1 . p l t 13 M o n F e b 08 1 0 : 5 8 : 5 8 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 S l o p e p a r a m e t e r is r e s t r i c t e d as s l o p e > = 1 27 28 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 29 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 30 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 This document is a draftfor review purposes only and does not constitute Agency policy. E-167 DRAFT--DO NOT CITE OR QUOTE 1 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 2 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 3 4 5 6 User has chosen the log transformed model 7 8 9 Default Initial Parameter Values 10 background = 0.0377358 11 intercept = -1.69494 12 slope = 1.12282 13 14 15 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 16 17 background intercept slope 18 19 ) a c k g r o u n d 1 -0.21 0.1 20 CO g o CO g o 21 i n t e r c e p t -0.21 1 22 23 s l o p e 0.1 1 24 25 26 27 P a r a m e t e r E s t i m a t e s 28 29 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 30 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 31 background 0.0373462 32 intercept -1.70923 33 slope 1.13164 34 35 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 36 37 38 39 A n a l y s i s o f D e v i a n c e T a b l e 40 41 Model Log(likelihood) # Param's Deviance Test d.f. P-value 42 Full model -152.615 6 43 F i t t e d m o d e l -153.279 3 1.32728 3 0.7227 44 R e d u c e d m o d e l -216.802 1 128.374 5 <.0001 45 46 AIC: 312.558 47 48 49 G o o d n e s s o f F i t 50 S c a l e d 51 Dose Est. Prob. Expected Observed Size Residua 52 53 0.0000 0.0373 1.979 2.000 53 0. 0 1 5 54 2.5565 0.3682 19.881 19.000 54 -0. 249 55 5.6937 0.5807 30.776 33.000 53 0. 6 1 9 56 9.7882 0.7162 37.243 35.000 52 -0. 690 57 16.5688 0.8197 43.446 45.000 53 0. 5 5 5 58 29.6953 0.8976 46.674 46.000 52 -0. 308 59 60 C h i ^ 2 = 1 . 3 3 d.f. = 3 P-value = 0.7232 61 62 63 B e n c h m a r k D o s e C o m p u t a t i o n 64 65 S p e c i f i e d e f f e c t = 0.1 66 67 R i s k T y p e = Extra risk 68 69 C o n f i d e n c e l e v e l = 0.95 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-168 DRAFT--DO NOT CITE OR QUOTE 1 BMD 0.64971 2 3 BMDL = 0.375051 4 5 6 E .2 .3 1 .3 . Figurefor Selected Model: Log-Logistic Log-Logistic Model with 0.95 Confidence Level Fraction Affected 7 10:58 02/08 2010 8 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-169 DRAFT--DO NOT CITE OR QUOTE 1 E.2.32. National Toxicology Program, 2006: Eosinophilic Focus, Liver 2 E.2.32.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 3 0.293 331.902 3.573E+00 2.225E+00 logistic negative intercept (intercept = 4 0.405 330.400 5.949E+00 5.137E+00 -2.043) log-logistic 3 0.152 333.515 4.139E+00 2.077E+00 log-probit slop e b ound hit (slop e = 1) 4 0.192 332.312 4.889E+00 3.980E+00 multistage, 5degree probit a 3 0.752 329.328 3.393E+00 2.466E+00 4 0.459 329.945 5.583E+00 4.864E+00 negative intercept (intercept = -1.235) W eibull 3 0.324 331.628 3.770E+00 2.249E+00 log-probit, unrestricted unrestricted (slope = 0.895) 3 0.116 334.150 4.146E+00 2.152E+00 aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2 .3 2 .2 . Outputfor Selected Model: Probit 6 National T oxicology Program, 2006: E osinophilic Focus, Liver 7 8 9 10 P r o b i t M o d e l . ( V e r s i o n : 3.1; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 4 5 _ N T P _ 2 0 0 6 _ L i v E o s F o c _ P r o b i t _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 4 5 _ N T P _ 2 0 0 6 _ L i v E o s F o c _ P r o b i t _ 1 . p l t 13 M o n F e b 08 11: 00: 54 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = C u m N o r m ( I n t e r c e p t + S l o p e * D o s e ) , 22 23 w h e r e C u m N o r m ( . ) is t h e c u m u l a t i v e n o r m a l d i s t r i b u t i o n f u n c t i o n 24 25 26 D e p e n d e n t v a r i a b l e = D i c h E f f 27 I n d e p e n d e n t v a r i a b l e = D o s e 28 S l o p e p a r a m e t e r is n o t r e s t r i c t e d 29 30 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 33 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 34 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 35 This document is a draftfor review purposes only and does not constitute Agency policy. E-170 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Default Initial (and Specified) Parameter Values 4 background = 0 Specified 5 intercept = -1.28017 6 slope = 0.0712441 7 8 9 Asymptotic Correlation Matrix of Parameter Estimates 10 11 ( * * * T h e m o d e l p a r a m e t e r ( s ) - b a c k g r o u n d 12 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 13 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 14 15 intercept slope 16 17 i n t e r c e p t 1 -0.77 18 19 slope -0.77 1 20 21 22 23 P a r a m e t e r E s t i m a t e s 24 25 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 26 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 27 intercept -1.23453 0.125132 -1.47979 -0.989279 28 slope 0.0688678 0.00823346 0.0527305 0.085005 29 30 31 32 A n a l y s i s o f D e v i a n c e T a b l e 33 34 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 35 F u l l m o d e l -161.07 6 36 F i t t e d m o d e l -162.972 2 3.80461 4 0.4331 37 R e d u c e d m o d e l -202.816 1 83.4925 5 <.0001 38 39 AIC: 329.945 40 41 42 G o o d n e s s o f F i t 43 44 Dose Est. Prob. Expected Observed Size Residual 1 CO 1o 45 46 0. 0 0 0 0 0. 1 5 5. 7 5 1 3. 0 0 0 53 -1 215 47 2. 5 5 6 5 0. 1 44 9 7. 8 2 6 8. 0 0 0 54 0. 067 48 5. 6 9 3 7 0. 1 99 8 10. 588 14 000 53 1. 1 7 2 49 9. 7 8 8 2 0. 2 87 6 15 242 17 000 53 0. 5 3 3 50 16. 5688 0. 4 62 8 24 526 22 000 53 -0 696 51 29. 6953 0. 7 91 2 41 932 42 000 53 0. 0 2 3 52 53 C h i ^ 2 = 3 . 6 2 d.f. = 4 P-value = 0.4593 54 55 56 B e n c h m a r k D o s e C o m p u t a t i o n 57 58 S p e c i f i e d e f f e c t : 0.1 59 60 R i s k T y p e Extra risk 61 62 C o n f i d e n c e l e v e l 0.95 63 64 BMD 5.58309 65 66 BMDL = 4.86394 67 This document is a draftfor review purposes only and does not constitute Agency policy. E-171 DRAFT--DO NOT CITE OR QUOTE 1 E.2.32.3. F igure f o r Selected M odel: Probit Probit Model with 0.95 Confidence Level Fraction Affected 2 11:00 02/08 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-172 DRAFT--DO NOT CITE OR QUOTE 1 E.2.33. National Toxicology Program, 2006: Fatty Change Diffuse, Liver 2 E.2.33.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 4 0.659 252.348 4.028E+00 2.923E+00 logistic negative intercept (intercept = 4 0.056 262.132 5.890E+00 5.042E+00 -2.825) log-logistic 4 0.359 254.413 4.254E+00 3.228E+00 log-probit multistage, 5degree probit Weibull a 4 0.367 254.428 4.204E+00 3.277E+00 3 0.581 254.045 3.524E+00 2.234E+00 negative intercept (intercept = 4 0.075 260.915 5.567E+00 4.784E+00 -1.665) 4 0.724 251.989 3.917E+00 2.856E+00 aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2 .3 3 .2 . Outputfor Selected Model: Weibull 6 National T oxicology Program, 2006: Fatty Change D iffuse, Liver 7 8 9 10 W e i b u l l M o d e l u s i n g W e i b u l l M o d e l ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 4 7 _ N T P _ 2 0 0 6 _ L i v F a t D i f f _ W e i b u l l _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 4 7 _ N T P _ 2 0 0 6 _ L i v F a t D i f f _ W e i b u l l _ 1 . p l t 13 M o n F e b 08 11: 01: 56 2 0 1 0 14 15 16 N T P l i v e r f a t t y c h a n g e d i f f u s e 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) * [ 1 - E X P ( - s l o p e * d o s e ^ p o w e r ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 P o w e r p a r a m e t e r is r e s t r i c t e d as p o w e r > = 1 27 28 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 29 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 30 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 31 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 32 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 33 34 35 36 D e f a u l t I n i t i a l (and S p e c i f i e d ) P a r a m e t e r V a l u e s 37 B a c k g r o u n d = 0 . 0 0 9 2 5 9 2 6 38 S l o p e = 0 . 0 0 7 2 1 3 5 5 This document is a draftfor review purposes only and does not constitute Agency policy. E-173 DRAFT--DO NOT CITE OR QUOTE 1 Power 1.69678 2 3 4 Asymptotic Correlation Matrix of Parameter Estimates 5 6 ( *** The model parameter(s) -Background 7 have been estimated at a boundary point, or have been specified by the user, 8 and do not appear in the correlation matrix ) 9 10 Slope Power 11 12 S l o p e 1 -0.98 13 14 Power -0.98 1 15 16 17 18 P a r a m e t e r E s t i m a t e s 19 20 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 21 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 22 B a c k g r o u n d 0 NA 23 Slope 0.0135075 0.00640459 0.00095478 0.0260603 24 Power 1.50444 0.168981 1.17324 1.83564 25 26 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 27 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 28 h a s n o s t a n d a r d e r r o r . 29 30 31 32 A n a l y s i s o f D e v i a n c e T a b l e 33 34 Model Log(likelihood) # Param's Deviance Test d.f. P-value 35 Full model -122.992 6 36 F i t t e d m o d e l -123.995 2 2.00444 4 0.7349 37 R e d u c e d m o d e l -204.846 1 163.708 5 <.0001 38 39 AIC: 251.989 40 41 42 G o o d n e s s o f F i t 43 S c a l e d 44 Dose Est. Prob. Expected Observed Size Residu 45 46 0. 0 0 0 0 0. 0 0 0 0 0. 0 0 0 0.000 53 0.000 47 2. 5 5 6 5 0. 0 5 3 9 2. 912 2.000 54 - 0 . 5 5 0 48 5. 6 9 3 7 0. 1 6 8 8 8. 949 12.000 53 1.119 49 9. 7 8 8 2 0. 3 4 1 5 18 102 17.000 53 -0.319 50 16. 5688 0. 6 0 2 4 31 929 30.000 53 -0.542 51 29. 6953 0. 8 9 1 3 47 2 3 8 48.000 53 0.336 52 53 C h i ^ 2 = 2 . 0 6 d.f. = 4 P-value = 0.7243 54 55 56 B e n c h m a r k D o s e C o m p u t a t i o n 57 58 S p e c i f i e d e f f e c t : 0.1 59 60 R i s k T y p e : Extra risk 61 62 C o n f i d e n c e l e v e l 0.95 63 64 BMD = 3.91723 65 66 BMDL = 2.85566 67 This document is a draftfor review purposes only and does not constitute Agency policy. E-174 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 E.2.33.3. F igure f o r Selected M odel: W eibull Weibull Model with 0.95 Confidence Level 2 11:01 02/08 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-175 DRAFT--DO NOT CITE OR QUOTE 1 E.2.34. National Toxicology Program, 2006: Gingival Hyperplasia, Squamous, 2 Years 2 E.2.34.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 4 0 .0 3 6 3 1 4 .9 8 5 7 .7 4 3 E + 0 0 5 .1 6 6 E + 0 0 p o w er b oun d hit (p ow er = 1) logistic negative intercept (intercept = 4 0.016 318.602 1.392E+01 1.056E+01 -1.859) log-logistic a 4 0.055 313.351 5.850E+00 3.730E+00 slope bound hit (slope = 1) log-probit 4 0.0 0 5 3 2 1 .4 2 6 1.535E +01 1.038E +01 slop e b ound hit (slop e = 1) multistage, 5degree probit 4 0.036 314.985 7.743E+00 5.166E+00 final B = 0 negative intercept (intercept = 4 0.018 318.240 1.318E+01 9.924E+00 -1.123) W eibull 4 0 .0 3 6 3 1 4 .9 8 5 7 .7 4 3 E + 0 0 5 .1 6 6 E + 0 0 p o w er b oun d hit (p ow er = 1) gamma, unrestricted log-logistic, unrestricted b log-probit, unrestricted W eibull, unrestricted 3 0.633 307.618 5.309E-01 9.859E -07 unrestricted (pow er = 0.282) 3 0.655 307.507 7.049E-01 1.260E-05 unrestricted (slope = 0.374) 3 0.668 307.444 8.357E-01 4.796E-05 unrestricted (slope = 0.22) 3 0.644 307.562 6.143E-01 3.872E -06 unrestricted (pow er = 0.325) aBest-fitting m odel, BM D S output presented in this appendix b Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .3 4 .2 . Outputfor Selected Model: Log-Logistic 6 National T oxicology Program, 2006: Gingival Hyperplasia, Squamous, 2 Years 7 8 9 10 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 4 2 _ N T P _ 2 0 0 6 _ G i n g H y p S q _ L o g L o g i s t i c _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 4 2 _ N T P _ 2 0 0 6 _ G i n g H y p S q _ L o g L o g i s t i c _ 1 . p l t 13 M o n F e b 08 1 0 : 5 9 : 5 7 2 0 1 0 14 15 16 [ i n s e r t s t u d y n o t e s ] 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e This document is a draftfor review purposes only and does not constitute Agency policy. E -176 D RAFT-- DO N O T CITE OR QUOTE 1 Slope parameter is restricted as slope >= 1 2 3 Total number of observations = 6 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 12 13 14 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 15 background = 0.0188679 16 intercept = -3.75308 17 slope = 1 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( * * * T h e m o d e l p a r a m e t e r ( s ) - s l o p e 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 24 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 25 26 background intercept 27 28 b a c k g r o u n d 1 -0.79 29 30 i n t e r c e p t -0.79 1 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 37 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf 38 b a c k g r o u n d 0 . 0 6 7 1 8 1 2 * * * 39 intercept -3.96371 * * * 40 s l o p e 1 41 42 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 43 44 45 46 A n a l y s i s o f D e v i a n c e T a b l e 47 48 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 49 F u l l m o d e l -149.95 6 50 F i t t e d m o d e l -154.675 2 9.45085 4 0.05077 51 R e d u c e d m o d e l -162.631 1 25.3627 5 0.0001186 52 53 AIC: 313.351 54 55 56 G o o d n e s s o f F i t 57 58 Dose Est. Prob. Expected Observed Size Residual 59 60 0.0000 0.0672 3.561 1. 0 0 0 53 -1.405 61 2.5565 0.1104 5.960 7. 0 0 0 54 0.452 62 5.6937 0.1582 8.385 14. 000 53 2.113 63 9.7882 0.2134 11.311 13. 000 53 0.566 64 16.5688 0.2905 15.394 15. 000 53 -0.119 65 29.6953 0.4036 21.389 16. 000 53 -1.509 66 67 C h i ^ 2 = 9 . 2 6 d.f. = 4 P-value = 0.0550 68 69 70 B e n c h m a r k D o s e C o m p u t a t i o n This document is a draftfor review purposes only and does not constitute Agency policy. E-177 DRAFT--DO NOT CITE OR QUOTE 1 2 Specified effect 3 4 Risk Type 5 6 Confidence level 7 8 BMD 9 10 B M D L 11 12 0.1 Extra risk 0.95 5.85026 3.7296 13 E .2 .3 4 .3 . Figurefor Selected Model: Log-Logistic Log-Logistic Model with 0.95 Confidence Level 14 15 16 17 E .2.34.4. Outputfor Additional Model Presented: Log-Logistic, Unrestricted 18 N ation al T o x ic o lo g y Program , 2006: G ingival H yperplasia, Squam ous, 2 Y ears 19 20 21 22 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 23 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 4 2 _ N T P _ 2 0 0 6 _ G i n g H y p S q _ L o g L o g i s t i c _ U _ 1 . ( d ) 24 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 4 2 _ N T P _ 2 0 0 6 _ G i n g H y p S q _ L o g L o g i s t i c _ U _ 1 . p l t 25 M o n F e b 08 1 0 : 5 9 : 5 7 2 0 1 0 26 27 28 [ i n s e r t s t u d y n o t e s ] This document is a draftfor review purposes only and does not constitute Agency policy. E-178 DRAFT--DO NOT CITE OR QUOTE 1 2 3 The form of the probability function is: 4 5 P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))] 6 7 8 Dependent variable = DichEff 9 Independent variable = Dose 10 S l o p e p a r a m e t e r is n o t r e s t r i c t e d 11 12 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 13 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 14 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 15 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 16 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 17 18 19 20 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 21 22 23 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 24 background = 0.0188679 25 intercept = -2.2 26 slope = 0.424326 27 28 29 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 30 31 background intercept slope 32 33 b a c k g r o u n d 1 -0.27 0.11 34 CO g o CO g o 35 i n t e r c e p t -0.27 1 36 37 s l o p e 0.11 1 38 39 40 41 P a r a m e t e r E s t i m a t e s 42 43 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 44 Variable Estimate Std. Err. Lower Conf. Limit Uppe 45 b a c k g r o u n d 0 . 0 1 8 5 1 3 8 * * 46 intercept -2.06653 * * 47 slope 0.373721 48 49 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 50 51 52 53 A n a l y s i s o f D e v i a n c e T a b l e 54 55 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 56 F u l l m o d e l -149.95 6 57 F i t t e d m o d e l -150.753 3 1.60697 3 0.6578 58 R e d u c e d m o d e l -162.631 1 25.3627 5 0.0001186 59 60 AIC: 307.507 61 62 63 G o o d n e s s o f F i t 64 65 Dose Est. Prob. Expected Observed Size Residual 66 67 0.0000 0.0185 0.981 1.000 53 0.019 68 2.5565 0.1681 9.078 7.000 54 - 0 . 7 5 6 69 5.6937 0.2101 11.136 14.000 53 0.966 70 9.7882 0.2433 12.893 13.000 53 0.034 This document is a draftfor review purposes only and does not constitute Agency policy. E-179 DRAFT--DO NOT CITE OR QUOTE 1 16.5688 0.2792 14.795 15.000 2 29.6953 0.3230 17.117 16.000 3 4 Chi^2 = 1.62 d.f. = 3 P-value = 0.6554 5 6 7 Benchmark Dose Computation 8 9 Specified effect : 0.1 10 11 R i s k T y p e : Extra risk 12 13 C o n f i d e n c e l e v e l : 0.95 14 15 BMD = 0.704898 16 17 BMDL = 1.26034e-005 18 19 53 53 0.063 -0.328 20 E .2.34.5. Figurefor Additional Model Presented: Log-Logistic, Unrestricted Log-Logistic Model with 0.95 Confidence Level 21 10:59 02/08 2010 22 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-180 DRAFT--DO NOT CITE OR QUOTE 1 E.2.35. National Toxicology Program, 2006: Hepatocyte Hypertrophy, 2 Years 2 E.2.35.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 5 0 .0 3 4 2 7 3 .8 7 5 9 .091E -01 7.8 6 8 E -0 1 p o w er b oun d hit (p o w er = 1) logistic negative intercept (intercept = 4 <0.001 297.895 2.475E+00 2.122E+00 -1.685) log-logistic 4 0.006 279.210 1.137E+00 6.491E-01 log-probit 5 0.006 277.800 1.530E+00 1.321E+00 multistage, 5degree a probit 4 0.018 275.693 9.272E-01 7.906E-01 negative intercept (intercept = 4 <0.001 299.731 2.453E+00 2.137E+00 -0.985) W eibull 5 0 .0 3 4 2 7 3 .8 7 5 9 .091E -01 7.8 6 8 E -0 1 p o w er b oun d hit (p o w er = 1) gamma, unrestricted log-probit, unrestricted W eibull, unrestricted 4 0.027 275.270 error error unrestricted (power = 0.844) 4 0.008 278.360 1.191E+00 7.038E-01 unrestricted (slope = 0.864) 4 0.024 275.439 7.345E-01 3.588E-01 unrestricted (power = 0.92) aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2.35.2. Outputfor Selected Model: Multistage, 5-Degree 6 National T oxicology Program, 2006: Hepatocyte Hypertrophy, 2 Years 7 8 9 10 M u l t i s t a g e M o d e l . ( V e r s i o n : 3.0; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 4 3 _ N T P _ 2 0 0 6 _ H e p H y p e r _ M u l t i 5 _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 4 3 _ N T P _ 2 0 0 6 _ H e p H y p e r _ M u l t i 5 _ 1 . p l t 13 M o n F e b 08 1 1 : 0 0 : 2 5 2 0 1 0 14 15 16 [ i n s e r t s t u d y n o t e s ] 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) * [ 1 - E X P ( 22 - b e t a i * d o s e ^ 1 - b e t a 2 * d o s e ^ 2 - b e t a 3 * d o s e ^ 3 - b e t a 4 * d o s e ^ 4 - b e t a 5 * d o s e ^ 5 ) ] 23 24 T h e p a r a m e t e r b e t a s a r e r e s t r i c t e d t o b e p o s i t i v e 25 26 27 D e p e n d e n t v a r i a b l e = D i c h E f f 28 I n d e p e n d e n t v a r i a b l e = D o s e 29 30 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 This document is a draftfor review purposes only and does not constitute Agency policy. E-181 D RAFT-- DO N O T CITE OR QUOTE 1 Total number of records with missing values = 0 2 Total number of parameters in model = 6 3 Total number of specified parameters = 0 4 Degree of polynomial = 5 5 6 7 Maximum number of = 250 8 Relative Function has b e e n set to: 1e-008 9 Parameter has b e e n set to: 1e-008 10 11 12 13 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 14 Background 0.112745 15 Beta(1) 0.0950808 16 Beta(2) 0 17 Beta(3) 0 18 Beta(4) 0 19 Beta(5) 4.39515e-008 20 21 22 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 23 24 ( *** The model parameter(s) -Background -Beta(2) -Beta(3) -Beta(4) 25 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 26 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 27 28 Beta(1) Beta(5) 29 30 B e t a ( 1 ) 1 -0.5 31 32 B e t a ( 5 ) -0.5 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 39 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 40 B a c k g r o u n d 0* * * 41 Beta(1) 0.113632 * * * 42 B e t a ( 2 ) 0* * * 43 B e t a ( 3 ) 0* * * 44 B e t a ( 4 ) 0* * * 45 Beta(5) 1. 1 3 2 2 e - 0 0 8 * * * 46 47 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 48 49 50 51 A n a l y s i s o f D e v i a n c e T a b l e 52 53 Model Log(likelihood) # Param's Deviance Test d.f. P-value 54 Full model -129.986 6 55 F i t t e d m o d e l -135.847 2 11.7216 4 0.01955 56 R e d u c e d m o d e l -219.97 1 179.968 5 <.0001 57 58 AIC: 275.693 59 60 61 G o o d n e s s o f F i t 62 S c a l e d 63 Dose Est. Prob. Expected Observed Size Residu 64 65 0.0000 0.0000 0.000 0.000 53 0. 0 0 0 66 2.5565 0.2521 13.614 19.000 54 1. 6 88 67 5.6937 0.4764 25.251 19.000 53 -1. 719 68 9.7882 0.6717 35.599 42.000 53 1. 8 72 69 16.5688 0.8510 45.106 41.000 53 -1. 584 70 29.6953 0.9769 51.778 52.000 53 0. 2 0 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-182 DRAFT--DO NOT CITE OR QUOTE 1 2 Chi^2 = 11.86 d.f. = 4 P-value = 0.0184 3 4 5 Benchmark Dose Computation 6 7 Specified effect 0.1 8 9 Risk Type Extra risk 10 11 C o n f i d e n c e l e v e l 0.95 12 13 BMD 0.92721 14 15 BMDL 0.790637 16 17 BMDU 1.14523 18 19 T a k e n t o g e t h e r , ( 0 . 7 9 0 6 3 7 , 1 . 1 4 5 2 3 ) is a 90 two-sided confidence 20 i n t e r v a l f o r t h e B M D 21 22 23 E .2 .3 5 .3 . Figurefor Selected Model: Multistage, 5-Degree Multistage Model with 0.95 Confidence Level 24 11:00 02/08 2010 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-183 DRAFT--DO NOT CITE OR QUOTE 1 E.2.36. National Toxicology Program, 2006: Necrosis, Liver 2 E.2.36.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 4 0 .9 3 9 2 3 4 .4 0 0 8 .6 5 5 E + 0 0 6 .3 4 0 E + 0 0 p o w er b oun d hit (p ow er = 1) logistic negative intercept (intercept 4 0.601 236.742 1.484E+01 1.240E+01 = -2.818) log-logistic 4 0.943 2 3 4 .3 8 2 7 .9 2 8 E + 0 0 5 .6 0 5 E + 0 0 slop e b ound hit (slop e = 1) log-probit 4 0 .5 7 2 2 3 6 .8 6 3 1.333E +01 1.024E +01 slop e b ound hit (slop e = 1) multistage, 5degree probit 4 0.939 234.400 8.655E+00 6.340E+00 final b = 0 negative intercept (intercept 4 0.666 236.293 1.393E+01 1.154E+01 = -1.626) W eibull 4 0 .9 3 9 2 3 4 .4 0 0 8 .6 5 5 E + 0 0 6 .3 4 0 E + 0 0 p o w er b oun d hit (p ow er = 1) gamma, unrestricted log-logistic, unrestricted log-probit, unrestricted a W eibull, unrestricted 3 0.883 236.290 7.726E +00 3.453E+00 unrestricted (power = 0.87) 3 0.860 236.377 7.733E +00 3.536E+00 unrestricted (slope = 0.974) 3 0.805 236.598 7.501E+00 3.504E+00 unrestricted (slope = 0.517) 3 0.879 236.302 7.763E +00 3.508E+00 unrestricted (power = 0.895) aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2.36.2. Outputfor Selected Model: Log-Probit, Unrestricted 6 National T oxicology Program, 2006: N ecrosis, Liver 7 8 9 10 P r o b i t M o d e l . ( V e r s i o n : 3.1; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 5 0 _ N T P _ 2 0 0 6 _ L i v N e c _ L o g P r o b i t _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 5 0 _ N T P _ 2 0 0 6 _ L i v N e c _ L o g P r o b i t _ U _ 1 . p l t 13 M o n F e b 08 1 1 : 2 9 : 3 0 2 0 1 0 14 15 16 N T P _ l i v e r _ n e c r o s i s 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = B a c k g r o u n d 22 + ( 1 - B a c k g r o u n d ) * C u m N o r m ( I n t e r c e p t + S l o p e * L o g ( D o s e ) ) , 23 24 w h e r e C u m N o r m ( . ) is t h e c u m u l a t i v e n o r m a l d i s t r i b u t i o n f u n c t i o n 25 26 27 Dependent variable = DichEff This document is a draftfor review purposes only and does not constitute Agency policy. E-184 DRAFT--DO NOT CITE OR QUOTE 1 Independent variable = Dose 2 Slope pa r a m e t e r is not restricted 3 4 Total number of observations = 6 5 Total number of records with missing values = 0 6 Maximum number of iterations = 250 7 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 8 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 9 10 11 12 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 13 14 15 D e f a u l t I n i t i a l (and S p e c i f i e d ) P a r a m e t e r V a l u e s 16 background = 0.0188679 17 intercept = -2.16223 18 slope = 0.457376 19 20 21 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 22 23 background intercept slope 24 Lf) o Lf) o 25 a c k g r o u n d 1 0.55 26 27 i n t e r c e p t 1 -0.97 28 29 s l o p e 0.55 -0.97 1 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 36 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 37 b a c k g r o u n d 0 . 0 2 2 1 1 5 1 0 . 0 2 2 1 3 5 1 -0.0212689 0.065499 38 intercept -2.32352 0.556343 -3.41393 -1.23311 39 slope 0.517104 0.185064 0.154385 0.879823 40 41 42 43 A n a l y s i s o f D e v i a n c e T a b l e 44 45 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 46 Full model -114.813 6 47 F i t t e d m o d e l -115.299 3 0.972184 3 0.808 48 R e d u c e d m o d e l -127.98 1 26.3331 5 <.0001 49 50 AIC: 236.598 51 52 53 G o o d n e s s o f F i t 54 S c a l e d 55 Dose Est . Prob. Expected Observed Size Residual 56 57 0 .0000 0. 0 2 2 1 1. 1 7 2 1.000 53 -0.161 58 2 .5565 0. 0 5 4 4 2. 938 4.000 54 0.637 59 5 .6937 0. 0 9 7 6 5. 174 4.000 53 -0.543 60 9 .7882 0. 1 4 5 7 7. 7 2 0 8.000 53 0.109 61 16 .5688 0. 2 0 9 6 11. 106 10.000 53 -0.373 62 29 .6953 0. 3 0 0 2 15. 908 17.000 53 0.327 63 64 ii^ 2 = 0. 99 d . f . = 3 P- value = 0 . 8 0 4 8 65 66 67 B e n c h m a r k D o s e C o m p u t a t i o n 68 69 S p e c i f i e d e f f e c t = 0.1 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-185 DRAFT--DO NOT CITE OR QUOTE 1 Risk Type 2 3 Confidence level 4 5 BMD 6 7 BMDL 8 9 Extra risk 0.95 7.50077 3.5039 10 E .2.36.3. Figurefor Selected Model: Log-Probit, Unrestricted LogProbit Model with 0.95 Confidence Level Fraction Affected 11 11:29 02/08 2010 12 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-186 DRAFT--DO NOT CITE OR QUOTE 1 E.2.37. National Toxicology Program, 2006: Oval Cell Hyperplasia 2 E.2.37.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 3 0.074 199.468 6.739E+00 5.074E+00 logistic negative intercept (intercept = 4 0.171 196.803 6.064E+00 5.145E+00 -3.834) log-logistic 3 0.042 201.659 6.936E+00 5.604E+00 log-probit multistage, 5degree probit a W eibull b 3 0.072 200.121 7.090E+00 5.931E+00 3 0.207 195.962 4.785E+00 3.105E+00 4 0.227 195.448 5.673E+00 4.793E+00 negative intercept (intercept = -2.19) 3 0.077 198.375 5.718E+00 4.088E+00 aBest-fitting m odel, BM D S output presented in this appendix b Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .3 7 .2 . Outputfor Selected Model: Probit 6 National T oxicology Program, 2006: Oval Cell Hyperplasia 7 8 9 10 P r o b i t M o d e l . ( V e r s i o n : 3.1; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 5 3 _ N T P _ 2 0 0 6 _ O v a l H y p e r _ P r o b i t _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 5 3 _ N T P _ 2 0 0 6 _ O v a l H y p e r _ P r o b i t _ 1 . p l t 13 M o n F e b 08 1 3 : 2 5 : 2 3 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = C u m N o r m ( I n t e r c e p t + S l o p e * D o s e ) , 22 23 w h e r e C u m N o r m ( . ) is t h e c u m u l a t i v e n o r m a l d i s t r i b u t i o n f u n c t i o n 24 25 26 D e p e n d e n t v a r i a b l e = D i c h E f f 27 I n d e p e n d e n t v a r i a b l e = D o s e 28 S l o p e p a r a m e t e r is n o t r e s t r i c t e d 29 30 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 33 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 34 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 35 36 37 This document is a draftfor review purposes only and does not constitute Agency policy. E-187 DRAFT--DO NOT CITE OR QUOTE 1 Default Initial (and Specified) Parameter Values 2 background = 0 Specified 3 intercept = -2.29925 4 slope = 0.169545 5 6 7 Asymptotic Correlation Matrix of Parameter Estimates 8 9 ( *** The model parameter(s) -background 10 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 11 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 12 13 intercept slope 14 15 i n t e r c e p t 1 -0.87 16 17 slope -0.87 1 18 19 20 21 P a r a m e t e r E s t i m a t e s 22 23 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 24 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 25 intercept -2.18988 0.208021 -2.5976 -1.78217 26 slope 0.172453 0.0182446 0.136694 0.208211 27 28 29 30 A n a l y s i s o f D e v i a n c e T a b l e 31 32 Model Log(likelihood) # Param's Deviance Test d.f. P-value 33 Full model -92.4898 6 34 F i t t e d m o d e l -95.7242 2 6.46873 4 0.1668 35 R e d u c e d m o d e l -210.191 1 235.402 5 <.0001 36 37 AIC: 195.448 38 39 40 G o o d n e s s o f F i t 41 S c a l e d 42 Dose Est. Prob. Expected Observed Size Residual 43 44 0.0000 0.0143 0.756 0.000 53 -0. 876 45 2.5565 0.0401 2.168 4.000 54 1. 2 7 0 46 5.6937 0.1135 6.017 3.000 53 -1. 306 47 9.7882 0.3079 16.317 20.000 53 1. 0 9 6 48 16.5688 0.7478 39.631 38.000 53 -0. 516 49 29.6953 0.9983 52.911 53.000 53 0. 2 9 9 50 51 C h i ^ 2 = 5 . 6 4 d.f. = 4 P -value = 0.2274 52 53 54 B e n c h m a r k D o s e C o m p u t a t i o n 55 56 S p e c i f i e d e f f e c t 0.1 57 58 R i s k T y p e Extra risk 59 60 C o n f i d e n c e l e v e l 0.95 61 62 BMD 5.67298 63 64 BMDL 4.79341 65 This document is a draftfor review purposes only and does not constitute Agency policy. E-188 DRAFT--DO NOT CITE OR QUOTE 1 E.2.37.3. F igure f o r Selected M odel: Probit Probit Model with 0.95 Confidence Level Fraction Affected 2 13:25 02/08 2010 3 dose 4 5 E .2.37.4. Outputfor Additional Model Presented: Weibull 6 National T oxicology Program, 2006: Oval Cell Hyperplasia 7 8 9 10 W e i b u l l M o d e l u s i n g W e i b u l l M o d e l ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 5 3 _ N T P _ 2 0 0 6 _ O v a l H y p e r _ W e i b u l l _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 5 3 _ N T P _ 2 0 0 6 _ O v a l H y p e r _ W e i b u l l _ 1 . p l t 13 M o n F e b 08 1 3 : 2 5 : 2 3 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) * [ 1 - E X P ( - s l o p e * d o s e ^ p o w e r ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 P o w e r p a r a m e t e r is r e s t r i c t e d as p o w e r > = 1 27 28 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 This document is a draftfor review purposes only and does not constitute Agency policy. E-189 DRAFT--DO NOT CITE OR QUOTE 1 Total number of records with missing values = 0 2 Maximum number of iterations = 250 3 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 4 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 5 6 7 8 Default Initial (and Specified) Parameter Values 9 Background = 0.00925926 10 Slope = 0.00296825 11 Power = 2.17092 12 13 14 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 15 16 Background Slope Power 17 18 B a c k g r o u n d 1 -0.72 0.7 19 20 Slope -0.72 1 -0.99 21 22 P o w e r 0.7 -0.99 1 23 24 25 26 P a r a m e t e r E s t i m a t e s 27 28 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 29 Variable Estimate Std. Err. Lower Conf. Limit Uppe Conf. Limit 30 B a c k g r o u n d 0 . 0 1 6 4 1 3 7 0 . 0 2 2 1 4 8 8 -0.0269971 0.0598245 31 Slope 0.00162074 0.00202897 -0.00235596 0.00559745 32 Power 2.39427 0.455116 1.50226 3.28628 33 34 35 36 A n a l y s i s o f D e v i a n c e T a b l e 37 38 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 39 Full model -92.4898 6 40 F i t t e d m o d e l -96.1875 3 7.3953 3 0.06031 41 R e d u c e d m o d e l -210.191 1 235.402 5 <.0001 42 43 AIC: 198.375 44 45 46 G o o d n e s s o f F i t 47 S c a l e d 48 Dose Est. Prob. Expected Observed Size Residual 49 50 0.0000 0.0164 0.870 0.000 53 -0. 940 51 2.5565 0.0314 1.695 4.000 54 1. 7 9 9 52 5.6937 0.1138 6.034 3.000 53 -1. 312 53 9.7882 0.3285 17.411 20.000 53 0. 7 57 54 16.5688 0.7440 39.431 38.000 53 -0. 450 55 29.6953 0.9957 52.774 53.000 53 0. 4 7 6 56 57 C h i ^ 2 = 6 85 d.f. = 3 P-value = 0.0770 58 59 60 B e n c h m a r k D o s e C o m p u t a t i o n 61 62 S p e c i f i e d e f f e c t 0.1 63 64 R i s k T y p e Extra risk 65 66 C o n f i d e n c e l e v e l 0.95 67 68 BMD 5.71754 69 70 BMDL 4.08823 This document is a draftfor review purposes only and does not constitute Agency policy. E-190 DRAFT--DO NOT CITE OR QUOTE 1 E.2.37.5. F igure f o r A d d itio n a l M o d el Presented: W eibull Weibull Model with 0.95 Confidence Level Fraction Affected 2 13:25 02/08 2010 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-191 DRAFT--DO NOT CITE OR QUOTE 1 E.2.38. National Toxicology Program, 2006: Pigmentation, Liver 2 E.2.38.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 3 0.552 196.971 2.172E+00 1.493E+00 logistic negative intercept (intercept = 4 0.247 197.066 1.853E+00 1.521E+00 -2.51) log-logistic 3 0.984 195.530 2.566E+00 1.937E+00 log-probit a multistage, 5degree probit W eibull 3 0.962 195.526 2.463E+00 1.890E+00 3 0.058 199.955 1.822E+00 9.916E-01 final b = 0 negative intercept (intercept = 4 0.004 200.504 1.710E+00 1.430E+00 -1.392) 3 0.219 199.007 1.756E+00 1.190E+00 aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2 .3 8 .2 . Outputfor Selected Model: Log-Probit 6 National T oxicology Program, 2006: Pigmentation, Liver 7 8 9 10 P r o b i t M o d e l . ( V e r s i o n : 3.1; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 5 4 _ N T P _ 2 0 0 6 _ P i g m e n t _ L o g P r o b i t _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 5 4 _ N T P _ 2 0 0 6 _ P i g m e n t _ L o g P r o b i t _ 1 . p l t 13 M o n F e b 08 13: 25: 55 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = B a c k g r o u n d 22 + ( 1 - B a c k g r o u n d ) * C u m N o r m ( I n t e r c e p t + S l o p e * L o g ( D o s e ) ) , 23 24 w h e r e C u m N o r m ( . ) is t h e c u m u l a t i v e n o r m a l d i s t r i b u t i o n f u n c t i o n 25 26 27 D e p e n d e n t v a r i a b l e = D i c h E f f 28 I n d e p e n d e n t v a r i a b l e = D o s e 29 S l o p e p a r a m e t e r is r e s t r i c t e d as s l o p e > = 1 30 31 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 32 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 33 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 34 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 35 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 36 37 38 This document is a draftfor review purposes only and does not constitute Agency policy. E-192 DRAFT--DO NOT CITE OR QUOTE 1 User has chosen the log transformed model 2 3 4 Default Initial (and Specified) Parameter Values 5 background = 0.0754717 6 intercept = -2.48683 7 slope = 1.53221 8 9 10 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 11 12 background intercept slope 13 14 b a c k g r o u n d 1 -0.42 0.33 15 16 i n t e r c e p t -0.42 1 -0.96 17 18 s l o p e 0.33 -0.96 1 19 20 21 22 P a r a m e t e r E s t i m a t e s 23 24 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 25 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 26 b a c k g r o u n d 0 . 0 7 2 5 4 7 3 0 . 0 3 3 8 8 5 6 0.00613263 0.138962 27 intercept -2.93268 0.487158 -3.8875 -1.97787 28 slope 1.83184 0.246868 1.34798 2.31569 29 30 31 32 A n a l y s i s o f D e v i a n c e T a b l e 33 34 Model Log(likelihood) # Param's Deviance Test d.f. P-value 35 Full model -94.6177 6 36 F i t t e d m o d e l -94.7632 3 0.291072 3 0.9617 37 R e d u c e d m o d e l -210.717 1 232.198 5 <.0001 38 39 AIC: 195.526 40 41 42 G o o d n e s s o f F i t 43 S c a l e d 44 Dose Est. Prob. Expected Observed Size Residual 45 46 0.0000 0.0725 3.845 4.000 53 0.082 47 2.5565 0.1769 9.553 9.000 54 -0. 1 9 7 48 5.6937 0.6291 33.342 34.000 53 0.187 49 9.7882 0.9013 47.771 48.000 53 0.105 50 16.5688 0.9874 52.334 52.000 53 -0.412 51 29.6953 0.9995 52.974 53.000 53 0.160 52 53 C h i ^ 2 = 0 . 2 9 d.f. = 3 P-value = 0.9624 54 55 56 B e n c h m a r k D o s e C o m p u t a t i o n 57 58 S p e c i f i e d e f f e c t 0.1 59 60 R i s k T y p e Extra risk 61 62 C o n f i d e n c e l e v e l 0.95 63 64 BMD 2.46293 65 66 BMDL 1.88981 67 This document is a draftfor review purposes only and does not constitute Agency policy. E-193 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 E.2.38.3. F igure f o r Selected M odel: Log-P robit LogProbit Model with 0.95 Confidence Level 2 13:25 02/08 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-194 DRAFT--DO NOT CITE OR QUOTE 1 E.2.39. National Toxicology Program, 2006: Toxic Hepatopathy 2 E.2.39.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2pValue AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 4 0.754 185.763 4.302E+00 3.463E+00 logistic negative intercept (intercept = 4 0.159 191.136 4.833E+00 4.068E+00 -3.756) log-logistic 3 0.391 189.577 4.697E+00 3.818E+00 log-probit multistage, 5degree a probit W eibull 3 0.394 189.580 4.972E+00 3.780E+00 4 0.693 185.924 3.980E+00 3.059E+00 final B = 0 negative intercept (intercept = 4 0.231 189.820 4.621E+00 3.860E+00 -2.172) 4 0.716 185.785 4.089E+00 3.215E+00 aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2.39.2. Outputfor Selected Model: Multistage, 5-Degree 6 National T oxicology Program, 2006: Toxic Hepatopathy 7 8 9 10 M u l t i s t a g e M o d e l . ( V e r s i o n : 3.0; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 5 5 _ N T P _ 2 0 0 6 _ T o x H e p a _ M u l t i 5 _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 5 5 _ N T P _ 2 0 0 6 _ T o x H e p a _ M u l t i 5 _ 1 . p l t 13 M o n _ F e b 08 1 3 : 2 6 : 2 8 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) * [ 1 - E X P ( 22 - b e t a i * d o s e ^ 1 - b e t a 2 * d o s e ^ 2 - b e t a 3 * d o s e ^ 3 - b e t a 4 * d o s e ^ 4 - b e t a 5 * d o s e ^ 5 ) ] 23 24 T h e p a r a m e t e r b e t a s a r e r e s t r i c t e d t o b e p o s i t i v e 25 26 27 D e p e n d e n t v a r i a b l e = D i c h E f f 28 I n d e p e n d e n t v a r i a b l e = D o s e 29 30 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 T o t a l n u m b e r o f p a r a m e t e r s i n m o d e l = 6 33 T o t a l n u m b e r o f s p e c i f i e d p a r a m e t e r s = 0 34 D e g r e e o f p o l y n o m i a l = 5 35 36 37 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 38 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 This document is a draftfor review purposes only and does not constitute Agency policy. E-195 DRAFT--DO NOT CITE OR QUOTE 1 Parameter Convergence has been set to: 1e-008 2 3 4 5 Default Initial Parameter Values 6 Background = 0 7 Beta(1) = 0 8 Beta(2) = 0 9 Beta(3) = 0 10 Beta(4) = 0 11 B e t a ( 5 ) = 4 . 3 6 9 6 3 e + 0 1 2 12 13 14 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 15 16 ( *** The model parameter(s) -Background -Beta(1) -Beta(4) -Beta(5) 17 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 18 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 19 20 Beta(2) Beta(3) 21 22 B e t a ( 2 ) 1 -0.95 23 24 B e t a ( 3 ) -0.95 1 25 26 27 28 P a r a m e t e r E s t i m a t e s 29 30 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 31 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 32 B a c k g r o u n d 0* * * 33 B e t a ( 1 ) 0* * * 34 Beta(2) 0.00639021 * * * 35 Beta(3) 6.5404e-005 * * * 36 B e t a ( 4 ) 0* * * 37 B e t a ( 5 ) 0* * * 38 39 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 40 41 42 43 A n a l y s i s o f D e v i a n c e T a b l e 44 45 Model Log(likelihood) # Param's Deviance Test d.f. P-value 46 Full model -89.8076 6 47 F i t t e d m o d e l -90.9619 2 2.30853 4 0.6792 48 R e d u c e d m o d e l -218.207 1 256.799 5 <.0001 49 50 AIC: 185.924 51 52 53 G o o d n e s s o f F i t 54 S c a l e d 55 Est. Prob. Expected Observed Residual 56 57 0.0000 0. 0 0 0 0 0.000 0. 0 0 0 53 0. 0 0 0 58 2.5565 0. 0 4 2 0 2.265 2. 0 0 0 54 -0. 180 59 5.6937 0. 1 9 6 9 10.434 8. 0 0 0 53 -0. 841 60 9.7882 0. 4 901 25.976 30. 000 53 1. 1 0 6 61 16.5688 0. 8 7 1 5 46.189 45. 000 53 -0. 488 62 29.6953 0. 9994 52.966 53. 000 53 0. 1 8 5 63 64 C h i ^ 2 = 2 . 2 3 d.f. = 4 P-value = 0.6928 65 66 67 B e n c h m a r k D o s e C o m p u t a t i o n 68 69 S p e c i f i e d e f f e c t = 0.1 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-196 DRAFT--DO NOT CITE OR QUOTE 1 Risk Type Extra risk 2 3 Confidence level 0.95 4 5 BMD 3.98025 6 7 BMDL 3.05855 8 9 BMDU 4.89735 10 11 T a k e n t o g e t h e r , ( 3 . 0 5 8 5 5 , 4 . 8 9 7 3 5 ) is a 90 12 i n t e r v a l f o r t h e B M D 13 14 two-sided confidence 15 E .2 .3 9 .3 . Figurefor Selected Model: Multistage, 5-Degree Multistage Model with 0.95 Confidence Level Fraction Affected 16 13:26 02/08 2010 17 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-197 DRAFT--DO NOT CITE OR QUOTE 1 E.2.40. Ohsako et al., 2001: Ano-Genital Length, PND 120 2 E.2.40.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 3 0.027 171.073 2.592E+01 1.750E+01 Notes exponential (M 3) 3 0.027 171.073 2.5 9 2 E + 0 1 1.750E +01 p o w er hit b oun d (d = 1) exponential (M 4) 2 0.106 168.392 2.248E+00 8.445E-01 exponential (M 5) Hill b 1 0.049 169.789 2.193E+00 9.382E-01 2 0.154 167.647 2.879E+00 8.028E-01 n lower bound hit (n = 1) linear polynom ial, 4degree power 3 0.025 171.258 2.700E+01 1.881E+01 3 0.025 171.258 2.700E+01 1.881E+01 3 0.025 1 71.258 2.7 0 0 E + 0 1 1.881E +01 p o w er b ound hit (p ow er = 1) H ill, unrestricted c 1 0.056 169.555 3.494E+00 3.046E-01 unrestricted (n = 0.591) power, unrestricted 2 0.153 167.654 4.151E +00 2.395E-01 unrestricted (power = 0.291) a C onstant variance m odel selected (p = 0.1 6 5 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .4 0 .2 . Outputfor Selected Model: Hill 6 O hsako et al., 2001: A no-G enital Length, P N D 120 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 5 6 _ O h s a k o _ 2 0 0 1 _ A n o g e n _ H i l l C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 5 6 _ O h s a k o _ 2 0 0 1 _ A n o g e n _ H i l l C V _ 1 . p l t 13 M o n F e b 08 1 3 7 2 7 : 0 2 2 0 1 0 14 15 16 F i g u r e 7 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 This document is a draftfor review purposes only and does not constitute Agency policy. E-198 DRAFT--DO NOT CITE OR QUOTE 1 Power parameter restricted to be greater than 1 2 A constant v ariance model is fit 3 4 Total number of dose groups = 5 5 Total number of records with missing values = 0 6 Maximum number of iterations = 250 7 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 8 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 9 10 11 12 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 13 alpha 7.27386 14 rho 0 Specified 15 intercept 28.905 16 - 5 . 1 0 6 5 17 1 . 5 7 0 4 6 18 2 . 4 3 1 7 19 20 21 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 22 23 ( *** The model parameter(s) -rho -n 24 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 25 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 26 27 alpha intercept v k 28 29 a l p h a 1 4.4e-008 -9.8e-008 7.2e-008 30 31 i n t e r c e p t 4.4e-008 1 -0.57 -0.52 32 33 v -9.8e-008 -0.57 1 -0.23 34 35 k 7.2e-008 -0.52 -0.23 1 36 37 38 39 P a r a m e t e r E s t i m a t e s 40 41 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 42 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 43 alpha 7.07394 1.36138 4.40568 9.7422 44 . n t e r c e p t 2 8 . 9 7 3 2 0 . 7 4 9 9 6 27.5034 30.4431 45 v -5.02686 1.05086 -7.08651 -2.9672 46 n 1 N A 47 k 2.56203 2.11462 -1.58255 6.70661 48 49 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 50 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 51 h a s n o s t a n d a r d e r r o r . 52 53 54 55 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 56 57 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 58 59 60 0 12 28.9 29 3.13 2.66 -0.0889 61 1.04 10 27.9 27.5 2.5 2.66 0.495 62 3 . 4 7 1 10 25.2 26.1 3.21 2.66 -1.09 63 1 1 . 3 6 10 26 24.9 2.85 2.66 1.35 64 3 8 . 4 2 12 23.8 24.3 1.56 2.66 -0.602 65 66 67 68 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-199 DRAFT--DO NOT CITE OR QUOTE 1 Model A1: Yij = Mu(i) + e(ij) 2 Var{ei;ij)} = Sigma^2 3 4 Model A2: Yij = Mu(i) + e(ij) 5 Var{ei;ij)} = Sigma(i) ^2 6 7 Model A3: Yij = Mu(i) + e(ij) 8 Var{ei;ij)} = Sigma^2 9 Model A3 uses any fixed variance parameters that 10 w e r e s p e c i f i e d b y t h e u s e r 11 12 M o d e l R: Yi = M u + e(i) 13 V a r { e ( i ) } = S i g m a ^ 2 14 15 16 L i k e l i h o o d s o f I n t e r e s t 17 18 Model Log(likelihood) # Param's AIC 19 A1 -77.952340 6 167.904680 20 A2 -74.703868 10 1 6 9.407736 21 A3 -77.952340 6 167.904680 22 fitted -79.823277 4 167.646555 23 R -89.824703 2 183.649405 24 25 26 E x p l a n a t i o n o f T e s t s 27 28 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 29 (A2 vs. R) 30 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 31 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 32 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 33 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 34 35 T e s t s o f I n t e r e s t 36 37 Test -2*log(Likelihood Ratio) Test df p-value 38 39 T e s t 1 30.2417 8 0.0001916 40 T e s t 2 6.49694 4 0.165 41 T e s t 3 6.49694 4 0.165 42 T e s t 4 3.74187 2 0.154 43 44 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 45 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 46 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 47 48 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 49 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 50 51 52 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 53 t o b e a p p r o p r i a t e h e r e 54 55 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 56 t o a d e q u a t e l y d e s c r i b e t h e d a t a 57 58 59 B e n c h m a r k D o s e C o m p u t a t i o n 60 61 S p e c i f i e d e f f e c t = 1 62 63 R i s k T y p e = Estimated standard deviations from the control mean 64 65 C o n f i d e n c e l e v e l = 0.95 66 67 BMD = 2.87863 68 69 BMDL = 0.802782 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-200 DRAFT--DO NOT CITE OR QUOTE 1 E .2 .4 0 .3 . Figurefor Selected Model: Hill Hill Model with 0.95 Confidence Level Mean Response 2 13:27 02/08 2010 3 dose 4 5 E .2.40.4. Outputfor Additional Model Presented: Hill, Unrestricted 6 O hsako et al., 2001: A no-G enital Length, P N D 120 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 5 6 _ O h s a k o _ 2 0 0 1 _ A n o g e n _ H i l l C V _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 5 6 _ O h s a k o _ 2 0 0 1 _ A n o g e n _ H i l l C V _ U _ 1 . p l t 13 M o n F e b 08 1 3 7 2 7 : 0 4 2 0 1 0 14 15 16 F i g u r e 7 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 P o w e r p a r a m e t e r is n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t This document is a draftfor review purposes only and does not constitute Agency policy. E-201 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 5 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha 7.27386 12 rho 0 Specified 13 intercept 28.905 14 - 5 . 1 0 6 5 15 1 . 5 7 0 4 6 16 2 . 4 3 1 7 17 18 19 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 20 21 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 22 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 23 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 24 25 alpha intercept v n k 26 27 a l p h a 1 -3.1e-008 7.5e-009 1.7e-008 -8.8e-009 28 29 intercept -3.1e-008 1 0.001 0.0016 -0.13 30 31 v 7.5e-009 0.001 1 0.98 -0.99 32 33 n 1.7e-008 0.0016 0.98 1 -0.97 34 35 k -8.8e-009 -0.13 -0.99 -0.97 1 36 37 38 39 P a r a m e t e r E s t i m a t e s 40 41 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 42 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 43 alpha 7.06192 1.35907 4.3982 9.72564 44 i n t e r c e p t 28.9618 0.754441 27.4831 30.4404 45 v -6.82284 11.1104 -28.5989 14.9532 46 n 0.591421 1.04 -1.44695 2.62979 47 k 7.47064 48.002 -86.6115 101.553 48 49 50 51 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 52 53 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 54 55 56 0 12 28.9 29 3.13 2.66 -0.074 57 1.04 10 27.9 27.3 2.5 2.66 0.71 58 3 . 4 7 1 10 25.2 26.3 3.21 2.66 -1.36 59 1 1 . 3 6 10 26 25.1 2.85 2.66 1.04 60 3 8 . 4 2 12 23.8 24 1.56 2.66 -0. 2 8 4 61 62 63 64 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 65 66 67 M o d e l A1: Yij = Mu(i) + e(ij) 68 V a r { e ( i j ) } = S i g m a ^ 2 69 70 M o d e l A2: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-202 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma(i)^2 2 3 Model A3: Yij = Mu(i) + e(ij) 4 Var{e(ij)} = Sigma^2 5 Model A3 uses any fixed variance parameters that 6 were specified by the user 7 8 M o d e l R: Yi = M u + e(i) 9 Var{e(i)} = Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) # Param's AIC 15 A1 -77.952340 6 167.904680 16 A2 -74.703868 10 1 6 9.407736 17 A3 -77.952340 6 167.904680 18 fitted -79.777354 5 169.554709 19 R -89.824703 2 183.649405 20 21 22 E x p l a n a t i o n o f T e s t s 23 24 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 25 (A2 vs. R) 26 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 27 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 28 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 29 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 30 31 T e s t s o f I n t e r e s t 32 33 Test -2*log(Likelihood Ratio) Test df p-value 34 35 T e s t 1 30.2417 8 0.0001916 36 T e s t 2 6.49694 4 0.165 37 T e s t 3 6.49694 4 0.165 38 T e s t 4 3.65003 1 0.05607 39 40 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 41 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 42 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 43 44 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 45 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 46 47 48 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 49 t o b e a p p r o p r i a t e h e r e 50 51 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 52 m o d e l 53 54 55 B e n c h m a r k D o s e C o m p u t a t i o n 56 57 S p e c i f i e d e f f e c t = 1 58 59 R i s k T y p e = Estimated standard deviations from the control mean 60 61 C o n f i d e n c e l e v e l = 0.95 62 63 BMD = 3.49389 64 65 BMDL = 0.304602 66 67 This document is a draftfor review purposes only and does not constitute Agency policy. E-203 DRAFT--DO NOT CITE OR QUOTE 1 E.2.40.5. F igure f o r A d d itio n a l M o d el Presented: H ill, U nrestricted Hill Model with 0.95 Confidence Level Mean Response 2 13:27 02/08 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-204 DRAFT--DO NOT CITE OR QUOTE 1 E.2.41. Sew all et al., 1995: T4 In Serum 2 E .2 .4 1 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of x2pFreedom Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M2) 3 0.722 204.495 1.869E+01 1.243E+01 Notes exponential (M3) 3 0 .7 2 2 2 0 4 .4 9 5 1.869E +01 1.243E +01 p o w er hit boun d (d = 1) exponential (M4) 2 0.854 205.483 1.106E+01 4.650E+00 exponential (M5) Hill b 2 0 .8 5 4 2 0 5 .4 8 3 1.106E +01 4 .6 5 0 E + 0 0 p o w er hit boun d (d = 1) 2 0.898 205.382 1.031E+01 3.603E+00 n lower bound hit (n = 1) linear polynom ial, 4degree power 3 0.576 205.150 2.238E+01 1.619E+01 3 0.576 205.150 2.238E+01 1.619E+01 3 0 .5 7 6 2 0 5 .1 5 0 2 .2 3 8 E + 0 1 1.619E +01 p o w er bound hit (p ow er = 1) Hill, unrestricted c 1 0.864 207.196 9.706E+00 1.973E+00 unrestricted (n = 0.569) power, unrestricted 2 0.985 205.197 9.726E +00 1.914E+00 unrestricted (power = 0.538) a Constant variance m odel selected (p = 0.4 0 7 8 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 E .2 .4 1 .2 . Outputfor Selected Model: Hill 5 Sew all et al., 1995: T4 In Serum 6 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 5 8 _ S e w a l l _ 1 9 9 5 _ T 4 _ H i l l C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 5 8 _ S e w a l l _ 1 9 9 5 _ T 4 _ H i l l C V _ 1 . p l t 13 M o n F e b 08 1 3 : 2 8 : 1 5 2 0 1 0 14 15 16 F i g u r e 1, S a l i n e n o n i n i t i a t e d 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 P o w e r p a r a m e t e r r e s t r i c t e d t o b e g r e a t e r t h a n 1 This document is a draftfor review purposes only and does not constitute Agency policy. E-205 DRAFT--DO NOT CITE OR QUOTE 1 A constant variance model is fit 2 3 Total number of dose groups = 5 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 12 alpha 33.0913 13 rho 0 Specified 14 intercept 30.6979 15 - 1 2 . 2 9 3 7 16 0 . 9 5 0 8 1 5 17 1 2 . 5 8 0 8 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( *** The model parameter(s) -rho -n 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 24 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 25 26 alpha intercept v k 27 28 a l p h a 1 -1.2e-009 -1.8e-008 1.5e-008 29 30 intercept -1.2e-009 1 0.3 0.65 31 32 -1.8e-008 0.3 1 0.89 33 34 k 1.5e-008 -0.65 -0.89 1 35 36 37 38 P a r a m e t e r E s t i m a t e s 39 40 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 41 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 42 alpha 29.5556 6.23087 17.3433 41.7679 43 . n t e r c e p t 3 0 . 3 9 5 7 1 . 6 8 7 4 7 27.0883 33.7031 44 v -18.2488 7.72836 -33.3961 -3.10154 45 n 1 N A 46 k 24.2883 26.743 -28.127 76.7035 47 48 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 49 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 50 h a s n o s t a n d a r d e r r o r . 51 52 53 54 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 55 56 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled R 57 58 59 09 30 .7 30.4 4.66 5.44 0.167 60 3 . 2 9 1 9 27 .9 28.2 7.17 5.44 -0.188 61 7 . 1 0 7 9 25 .9 26.3 6.81 5.44 -0.204 62 1 6 . 6 3 9 23 .6 23 5.38 5.44 0.319 63 4 4 . 6 6 9 18 .4 18.6 4.12 5.44 -0.0942 64 65 66 67 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 68 69 70 M o d e l A1: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-206 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma^2 2 3 Model A2: Yij = Mu(i) + e(ij) 4 Var{e(ij)} = Sigma(i)^2 5 6 Model A3: Yij = Mu(i) + e(ij) 7 Var{e(ij)} = Sigma^2 8 Model A3 uses any fixed variance parameters that 9 were specified by the user 10 11 M o d e l R: Yi = M u + e(i) 12 V a r { e ( i ) } = S i g m a ^ 2 13 14 15 L i k e l i h o o d s o f I n t e r e s t 16 17 Model Log(likelihood) # Param's AIC 18 A1 -98.583448 6 209.166896 19 A2 -96.590204 10 213.180407 20 A3 -98.583448 6 209.166896 21 fitted -98.691143 4 205.382286 22 R -109.013252 2 222.026503 23 24 25 E x p l a n a t i o n o f T e s t s 26 27 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 28 (A2 vs. R) 29 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 30 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 31 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 32 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 33 34 T e s t s o f I n t e r e s t 35 36 Test -2*log(Likelihood Ratio) Test df p-value 37 38 T e s t 1 24.8461 8 0.001651 39 T e s t 2 3.98649 4 0.4078 40 T e s t 3 3.98649 4 0.4078 41 T e s t 4 0.21539 2 0.8979 42 43 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 44 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 45 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 46 47 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 48 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 49 50 51 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 52 t o b e a p p r o p r i a t e h e r e 53 54 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 55 t o a d e q u a t e l y d e s c r i b e t h e d a t a 56 57 58 B e n c h m a r k D o s e C o m p u t a t i o n 59 60 S p e c i f i e d e f f e c t = 1 61 62 R i s k T y p e = Estimated standard deviations from the control mean 63 64 C o n f i d e n c e l e v e l = 0.95 65 66 BMD = 10.306 67 68 BMDL = 3.60269 This document is a draftfor review purposes only and does not constitute Agency policy. E-207 DRAFT--DO NOT CITE OR QUOTE 1 E .2 .4 1 .3 . Figurefor Selected Model: Hill Hill Model with 0.95 Confidence Level Mean Response 2 13:28 02/08 2010 3 dose 4 5 E .2.41.4. Outputfor Additional Model Presented: Hill, Unrestricted 6 Sew all et al., 1995: T4 In Serum 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 5 8 _ S e w a l l _ 1 9 9 5 _ T 4 _ H i l l C V _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 5 8 _ S e w a l l _ 1 9 9 5 _ T 4 _ H i l l C V _ U _ 1 . p l t 13 M o n F e b 08 1 3 : 2 8 : 1 5 2 0 1 0 14 15 16 F i g u r e 1, S a l i n e n o n i n i t i a t e d 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 P o w e r p a r a m e t e r is n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t This document is a draftfor review purposes only and does not constitute Agency policy. E-208 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 5 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha 33.0913 12 rho 0 Specified 13 intercept 30.6979 14 - 1 2 . 2 9 3 7 15 0 . 9 5 0 8 1 5 16 1 2 . 5 8 0 8 17 18 19 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 20 21 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 22 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 23 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 24 25 alpha intercept v n k 26 27 a l p h a 1 -3.9e-005 0.00022 0.00021 -0.00022 28 29 intercept -3.9e-005 1 -0.17 -0.31 0.18 30 31 v 0.00022 -0.17 1 0.97 -1 32 33 n 0.00021 -0.31 0.97 1 -0.98 34 35 k -0.00022 0.18 -1 -0.98 1 36 37 38 39 P a r a m e t e r E s t i m a t e s 40 41 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 42 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 43 alpha 29.4337 6.20518 17.2718 41.5957 44 i n t e r c e p t 30.7096 1.79801 27.1855 34.2336 45 v -143.244 3972.28 -7928.78 7642.29 46 n 0.569063 0.947248 -1.28751 2.42564 47 k 2856.29 171186 -332662 338374 48 49 50 51 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 52 53 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 54 55 56 09 30.7 30 .7 4.66 5.43 -0.00646 57 3 . 2 9 1 9 27.9 27 .7 7.17 5.43 0.0842 58 7 . 1 0 7 9 25.9 26 .1 6.81 5.43 -0.134 59 1 6 . 6 3 9 23.6 23 .4 5.38 5.43 0.0657 60 4 4 . 6 6 9 18.4 18 .4 4.12 5.43 -0.00948 61 62 63 64 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 65 66 67 M o d e l A1: Yij = M u ( i ) + e (ij) 68 V a r { e i j ) } = S i g m a ^ 2 69 70 M o d e l A2: Yij = M u ( i ) + e (ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-209 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma(i)^2 2 3 Model A3: Yij = Mu(i) + e(ij) 4 Var{e(ij)} = Sigma^2 5 Model A3 uses any fixed variance parameters that 6 were specified by the user 7 8 M o d e l R: Yi = M u + e(i) 9 Var{e(i)} = Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) # Param's AIC 15 A1 -98.583448 6 209.166896 16 A2 -96.590204 10 213.180407 17 A3 -98.583448 6 209.166896 18 fitted -98.598183 5 207.196367 19 R -109.013252 2 222.026503 20 21 22 E x p l a n a t i o n o f T e s t s 23 24 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 25 (A2 vs. R) 26 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 27 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 28 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 29 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 30 31 T e s t s o f I n t e r e s t 32 33 Test -2*log(Likelihood Ratio) Test df p-value 34 35 T e s t 1 24.8461 8 0.001651 36 T e s t 2 3.98649 4 0.4078 37 T e s t 3 3.98649 4 0.4078 38 T e s t 4 0.0294713 1 0.8637 39 40 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 41 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 42 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 43 44 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 45 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 46 47 48 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 49 t o b e a p p r o p r i a t e h e r e 50 51 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 52 t o a d e q u a t e l y d e s c r i b e t h e d a t a 53 54 55 B e n c h m a r k D o s e C o m p u t a t i o n 56 57 S p e c i f i e d e f f e c t = 1 58 59 R i s k T y p e = Estimated standard deviations from the control mean 60 61 C o n f i d e n c e l e v e l = 0.95 62 63 BMD = 9.70574 64 65 BMDL = 1.97319 66 This document is a draftfor review purposes only and does not constitute Agency policy. E-210 DRAFT--DO NOT CITE OR QUOTE 1 E.2.41.5. F igure f o r A d d itio n a l M o d el Presented: H ill, U nrestricted Hill Model with 0.95 Confidence Level Mean Response 2 13:28 02/08 2010 3 4 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-211 DRAFT--DO NOT CITE OR QUOTE 1 E.2.42. Shi et al., 2007: E stradiol 17B, PE9 2 E .2 .4 2 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 3 0.010 391.638 6.976E+00 3.761E+00 Notes exponential (M 3) exponential (M4) b exponential (M 5) 3 0 .0 1 0 3 9 1 .6 3 8 6 .9 7 6 E + 0 0 3 .7 6 1 E + 0 0 p o w er hit bound (d = 1) 2 0.690 382.969 8.068E-01 3.544E-01 2 0 .6 9 0 3 8 2 .9 6 9 8 .068E -01 3 .544E -01 p o w er hit bound (d = 1) H ill 2 0.975 382.278 7.239E -01 error n lo w er b ound hit (n = 1) linear polynom ial, 4degree power 3 0.003 394.308 9.841E+00 6.687E+00 3 0.003 394.308 9.841E+00 6.687E+00 3 0.003 3 9 4 .3 0 8 9 .8 4 1 E + 0 0 6 .6 8 7 E + 0 0 p o w er b oun d hit (p ow er = 1) H ill, unrestricted 1 0.897 384.243 7.086E -01 error unrestricted (n = 0.875) power, unrestricted 2 0.506 383.590 6.280E-01 3.304E-02 unrestricted (pow er = 0.222) a N on-constant variance m odel selected (p = 0 .0521) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2 .4 2 .2 . Outputfor Selected Model: Exponential (M4) 6 Shi et al., 2007: Estradiol 17B, PE9 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 5 9 _ S h i _ 2 0 0 7 _ E s t r a d i o l _ E x p _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 M o n F e b 08 1 3 : 2 8 : 5 2 2 0 1 0 14 15 16 F i g u r e 4 P E 9 o n l y 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-212 DRAFT--DO NOT CITE OR QUOTE 1 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 2 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 3 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 4 5 6 Dependent variable = Mean 7 Independent variable = Dose 8 Data are assumed to be distributed: normally 9 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 10 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 11 12 T o t a l n u m b e r o f d o s e g r o u p s = 5 13 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 14 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 15 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 16 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 17 18 M L E s o l u t i o n p r o v i d e d : E x a c t 19 20 21 I n i t i a l P a r a m e t e r V a l u e s 22 23 Variable Model 4 24 25 lnalpha 2.65881 26 rho 0.913414 27 a 108 28 b 0 . 2 7 7 6 3 7 29 c 0 . 3 4 0 1 3 6 30 d 1 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 Variable Model 4 37 38 lnalpha 1.66773 39 rho 1.15314 40 a 1 0 3 . 1 4 6 41 b 1 . 0 0 6 8 5 42 c 0 . 4 1 8 7 4 2 43 d 1 44 45 46 T a b l e o f S t a t s F r o m I n p u t D a t a 47 48 Dose N Obs Mean Obs Std 49 50 0 10 102.9 41.41 51 0.3418 10 86.19 19.58 52 1.075 10 63.33 29.36 53 5.23 10 48.1 18.82 54 13.91 10 38.57 22.59 55 56 57 E s t i m a t e d V a l u e s o f I n t e r e s t 58 59 Dose Est Mean Est Std Scaled Residual 60 61 0 103.1 33.35 -0.02738 62 0 . 3 4 1 8 85.69 29.96 0.05296 63 1 . 0 7 5 63.51 25.21 -0.02238 64 5 . 2 3 43.5 20.27 0.7167 65 1 3 . 9 1 43.19 20.19 -0.7237 66 67 68 69 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d : 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-213 DRAFT--DO NOT CITE OR QUOTE 1 Model A1: Yij = Mu(i) + e(ij) 2 V a r { e i ;ij)} = S i g m a ^ 2 3 4 Model A2: Yij = Mu(i) + e(ij) 5 V a r { e i :ij)} = S i g m a ( i ) ^ 2 6 7 Model A3: Yij = Mu(i) + e(ij) 8 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 9 10 M o d e l R: Yij = M u + e(i) 11 V a r { e i ij)} = S i g m a ^ 2 12 13 14 L i k e l i h o o d s o f I n t e r e s t 15 16 Model Log(likelihood) DF AIC 17 18 A1 -188.3615 6 388.7231 19 A2 -183.667 10 387.3339 20 A3 -186.1132 7 386.2263 21 R -203.3606 2 410.7211 22 4 -186.4844 5 382.9687 23 24 25 Additive constant for all log-likelihoods = -45.95. This constant added to the 26 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 27 d e p e n d o n t h e m o d e l p a r a m e t e r s . 28 29 30 E x p l a n a t i o n o f T e s t s 31 32 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 33 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 34 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 35 36 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 37 38 39 T e s t s o f I n t e r e s t 40 41 T e s t -2*log(Likelihood Ratio) D. F. p-value 42 43 T e s t 1 39.39 8 < 0.0001 44 T e s t 2 9.389 4 0.05208 45 T e s t 3 4.892 3 0.1798 46 T e s t 6a 0.7424 2 0.6899 47 48 49 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 50 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 51 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 52 53 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 54 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e . 55 56 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 57 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 58 59 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 60 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 61 62 63 B e n c h m a r k D o s e C o m p u t a t i o n s : 64 65 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 66 67 R i s k T y p e = E s t i m a t e d s t a n d a r d d e v i a t i o n s f r o m c o n t r o l 68 69 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-214 DRAFT--DO NOT CITE OR QUOTE 1 BMD 0.806817 2 3 BMDL = 0.354366 4 5 6 E .2 .4 2 .3 . Figurefor Selected Model: Exponential (M4) Exponential Model 4 with 0.95 Confidence Level Mean Response 7 13:28 02/08 2010 8 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-215 DRAFT--DO NOT CITE OR QUOTE 1 E .2 .4 3 . S m ia lo w ic z et a l., 2 0 0 8 : P F C p er 1 0 A6 C ells 2 E .2 .4 3 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 3 0.101 901.897 8.343E+00 5.064E+00 Notes exponential (M 3) 3 0.101 9 0 1 .8 9 7 8 .3 4 3 E + 0 0 5 .0 6 4 E + 0 0 p ow er hit b ound (d = 1) exponential (M 4) 2 0.044 903.897 8.325E+00 1.465E+00 exponential (M 5) 2 0 .0 4 4 9 0 3 .8 9 7 8 .3 2 5 E + 0 0 1 .465E + 00 p ow er hit b ound (d = 1) H ill 2 0.063 9 0 3 .1 9 2 3 .6 6 9 E + 0 0 6.9 7 0 E -0 1 n lo w er b oun d hit (n = 1) linear polynom ial, 4degree power 3 0.048 903.585 1.373E+01 1.053E+01 3 0.048 903.585 1.374E+01 1.053E+01 3 0 .0 4 8 9 0 3 .5 8 5 1.373E +01 1.053E +01 p ow er b oun d hit (p ow er = 1) H ill, unrestricted 1 0.213 901.219 1.928E+00 2.208E-01 unrestricted (n = 0.35) power, unrestricted b 2 0.481 899.130 1.902E+00 2.158E-01 unrestricted (power = 0.333) a C onstant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2.43.2. Outputfor Selected Model: Power, Unrestricted 6 S m ia lo w icz et al., 2008: PFC per 10A6 C ells 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 6 0 _ S m i a l _ 2 0 0 8 _ P F C c e l l s _ P w r C V _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 6 0 _ S m i a l _ 2 0 0 8 _ P F C c e l l s _ P w r C V _ U _ 1 . p l t 13 M o n F e b 08 1 3 : 2 9 : 3 8 2 0 1 0 14 15 16 A n t i R e s p o n s e t o S R B C s , P F C p e r 1 0 t o 6 c e l l s , T a b l e 4 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 T h e p o w e r is n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t This document is a draftfor review purposes only and does not constitute Agency policy. E-216 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 5 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha = 232385 12 rho = 0 Specified 13 control = 1491 14 slope = -491.716 15 power = 0.288021 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 21 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 22 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 23 24 alpha control slope power 25 26 a l p h a 1 -3.4e-009 1.8e-009 -1.2e-010 27 o CO K> o CO K> Lf) CD O Lf) CD O 28 control -3.4e-009 1 29 30 slope 1.8e-009 1 0.94 31 32 power -1.2e-010 0.94 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 39 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 40 alpha 219793 37974.5 145365 294222 41 control 1470.48 123.73 1227.98 1712.99 42 slope -378.406 157.002 -686.125 -70.6872 43 power 0.333124 0.113501 0.110666 0.555581 44 45 46 47 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 48 49 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res. 50 51 52 0 15 1.49e+003 1.47e+003 716 469 0.169 53 0 . 4 3 8 14 1 . 1 3 e + 0 0 3 1.18e+003 171 469 -0.431 54 2 . 4 6 4 15 945 959 516 469 -0.12 55 13.4 15 677 572 465 469 0.867 56 3 1 . 6 5 8 161 274 117 469 -0.684 57 58 59 60 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 61 62 63 M o d e l A1: Yij = Mu(i) + e(ij ) 64 V a r { e ( i j ) } = S i g m a ^ 2 65 66 M o d e l A2: Yij Mu(i) + e(ij) 67 Var{e(ij)} Sigma(i)^2 68 69 M o d e l A3: Yij Mu(i) + e(ij) 70 Var{e(ij)} Sigma^2 This document is a draftfor review purposes only and does not constitute Agency policy. E-217 DRAFT--DO NOT CITE OR QUOTE 1 Model A3 uses any fixed variance parameters that 2 were specified by the user 3 4 M o d e l R: Yi = M u + e(i) 5 Var{e(i)} = Sigma^2 6 7 8 Likelihoods of Interest 9 10 Model Log(likelihood) # Param's AIC 11 A1 -444.832859 6 901.665718 12 A2 -425.402825 10 870.805651 13 A3 -444.832859 6 901.665718 14 fitted -445.564823 4 899.129647 15 R -463.753685 2 931.507371 16 17 18 E x p l a n a t i o n o f T e s t s 19 20 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 21 (A2 vs. R) 22 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 23 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 24 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 25 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 26 27 T e s t s o f I n t e r e s t 28 29 Test -2*log(Likelihood Ratio) Test df p-value 30 31 T e s t 1 76.7017 8 <.0001 32 T e s t 2 38.8601 4 <.0001 33 T e s t 3 38.8601 4 <.0001 34 T e s t 4 1.46393 2 0.481 35 36 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 37 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 38 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 39 40 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. C o n s i d e r r u n n i n g a 41 n o n - h o m o g e n e o u s v a r i a n c e m o d e l 42 43 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t t o c o n s i d e r a 44 d i f f e r e n t v a r i a n c e m o d e l 45 46 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 47 t o a d e q u a t e l y d e s c r i b e t h e d a t a 48 49 50 B e n c h m a r k D o s e C o m p u t a t i o n 51 52 S p e c i f i e d e f f e c t = 1 53 54 R i s k T y p e = Estimated standard deviations from the control mean 55 56 C o n f i d e n c e l e v e l = 0.95 57 58 B M D = 1 . 9 0 2 4 9 59 60 61 B M D L = 0 . 2 1 5 8 4 3 62 This document is a draftfor review purposes only and does not constitute Agency policy. E-218 DRAFT--DO NOT CITE OR QUOTE 1 E.2.43.3. F igure f o r Selected M odel: Power, U nrestricted Power Model with 0.95 Confidence Level Mean Response 2 13:29 02/08 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-219 DRAFT--DO NOT CITE OR QUOTE 1 E.2.44. Sm ialow icz et al., 2008: PFC per Spleen 2 E .2 .4 4 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 3 0.124 377.565 1.334E+01 8.593E+00 Notes exponential (M 3) 2 0.069 379.138 1.536E+01 8.895E+00 exponential (M 4) 3 0.124 377.565 1.334E+01 8.593E+00 exponential (M 5) 1 0.021 381.138 1.536E+01 8.895E+00 H ill 2 0.116 378.108 1.568E+01 error n lo w er b oun d hit (n = 1) linear polynom ial, 4degree power 3 0.126 377.522 2.055E+01 1.624E+01 3 0.126 377.522 2.055E+01 1.624E+01 3 0 .1 2 6 3 7 7 .5 2 2 2 .0 5 5 E + 0 1 1.624E +01 p o w er b oun d hit (p ow er = 1) H ill, unrestricted 1 0.103 378.463 1.202E+01 error unrestricted (n = 0.544) power, unrestricted b 2 0.270 376.420 1.187E+01 3.762E+00 unrestricted (power = 0.531) a N on-con stant variance m odel selected (p = 0 .0 0 1 1 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .2.44.2. Outputfor Selected Model: Power, Unrestricted 6 Sm ialow icz et al., 2008: PFC per Spleen 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 6 1 _ S m i a l _ 2 0 0 8 _ P F C s p l e e n _ P w r _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 6 1 _ S m i a l _ 2 0 0 8 _ P F C s p l e e n _ P w r _ U _ 1 . p l t 13 M o n F e b 08 1 3 : 3 0 : 1 6 ~ 2 0 1 0 14 15 16 A n t i R e s p o n s e t o S R B C s - P F C x 10 t o t h e 4 p e r s p l e e n , T a b l e 4 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 T h e p o w e r is n o t r e s t r i c t e d 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-220 DRAFT--DO NOT CITE OR QUOTE 1 Total number of dose groups = 5 2 Total number of records with missing values = 0 3 Maximum number of iterations = 250 4 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 5 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 6 7 8 9 Default Initial Parameter Values 10 lalpha = 4.76607 11 rho = 0 12 control = 27.8 13 slope = -9.21898 14 power = 0.286443 15 16 17 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 18 19 lalpha rho control slope power 20 21 l a l p h a 1 -0.98 0.25 -0.28 -0.22 22 23 rho -0.98 1 -0.3 0.28 0.22 24 CO CO o CO CO o 25 c o n t r o l 0.25 -0.3 1 -0.74 26 27 slope -0.28 0.28 1 0.99 28 29 power -0.22 0.22 -0.74 0.99 1 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 36 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 37 lalpha 0.746922 1.02058 -1.25337 2.74721 38 rho 1.36826 0.355827 0.67085 2.06567 39 control 25.3816 2.96691 19.5666 31.1967 40 slope -3.5662 2.52558 -8.51626 1.38385 41 power 0.531216 0.175728 0.186796 0.875637 42 43 44 45 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 46 47 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res. 48 49 50 0 15 27.8 25.4 13. 4 13.3 0.706 51 0 . 4 3 8 14 21 23.1 13. 6 12.4 -0.626 52 2 . 4 6 4 15 17.6 19.6 9. 4 11.1 -0.704 53 13.4 15 12.6 11.2 8. 7 7.6 0.702 54 3 1 . 6 5 8 3 3.03 3. 1 3.1 -0.0313 55 56 57 58 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 59 60 61 M o d e l A1: Yij = Mu(i) + e(ij ) 62 V a r { e ( i j ) } = S i g m a ^ 2 63 64 M o d e l A2: Yij Mu(i) + e(ij) 65 Var{e(ij)} Sigma(i)^2 66 67 M o d e l A3: Yij = Mu(i) + e(ij) 68 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 69 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 70 w e r e s p e c i f i e d b y t h e u s e r This document is a draftfor review purposes only and does not constitute Agency policy. E-221 DRAFT--DO NOT CITE OR QUOTE 1 2 M o d e l R: Yi = M u + e(i) 3 Var{e(i)} = Sigma^2 4 5 6 Likelihoods of Interest 7 8 Model Log(likelihood) # Param's AIC 9 A1 -190.565019 6 393.130038 10 A2 -181.476284 10 3 8 2.952569 11 A3 -181.900030 7 377.800059 12 fitted -183.210137 5 376.420274 13 R -204.636496 2 413.272993 14 15 16 E x p l a n a t i o n o f T e s t s 17 18 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 19 (A2 vs. R) 20 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 21 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 22 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 23 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 24 25 T e s t s o f I n t e r e s t 26 27 Test -2*log(Likelihood Ratio) Test df p-value 28 29 T e s t 1 46.3204 8 <.0001 30 T e s t 2 18.1775 4 0.001139 31 T e s t 3 0.84749 3 0.8381 32 T e s t 4 2.62021 2 0.2698 33 34 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 35 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 36 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 37 38 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 39 m o d e l a p p e a r s t o b e a p p r o p r i a t e 40 41 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 42 t o b e a p p r o p r i a t e h e r e 43 44 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 45 t o a d e q u a t e l y d e s c r i b e t h e d a t a 46 47 48 B e n c h m a r k D o s e C o m p u t a t i o n 49 50 S p e c i f i e d e f f e c t = 1 51 52 R i s k T y p e = Estimated standard deviations from the control mean 53 54 C o n f i d e n c e l e v e l = 0.95 55 56 B M D = 1 1 . 8 7 4 8 57 58 59 B M D L = 3 . 7 6 1 6 1 60 This document is a draftfor review purposes only and does not constitute Agency policy. E-222 DRAFT--DO NOT CITE OR QUOTE 1 E.2.44.3. F igure f o r Selected M odel: Power, U nrestricted Power Model with 0.95 Confidence Level Mean Response 2 13:30 02/08 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-223 DRAFT--DO NOT CITE OR QUOTE 1 E .2.45. Toth et al., 1979: A m yloidosis 2 E .2 .4 5 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 2 0.040 1 49.120 1.965E +01 1.283E +01 p o w er b oun d hit (p ow er = 1) logistic negative intercept (intercept = 2 0.019 151.340 3.701E+01 2.858E+01 -2.16) log-logistic a 2 0.053 148.269 1.503E+01 8.747E+00 slope bound hit (slope = 1) log-probit 2 0.009 152.855 3.7 8 2 E + 0 1 2 .5 0 2 E + 0 1 slop e b ound hit (slop e = 1) multistage, 3degree probit 2 0.040 149.120 1.965E+01 1.283E+01 final B = 0 negative intercept (intercept = 2 0.021 151.115 3.467E+01 2.657E+01 -1.276) W eibull 2 0.040 1 49.120 1.965E +01 1.283E +01 p o w er b oun d hit (p ow er = 1) gamma, unrestricted log-logistic, unrestricted b log-probit, unrestricted W eibull, unrestricted 2 0.959 140.119 4.349E-01 2.891E-03 unrestricted (pow er = 0.254) 2 0.903 140.240 4.843E-01 5.312E-03 unrestricted (slope = 0.326) 2 0.870 140.315 4.960E-01 7.292E-03 unrestricted (slope = 0.186) 2 0.933 140.174 4.641E-01 4.069E-03 unrestricted (pow er = 0.289) aBest-fitting m odel, BM D S output presented in this appendix b Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .4 5 .2 . Outputfor Selected Model: Log-Logistic 6 Toth et al., 1979: A m yloidosis 7 8 9 10 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 6 2 _ T o t h _ 1 9 7 9 _ A m y 1 y r _ L o g L o g i s t i c _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 6 2 _ T o t h _ 1 9 7 9 _ A m y 1 y r _ L o g L o g i s t i c _ 1 . p l t 13 M o n F e b 08 1 3 : 3 0 : 5 4 2 0 1 0 14 15 16 T a b l e 2 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e This document is a draftfor review purposes only and does not constitute Agency policy. E -224 D RAFT-- DO N O T CITE OR QUOTE 1 Slope pa r a m e t e r is r estricted as slope >= 1 2 3 Total number of observations = 4 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 12 13 14 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 15 background = 0 16 intercept = -4.54593 17 slope = 1 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( * * * T h e m o d e l p a r a m e t e r ( s ) - s l o p e 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 24 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 25 26 background intercept 27 28 b a c k g r o u n d 1 -0.49 29 30 i n t e r c e p t -0.49 1 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 37 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf 38 b a c k g r o u n d 0 . 0 6 9 9 9 1 8 * * * 39 intercept -4.90704 * * * 40 s l o p e 1 41 42 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 43 44 45 46 A n a l y s i s o f D e v i a n c e T a b l e 47 48 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 49 F u l l m o d e l -68.017 4 50 F i t t e d m o d e l -72.1346 2 8.23525 2 0.01628 51 R e d u c e d m o d e l -82.0119 1 27.99 3 <.0001 52 53 AIC: 148.269 54 55 56 G o o d n e s s o f F i t 57 58 Dose Est. Prob. Expected Observed Size Residual 59 60 0.0000 0.0700 2.660 0.000 38 -1.691 61 0.5732 0.0739 3.252 5.000 44 1 . 0 0 7 62 14.2123 0.1584 6.971 10.000 44 1 . 2 5 1 63 91.2070 0.4446 19.117 17.000 43 - 0 .650 64 65 C h i ^ 2 = 5 . 8 6 d.f. = 2 P-value = 0.0534 66 67 68 B e n c h m a r k D o s e C o m p u t a t i o n 69 70 S p e c i f i e d e f f e c t = 0.1 This document is a draftfor review purposes only and does not constitute Agency policy. E-225 DRAFT--DO NOT CITE OR QUOTE 1 2 Risk Type 3 4 Confidence level 5 6 BMD 7 8 BMDL 9 10 Extra risk 0.95 15.0264 8.74665 11 E .2 .4 5 .3 . Figurefor Selected Model: Log-Logistic Log-Logistic Model with 0.95 Confidence Level 12 13 14 15 E .2.45.4. Outputfor Additional Model Presented: Log-Logistic, Unrestricted 16 Toth et al., 1979: A m y lo id o sis 17 18 19 20 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 21 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 6 2 _ T o t h _ 1 9 7 9 _ A m y 1 y r _ L o g L o g i s t i c _ U _ 1 . ( d ) 22 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 6 2 _ T o t h _ 1 9 7 9 _ A m y 1 y r _ L o g L o g i s t i c _ U _ 1 . p l t 23 M o n F e b 08 1 3 : 3 0 : 5 4 2 0 1 0 24 25 26 T a b l e 2 27 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-226 DRAFT--DO NOT CITE OR QUOTE 1 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 2 3 P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))] 4 5 6 Dependent variable = DichEff 7 Independent variable = Dose 8 Slope pa r a m e t e r is not restricted 9 10 T o t a l n u m b e r o f o b s e r v a t i o n s = 4 11 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 12 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 13 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 16 17 18 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 19 20 21 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 22 background = 0 23 intercept = -1.92722 24 slope = 0.314472 25 26 27 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 28 29 ( * * * T h e m o d e l p a r a m e t e r ( s ) - b a c k g r o u n d 30 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 31 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 32 33 intercept slope 34 CO o CO o 35 i n t e r c e p t 1 36 37 s l o p e 1 38 39 40 41 P a r a m e t e r E s t i m a t e s 42 43 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 44 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf 45 b a c k g r o u n d 0* * * 46 intercept -1.96073 * * * 47 slope 0.326156 48 49 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 50 51 52 53 A n a l y s i s o f D e v i a n c e T a b l e 54 55 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 56 F u l l m o d e l -68.017 4 57 F i t t e d m o d e l -68.1201 2 0.206341 2 0.902 58 R e d u c e d m o d e l -82.0119 1 27.99 3 <.0001 59 60 AIC: 140.24 61 62 63 G o o d n e s s o f F i t 64 65 Dose Est. Prob. Expected Observed Size Residu. 66 67 0.0000 0.0000 0.000 0.000 38 0.000 68 0.5732 0.1051 4.623 5.000 44 0 . 1 8 6 69 14.2123 0.2507 11.029 10.000 44 - 0 . 3 5 8 70 91.2070 0.3802 16.348 17.000 43 0.205 This document is a draftfor review purposes only and does not constitute Agency policy. E-227 DRAFT--DO NOT CITE OR QUOTE 1 2 Chi^2 = 0.20 d.f. = 2 P-value = 0.9028 3 4 5 Benchmark Dose Computation 6 7 Specified effect : 0.1 8 9 Risk Type : Extra risk 10 11 C o n f i d e n c e l e v e l : 0.95 12 13 BMD ^ 0.484272 14 15 BMDL ^ 0.00531211 16 17 18 E .2.45.5. Figurefor Additional Model Presented: Log-Logistic, Unrestricted Log-Logistic Model with 0.95 Confidence Level 19 13:30 02/08 2010 20 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-228 DRAFT--DO NOT CITE OR QUOTE 1 E .2.46. Toth et al., 1979: Skin L esions 2 E .2 .4 6 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes gamma 2 0.032 1 56.346 1.037E +01 7 .4 7 0 E + 0 0 p o w er b oun d hit (p ow er = 1) logistic negative intercept (intercept = 2 0.005 161.421 2.487E+01 1.982E+01 -1.999) log-logistic a 2 0.078 153.963 6.413E+00 4.025E+00 slope bound hit (slope = 1) log-probit 2 0.003 1 61.788 1.887E +01 1.280E +01 slop e b ound hit (slop e = 1) multistage, 3degree probit 2 0.032 156.346 1.037E+01 7.470E+00 final B = 0 negative intercept (intercept = 2 0.006 160.991 2.309E+01 1.858E+01 -1.198) W eibull 2 0.032 1 56.346 1.037E +01 7 .4 7 0 E + 0 0 p o w er b oun d hit (p ow er = 1) gamma, unrestricted log-logistic, unrestricted b log-probit, unrestricted W eibull, unrestricted 2 0.945 147.148 error error unrestricted (pow er = 0.341) 2 0.744 147.631 5.969E-01 6.773E -02 unrestricted (slope = 0.48) 2 0.670 147.844 5.939E-01 8.147E-02 unrestricted (slope = 0.279) 2 0.866 147.324 5.539E-01 5.181E -02 unrestricted (pow er = 0.405) aBest-fitting m odel, BM D S output presented in this appendix b Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .4 6 .2 . Outputfor Selected Model: Log-Logistic 6 Toth et al., 1979: Skin L esions 7 8 9 10 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 6 3 _ T o t h _ 1 9 7 9 _ S k i n L e s _ L o g L o g i s t i c _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 6 3 _ T o t h _ 1 9 7 9 _ S k i n L e s _ L o g L o g i s t i c _ 1 . p l t 13 _ W e d F e b 10 1 4 : 4 7 : 5 3 2 0 1 0 ~ 14 15 16 T a b l e 2 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e This document is a draftfor review purposes only and does not constitute Agency policy. E -229 D RAFT-- DO N O T CITE OR QUOTE 1 Slope pa r a m e t e r is r estricted as slope >= 1 2 3 Total number of observations = 4 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 12 13 14 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 15 background = 0 16 intercept = -3.94312 17 slope = 1 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( * * * T h e m o d e l p a r a m e t e r ( s ) - s l o p e 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 24 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 25 26 background intercept 27 28 b a c k g r o u n d 1 -0.43 29 30 i n t e r c e p t -0.43 1 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 37 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 38 b a c k g r o u n d 0 . 0 5 6 4 5 6 2 * * * 39 intercept -4.05558 * * * 40 s l o p e 1* * * 41 42 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 43 44 45 46 A n a l y s i s o f D e v i a n c e T a b l e 47 48 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 49 Full model -71.5177 4 50 F i t t e d m o d e l -74.9813 2 6.92722 2 0.03132 51 R e d u c e d m o d e l -95.8498 1 48.6642 3 <.0001 52 53 AIC: 153.963 54 55 56 G o o d n e s s o f F i t 57 S c a l e d 58 Dose Est. Prob. Expected Observed Size Residual 59 60 0.0000 0.0565 2.145 0.000 38 -1.508 61 0.5732 0.0657 2.892 5.000 44 1 . 2 8 2 62 14.2123 0.2429 10.687 13.000 44 0 . 8 1 3 63 91.2070 0.6343 27.275 25.000 43 - 0 .720 64 65 C h i ^ 2 = 5 . 1 0 d.f. = 2 P-value = 0.0782 66 67 68 B e n c h m a r k D o s e C o m p u t a t i o n 69 70 S p e c i f i e d e f f e c t = 0.1 This document is a draftfor review purposes only and does not constitute Agency policy. E-230 DRAFT--DO NOT CITE OR QUOTE 1 2 Risk Type = 3 4 Confidence level = 5 6 BMD = 7 8 BMDL = 9 10 Extra risk 0.95 6.4132 4.0249 11 E .2 .4 6 .3 . Figurefor Selected Model: Log-Logistic Log-Logistic Model with 0.95 Confidence Level 12 13 14 15 E .2.46.4. Outputfor Additional Model Presented: Log-Logistic, Unrestricted 16 Toth et al., 1979: Skin L esio n s 17 18 19 20 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 21 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 6 3 _ T o t h _ 1 9 7 9 _ S k i n L e s _ L o g L o g i s t i c _ U _ 1 . ( d ) 22 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 6 3 _ T o t h _ 1 9 7 9 _ S k i n L e s _ L o g L o g i s t i c _ U _ 1 . p l t 23 _ W e d F e b 10 1 4 : 4 7 : 5 4 2 0 1 0 24 25 26 T a b l e 2 27 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-231 DRAFT--DO NOT CITE OR QUOTE 1 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 2 3 P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))] 4 5 6 Dependent variable = DichEff 7 Independent variable = Dose 8 Slope pa r a m e t e r is not restricted 9 10 T o t a l n u m b e r o f o b s e r v a t i o n s = 4 11 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 12 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 13 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 16 17 18 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 19 20 21 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 22 background = 0 23 intercept = -1.87608 24 slope = 0.458888 25 26 27 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 28 29 ( * * * T h e m o d e l p a r a m e t e r ( s ) - b a c k g r o u n d 30 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 31 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 32 33 intercept slope 34 CO CO o co CO o 35 i n t e r c e p t 1 36 37 s l o p e 1 38 39 40 41 P a r a m e t e r E s t i m a t e s 42 43 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 44 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 45 b a c k g r o u n d 0* * * 46 intercept -1.94946 * * * 47 slope 0.4802 * * * 48 49 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 50 51 52 53 A n a l y s i s o f D e v i a n c e T a b l e 54 55 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 56 Full model -71.5177 4 57 F i t t e d m o d e l -71.8153 2 0.59526 2 0.7426 58 R e d u c e d m o d e l -95.8498 1 48.6642 3 <.0001 59 60 AIC: 147.631 61 62 63 G o o d n e s s o f F i t 64 S c a l e d 65 Dose Est. Prob. Expected Observed Size Residu. 66 67 0.0000 0.0000 0.000 0.000 38 0.000 68 0.5732 0.0983 4.323 5.000 44 0 . 3 4 3 69 14.2123 0.3374 14.845 13.000 44 - 0 . 5 8 8 70 91.2070 0.5542 23.832 25.000 43 0.358 This document is a draftfor review purposes only and does not constitute Agency policy. E-232 DRAFT--DO NOT CITE OR QUOTE 1 2 Chi^2 = 0.59 d.f. = 2 P-value = 0.7438 3 4 5 Benchmark Dose Computation 6 7 Specified effect : 0.1 8 9 Risk Type : Extra risk 10 11 C o n f i d e n c e l e v e l : 0.95 12 13 BMD ^ 0.596932 14 15 BMDL = 0.06773 16 17 18 E .2.46.5. Figurefor Additional Model Presented: Log-Logistic, Unrestricted Log-Logistic Model with 0.95 Confidence Level 19 14:47 02/10 2010 20 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-233 DRAFT--DO NOT CITE OR QUOTE 1 E .2.47. V an B irgelen et al., 1995a: H epatic R etinol 2 E .2 .4 7 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 4 <0.0001 159.735 7.790E+00 4.150E+00 Notes exponential (M 3) 4 <0.0001 3222.700 5.542E +01 error p o w er hit bound (d = 1) exponential (M4) b 3 <0.001 141.454 2.488E+01 3.363E+00 exponential (M 5) 3 < 0 .0 0 1 1 41.454 2.4 8 8 E + 0 1 3 .3 6 3 E + 0 0 p o w er hit bound (d = 1) H ill 3 0.239 124.865 5.316E +00 error n lo w er b oun d hit (n = 1) linear polynom ial, 5degree power 4 <0.0001 176.828 1.877E+02 1.437E+02 4 <0.0001 176.828 1.877E+02 1.437E+02 4 < 0 .0 0 0 1 1 76.828 1.877E + 02 1 .437E + 02 p o w er b oun d hit (p ow er = 1) H ill, unrestricted 2 0.241 125.495 3.595E +00 error unrestricted (n = 0.763) power, unrestricted c 3 0.011 131.771 3.802E-01 1.393E-02 unrestricted (power = 0.14) a N on-con stant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .4 7 .2 . Outputfor Selected Model: Exponential (M4) 6 Van B irgelen et al., 1995a: H epatic Retinol 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 6 5 _ V a n B _ 1 9 9 5 a _ H e p R e t _ E x p _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 M o n F e b 08 1 3 : 3 2 : 0 0 2 0 1 0 14 15 16 T b l 3 , h e p a t i c r e t i n o l 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; This document is a draftfor review purposes only and does not constitute Agency policy. E-234 DRAFT--DO NOT CITE OR QUOTE 1 sign = -1 for d e c r e a s i n g trend. 2 3 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 4 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 5 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 6 7 8 Dependent variable = Mean 9 Independent variable = Dose 10 D a t a a r e a s s u m e d t o b e d i s t r i b u t e d : n o r m a l l y 11 V a r i a n c e M o d e l : e x p ( l n a l p h a + r h o * l n ( Y [ d o s e ] ) ) 12 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 13 14 T o t a l n u m b e r o f d o s e g r o u p s = 6 15 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 16 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 17 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 18 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 19 20 M L E s o l u t i o n p r o v i d e d : E x a c t 21 22 23 I n i t i a l P a r a m e t e r V a l u e s 24 25 Variable Model 4 26 27 lnalpha -1.16065 28 rho 1.53688 29 a 1 5 . 6 4 5 30 b 0 . 0 2 5 4 3 5 1 31 c 0 . 0 3 6 5 2 4 7 32 d 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 Variable Model 4 39 40 lnalpha -0.92683 41 rho I. 77262 42 a I I . 5 0 4 9 43 b 0 . 0 2 8 6 5 9 8 44 c 0 . 0 6 5 3 0 4 3 45 d 1 46 47 48 T a b l e o f S t a t s F r o m I n p u t D a t a 49 50 Dose N Obs Mean Obs Std Dev 51 52 08 14.9 8.768 53 7.204 8 8.4 3.394 54 11.76 8 8.2 2.263 55 18.09 8 5.1 0.8485 56 86.41 8 2.2 0.8485 57 250.2 8 0.6 0.5657 58 59 60 E s t i m a t e d V a l u e s o f I n t e r e s t 61 62 Dose Est Mean Est Std Scaled Residual 63 64 0 11.5 5.483 1.751 65 7 . 2 0 4 9.499 4.627 -0.6719 66 1 1 . 7 6 8.428 4.161 -0.1552 67 1 8 . 0 9 7.154 3.599 -1.615 68 8 6 . 4 1 1.655 0.9832 1.568 69 250.2 0.7596 0.4931 -0.9155 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-235 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Other models for which likelihoods are calculated: 4 5 Model A1: Yij = Mu(i) + e(ij) 6 V a r { e i ;ij)} = S i g m a ^ 2 7 8 Model A2: Yij = Mu(i) + e(ij) 9 V a r { e i :ij)} = S i g m a ( i ) ^ 2 10 11 M o d e l A3: Yij = M u ( i ) + e(i j ) 12 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 13 14 M o d e l R: Yij = M u + e(i) 15 V a r { e i ij)} = S i g m a ^ 2 16 17 18 L i k e l i h o o d s o f I n t e r e s t 19 20 Model Log(likelihood) DF AIC 21 22 A1 -87.1567 7 188.3134 23 A2 -47.28742 12 11 8 . 5 7 4 8 24 A3 -55.32422 8 126.6484 25 R -109.967 2 223.934 26 4 -65.72714 5 141.4543 27 28 29 Additive constant for all log-likelihoods = -44.11. This constant added to the 30 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 31 d e p e n d o n t h e m o d e l p a r a m e t e r s . 32 33 34 E x p l a n a t i o n o f T e s t s 35 36 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 37 T e s t 2: A r e H o m o g e n e o u s ? (A2 vs. A1) 38 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 39 40 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 41 42 43 T e s t s o f I n t e r e s t 44 45 T e s t -2*log(Likelihood Ratio) D. F. p-value 46 47 T e s t 1 125.4 10 < 0.0001 48 T e s t 2 79.74 5 < 0.0001 49 T e s t 3 16.07 4 0.002922 50 T e s t 6a 20.81 3 0.0001155 51 52 53 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 54 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 55 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 56 57 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 58 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e . 59 60 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t to 61 c o n s i d e r a d i f f e r e n t v a r i a n c e m o d e l . 62 63 T h e p - v a l u e f o r T e s t 6a is l e s s t h a n .1. M o d e l 4 m a y n o t a d e q u a t e l y 64 d e s c r i b e t h e d a t a ; y o u m a y w a n t t o c o n s i d e r a n o t h e r m o d e l . 65 66 67 B e n c h m a r k D o s e C o m p u t a t i o n s : 68 69 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-236 DRAFT--DO NOT CITE OR QUOTE 1 Risk Type = Estimated standard deviations from control 2 3 Confidence Level = 0.950000 4 5 BMD = 24.8811 6 7 BMDL = 3.36281 8 9 10 E .2 .4 7 .3 . Figurefor Selected Model: Exponential (M4) Exponential Model 4 with 0.95 Confidence Level Mean Response 11 12 13 14 E .2.47.4. Outputfor Additional Model Presented: Power, Unrestricted 15 V an B irg elen et al., 1995a: H epatic R etinol 16 17 18 19 20 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 21 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 6 5 _ V a n B _ 1 9 9 5 a _ H e p R e t _ P w r _ U _ 1 . ( d ) 22 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 6 5 _ V a n B _ 1 9 9 5 a _ H e p R e t _ P w r _ U _ 1 . p l t 23 M o n F e b 08 1 3 : 3 2 : 0 3 2 0 1 0 24 25 26 T b l 3 , h e p a t i c r e t i n o l 27 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-237 DRAFT--DO NOT CITE OR QUOTE 1 The f o r m of the r e s p o n s e f u n c t i o n is: 2 3 Y[dose] = control + slope * doseApower 4 5 6 Dependent variable = Mean 7 Independent variable = Dose 8 The power is not restricted 9 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) * rho) 10 11 T o t a l n u m b e r o f d o s e g r o u p s = 6 12 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 13 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 14 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 16 17 18 19 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 20 lalpha = 2.76506 21 rho = 0 22 control = 14.9 23 slope = -3.98831 24 p o w e r = 25 26 27 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 28 29 lalpha rho control slope power 30 Lf) g o Lf) g o CO o CO o 31 l a l p h a 1 -0.042 0.038 0.063 32 33 rh o 1 -0.089 0.0044 -0.1 34 35 c o n t r o l -0.042 -0.089 1 -0.81 36 37 slope 0.038 0.0044 1 0.95 38 39 p o w e r 0 . 0 6 3 -0.1 -0.81 0.95 1 40 41 42 43 P a r a m e t e r E s t i m a t e s 44 45 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 46 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 47 lalpha -0.986251 0.394722 -1.75989 -0.212609 48 rho 1.67858 0.202896 1.28091 2.07625 49 control 16.9266 2.23237 12.5513 21.302 50 slope -7.51118 2.04379 -11.5169 -3.50543 51 power 0.139871 0.0269576 0.0870351 0.192707 52 53 54 55 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 56 57 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 58 59 60 08 14.9 16.9 8.77 6.56 -0.874 61 7 . 2 0 4 8 8.4 7.03 3.39 3.14 1.24 62 1 1 . 7 6 8 8.2 6.32 2.26 2.87 1.85 63 1 8 . 0 9 8 5.1 5.67 0.849 2.62 -0.611 64 8 6 . 4 1 8 2.2 2.91 0.849 1.5 -1.34 65 2 5 0 . 2 8 0.6 0.666 0.566 0.434 -0.427 66 67 68 69 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-238 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A1: Yij Mu(i) + e(ij) 3 Var{e(ij)} Sigma^2 4 5 Model A2: Yij Mu(i) + e(ij) 6 Var{e(ij)} Sigma(i)^2 7 8 Model A3: Yij = Mu(i) + e(ij) 9 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 10 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 11 w e r e s p e c i f i e d b y t h e u s e r 12 13 M o d e l R: Yi = M u + e(i) 14 V a r { e ( i ) } = S i g m a ^ 2 15 16 17 L i k e l i h o o d s o f I n t e r e s t 18 19 Model Log(likelihood) # Param's AIC 20 A1 -87.156698 7 188.313395 21 A2 -47.287416 12 1 1 8 . 5 7 4 8 3 3 22 A3 -55.324218 8 126.648436 23 fitted -60.885746 5 131.771493 24 R -109.967018 2 223.934036 25 26 27 E x p l a n a t i o n o f T e s t s 28 29 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 30 (A2 vs. R) 31 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 32 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 33 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 34 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 35 36 T e s t s o f I n t e r e s t 37 38 Test -2*log(Likelihood Ratio) Test df p-value 39 40 T e s t 1 125.359 10 <.0001 41 T e s t 2 79.7386 5 <.0001 42 T e s t 3 16.0736 4 0.002922 43 T e s t 4 11.1231 3 0.01108 44 45 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 46 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 47 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 48 49 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 50 m o d e l a p p e a r s t o b e a p p r o p r i a t e 51 52 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t t o c o n s i d e r a 53 d i f f e r e n t v a r i a n c e m o d e l 54 55 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 56 m o d e l 57 58 59 B e n c h m a r k D o s e C o m p u t a t i o n 60 61 S p e c i f i e d e f f e c t = 1 62 63 R i s k T y p e = Estimated standard deviations from the control mean 64 65 C o n f i d e n c e l e v e l = 0.95 66 67 B M D = 0 . 3 8 0 2 0 8 68 69 70 B M D L = 0 . 0 1 3 9 2 7 This document is a draftfor review purposes only and does not constitute Agency policy. E-239 DRAFT--DO NOT CITE OR QUOTE 1 E.2.47.5. F igure f o r A d d itio n a l M o d el Presented: Power, U nrestricted Power Model with 0.95 Confidence Level Mean Response 2 13:32 02/08 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-240 DRAFT--DO NOT CITE OR QUOTE 1 E.2.48. Van Birgelen et al., 1995a: Hepatic Retinol Palmitate 2 E.2.48.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 4 <0.0001 460.282 error error Notes exponential (M 3) exponential (M4) b exponential (M 5) 4 <0.0001 460.282 error error p o w er hit b ound (d = 1) 3 <0.0001 446.995 1.415E+02 3.647E+01 3 < 0 .0 0 0 1 4 4 6 .9 9 5 1 .415E + 02 3.6 4 7 E + 0 1 p o w er hit b ound (d = 1) H ill 3 0.009 416.233 3.657E +00 error n lo w er bound hit (n = 1) linear polynom ial, 5degree power 4 <0.0001 486.375 3.487E+02 2.412E+02 0 N /A 584.170 error 5 .6 1 7 E + 0 2 4 < 0 .0 0 0 1 4 8 6 .3 7 5 3 .4 8 7 E + 0 2 2 .4 1 2 E + 0 2 p o w er b oun d hit (p ow er = 1) H ill, unrestricted 3 <0.0001 527.310 6.875E-14 6.875E -14 unrestricted (n = 0.613) power, unrestricted c 3 0.239 408.982 5.262E-02 5.889E-05 unrestricted (pow er = 0.064) a N on-con stant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .4 8 .2 . Outputfor Selected Model: Exponential (M4) 6 Van B irgelen et al., 1995a: H epatic R etinol Palmitate 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a Fil e : C : \ 1 \ B l o o d \ 6 6 V a n B 1 9 9 5 a H e p R e t P a l m E x p 1.( 12 G n u p l o t P l o t t i n g Fil e : 13 M o n F e b 08 1 3 : 3 2 : 14 15 16 T b l 3 , h e p a t i c r e t i n o l p a l m i t a t e 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; This document is a draftfor review purposes only and does not constitute Agency policy. E-241 DRAFT--DO NOT CITE OR QUOTE 1 sign = -1 for d e c r e a s i n g trend. 2 3 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 4 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 5 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 6 7 8 Dependent variable = Mean 9 Independent variable = Dose 10 D a t a a r e a s s u m e d t o b e d i s t r i b u t e d : n o r m a l l y 11 V a r i a n c e M o d e l : e x p ( l n a l p h a + r h o * l n ( Y [ d o s e ] ) ) 12 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 13 14 T o t a l n u m b e r o f d o s e g r o u p s = 6 15 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 16 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 17 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 18 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 19 20 M L E s o l u t i o n p r o v i d e d : E x a c t 21 22 23 I n i t i a l P a r a m e t e r V a l u e s 24 25 Variable Model 4 26 27 lnalpha 0.284674 28 rho 1.77158 29 a 4 9 5 . 6 30 b 0 . 0 3 3 7 8 2 6 31 c 0 . 0 0 5 7 6 5 0 2 32 d 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 Variable Model 4 39 40 lnalpha -0.241601 41 rho 2.03456 42 a 2 2 3 . 8 4 8 43 b 0 . 0 3 0 0 7 3 7 44 c 0 . 0 1 2 9 2 5 3 45 d 1 46 47 N C = N o C o n v e r g e n c e 48 49 50 T a b l e o f S t a t s F r o m I n p u t D a t a 51 52 Dose N Obs Mean Obs Std Dev 53 54 08 472 271.5 55 7.204 8 94 6 7 . 8 8 56 11.76 8 107 76.37 57 18.09 8 74 39.6 58 86.41 8 22 22.63 59 250.2 8 3 2.828 60 61 62 E s t i m a t e d V a l u e s o f I n t e r e s t 63 64 Dose Est Mean Est Std Scaled Residual 65 66 0 223.8 217.8 3.222 67 7 . 2 0 4 180.8 175.3 -1.401 68 1 1 . 7 6 158 152.9 -0.9443 69 1 8 . 0 9 131.1 126.4 -1.278 70 8 6 . 4 1 19.33 18.03 0.4197 This document is a draftfor review purposes only and does not constitute Agency policy. E-242 DRAFT--DO NOT CITE OR QUOTE 1 250.2 3.013 2.721 -0.01317 2 3 4 5 Other models for which likelihoods are calculated: 6 7 Model A1: Yij = Mu(i) + e(ij) 8 V a r { e i ;ij)} = S i g m a ^ 2 9 10 M o d e l A2: Yij = M u ( i ) + e(i j ) 11 V a r { e i :ij)} = S i g m a ( i ) ^ 2 12 13 M o d e l A3: Yij = M u ( i ) + e(i j ) 14 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 15 16 M o d e l R: Yij = M u + e(i) 17 V a r { e i ij)} = S i g m a ^ 2 18 19 20 L i k e l i h o o d s o f I n t e r e s t 21 22 Model Log(likelihood) DF AIC 23 24 A1 -250.5548 7 515.1096 25 A2 -196.7557 12 41 7 . 5 1 1 5 26 A3 -197.3832 8 410.7663 27 R -276.7896 2 557.5793 28 4 -218.4977 5 446.9954 29 30 31 Additive constant for all log-likelihoods = -44.11. This constant added to the 32 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 33 d e p e n d o n t h e m o d e l p a r a m e t e r s . 34 35 36 E x p l a n a t i o n o f T e s t s 37 38 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 39 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 40 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs . A3) 41 42 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 vs 4) 43 44 45 T e s t s o f I n t e r e s t 46 47 T e s t -2*log(Likelihood Ratio) D. F. p-value 48 49 T e s t 1 160.1 10 < 0.0001 50 T e s t 2 107.6 5 < 0.0001 51 T e s t 3 1.255 4 0.869 52 T e s t 6a 42.23 3 < 0.0001 53 54 55 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s to b e a 56 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 57 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 58 59 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 60 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e . 61 62 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 63 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 64 65 T h e p - v a l u e f o r T e s t 6a is l e s s t h a n .1. M o d e l 4 m a y n o t a d e q u a t e l y 66 d e s c r i b e t h e d a t a ; y o u m a y w a n t t o c o n s i d e r a n o t h e r m o d e l . 67 68 69 B e n c h m a r k D o s e C o m p u t a t i o n s : 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-243 DRAFT--DO NOT CITE OR QUOTE 1 Specified Effect 1.000000 2 3 Risk Type = Estimated standard deviations from control 4 5 Confidence Level = 0.950000 6 7 BMD = 141.528 8 9 BMDL = 36.4721 10 11 12 E .2 .4 8 .3 . Figurefor Selected Model: Exponential (M4) Exponential Model 4 with 0.95 Confidence Level Mean Response 13 14 15 16 E .2.48.4. Outputfor Additional Model Presented: Power, Unrestricted 17 V an B irg elen et al., 1995a: H epatic R etin ol Palm itate 18 19 20 21 22 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 23 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 6 6 _ V a n B _ 1 9 9 5 a _ H e p R e t P a l m _ P w r _ U _ 1 . ( d ) 24 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 6 6 _ V a n B _ 1 9 9 5 a _ H e p R e t P a l m _ P w r _ U _ 1 . p l t 25 _ M o n F e b 08 1 3 : 3 2 7 4 7 2 0 1 0 26 27 28 T b l 3 , h e p a t i c r e t i n o l p a l m i t a t e This document is a draftfor review purposes only and does not constitute Agency policy. E-244 DRAFT--DO NOT CITE OR QUOTE 1 2 3 The f o r m of the r e s p o n s e f u n c t i o n is: 4 5 Y[dose] = control + slope * doseApower 6 7 8 Dependent variable = Mean 9 Independent variable = Dose 10 T h e p o w e r is n o t r e s t r i c t e d 11 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 12 13 T o t a l n u m b e r o f d o s e g r o u p s = 6 14 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 15 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 16 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 17 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 18 19 20 21 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 22 lalpha = 9.57332 23 rho = 0 24 control = 472 25 slope = -320.514 26 p o w e r = 0 . 0 7 1 1 1 7 3 27 28 29 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 30 31 lalpha rho control slope power 32 33 l a l p h a 1 -0.95 0.3 -0.31 -0.3 34 35 rho -0.95 1 -0.41 0.39 0.29 36 37 c o n t r o l 0.3 -0.41 1 -0.98 -0.82 38 39 slope -0.31 0.39 -0.98 1 0.9 40 41 p o w e r -0.3 0.29 -0.82 0.9 1 42 43 44 45 P a r a m e t e r E s t i m a t e s 46 47 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 48 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 49 lalpha 0.0640168 0.859472 -1.62052 1.74855 50 rho 1.81132 0.197468 1.42429 2.19835 51 control 464.29 87.5705 292.655 635.925 52 slope -324.216 83.3327 -487.545 -160.887 53 power 0.0639088 0.0139778 0.0365129 0.0913048 54 55 56 57 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 58 59 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 60 61 62 0 8 472 464 272 269 0.0812 63 7 . 2 0 4 8 94 96.5 67.9 64.7 -0.108 64 1 1 . 7 6 8 107 84.8 76.4 57.6 1.09 65 1 8 . 0 9 8 74 74.2 39.6 51 -0 . 0 0 9 4 1 66 8 6 . 4 1 8 22 33.2 22.6 24.6 -1.28 67 2 5 0 . 2 8 3 2.86 2.83 2.68 0.145 68 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-245 DRAFT--DO NOT CITE OR QUOTE 1 Model Descriptions for likelihoods calculated 2 3 4 Model A1: Yij Mu(i) + e(ij) 5 Var{e(ij)} Sigma^2 6 7 Model A2: Yij Mu(i) + e(ij) 8 Var{e(ij)} Sigma(i)^2 9 10 M o d e l A3: Yij = Mu(i) + e(ij) 11 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 12 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 13 w e r e s p e c i f i e d b y t h e u s e r 14 15 M o d e l R: Yi = M u + e(i) 16 V a r { e ( i ) } = S i g m a ^ 2 17 18 19 L i k e l i h o o d s o f I n t e r e s t 20 21 Model Log(likelihood) # Param's AIC 22 A1 -250.554817 7 515.109634 23 A2 -196.755746 12 4 1 7 . 5 1 1 4 9 1 24 A3 -197.383174 8 410.766347 25 fitted -199.490808 5 408.981615 26 R -276.789644 2 557.579287 27 28 29 E x p l a n a t i o n o f T e s t s 30 31 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 32 (A2 vs. R) 33 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 34 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 35 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 36 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 37 38 T e s t s o f I n t e r e s t 39 40 Test -2*log(Likelihood Ratio) Test df p-value 41 42 T e s t 1 160.068 10 <.0001 43 T e s t 2 107.598 5 <.0001 44 T e s t 3 1.25486 4 0.869 45 T e s t 4 4.21527 3 0.2391 46 47 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 48 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 49 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 50 51 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 52 m o d e l a p p e a r s t o b e a p p r o p r i a t e 53 54 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 55 t o b e a p p r o p r i a t e h e r e 56 57 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 58 t o a d e q u a t e l y d e s c r i b e t h e d a t a 59 60 61 B e n c h m a r k D o s e C o m p u t a t i o n 62 63 S p e c i f i e d e f f e c t = 1 64 65 R i s k T y p e = Estimated standard deviations from the control mean 66 67 C o n f i d e n c e l e v e l = 0.95 68 69 B M D = 0 . 0 5 2 6 2 4 7 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-246 DRAFT--DO NOT CITE OR QUOTE 1 2 BMDL = 5.88883e-005 3 4 5 E .2.48.5. Figurefor Additional Model Presented: Power, Unrestricted Power Model with 0.95 Confidence Level Mean Response 700 600 500 400 300 200 100 0 0 6 13:32 02/08 2010 7 50 100 150 200 dose 250 This document is a draftfor review purposes only and does not constitute Agency policy. E-247 DRAFT--DO NOT CITE OR QUOTE 1 E.2.49. W hite et al., 1986: C H 50 2 E .2 .4 9 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg) BMDL (ng/kg) exponential (M 2) 5 0.002 389.664 1.957E+01 1.261E+01 Notes exponential (M 3) 5 0 .0 0 2 3 8 9 .6 6 4 1.957E +01 1.261E +01 p o w er hit bound (d = 1) exponential (M 4) 4 0.001 390.632 1.411E+01 5.177E+00 exponential (M 5) Hill b 4 0.001 3 9 0 .6 3 2 1.411E +01 5 .1 7 7 E + 0 0 p o w er hit bound (d = 1) 4 0.002 389.601 8.632E+00 1.498E+00 n lower bound hit (n = 1) linear polynom ial, 6degree power 5 <0.001 394.446 3.497E+01 2.568E+01 5 <0.001 394.446 3.497E+01 2.568E+01 5 < 0 .001 3 9 4 .4 4 6 3.4 9 7 E + 0 1 2 .5 6 8 E + 0 1 p o w er b oun d hit (p ow er = 1) H ill, unrestricted c 3 0.071 381.520 1.481E-01 4.351E-03 unrestricted (n = 0.246) power, unrestricted 4 0.148 379.265 1.211E-01 1.225E-03 unrestricted (pow er = 0.227) a N on-constant variance m odel selected (p = 0 .0871) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .2 .4 9 .2 . Outputfor Selected Model: Hill 6 W hite et al., 1986: CH 50 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 7 1 _ W h i t e _ 1 9 8 6 _ C H 5 0 _ H i l l _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 7 1 _ W h i t e _ 1 9 8 6 _ C H 5 0 _ H i l l _ 1 . p l t 13 M o n F e b 08 1 3 : 3 5 : 5 6 2 0 1 0 14 15 16 [ i n s e r t s t u d y n o t e s ] 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 P o w e r p a r a m e t e r r e s t r i c t e d t o b e g r e a t e r t h a n 1 This document is a draftfor review purposes only and does not constitute Agency policy. E-248 DRAFT--DO NOT CITE OR QUOTE 1 The variance is to be modeled as Var(i) = exp(lalpha + rho ln(mean(i))) 2 3 Total number of dose groups = 7 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 12 lalpha 5.60999 13 rho 0 14 intercept 91 15 -74 16 0 . 1 1 8 0 3 6 17 1 . 0 9 4 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( * * * T h e m o d e l p a r a m e t e r ( s ) -n 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r 24 a n d d o n o t a p p e a r in t h e c o r r e l a t i o n m a t r i x ) 25 26 lalpha rho intercept v k 27 28 l a l p h a 1 -0.99 0.27 0.23 -0.32 29 30 rho -0.99 1 -0.28 -0.24 0.33 31 32 i n t e r c e p t 0.27 -0.28 1 0.39 -0.78 33 Lf) CO o 34 v 0.23 -0.24 0.39 1 35 36 k -0.32 0.33 -0.78 -0.85 1 37 38 39 40 P a r a m e t e r E s t i m a t e s 41 42 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 43 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 44 lalpha 4.581 1.66273 1.32211 7.83989 45 rho 0.31293 0.431616 -0.533022 1.15888 46 i n t e r c e p t 74.6365 6.33673 62.2167 87.0562 47 v -66.2096 14.7876 -95.1928 -37.2264 48 n 1 N A 49 k 20.8286 21.3237 -20.965 62.6223 50 51 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 52 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 53 h a s n o s t a n d a r d e r r o r . 54 55 56 57 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 58 59 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 60 61 62 0 8 91 7 4 . 6 1 4 . 1 19. 4 2 .39 63 1 . 0 9 4 8 54 71.3 8.49 19. 3 -2 .54 64 4 . 0 8 5 8 63 63.8 11.3 18. 9 -0. 117 65 7 . 1 4 8 56 57.7 25.5 18. 6 -0. 263 66 2 6 . 8 1 8 41 37.4 17 17. 4 0. 5 8 9 67 4 8 . 7 2 8 32 28.3 17 16. 7 0. 6 3 6 68 9 0 . 5 6 8 17 20.8 17 15. 9 -0. 678 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-249 DRAFT--DO NOT CITE OR QUOTE 1 2 Model Descriptions for likelihoods calculated 3 4 5 Model A1: Yij Mu(i) + e(ij) 6 Var{e(ij)} Sigma^2 7 8 Model A2: Yij Mu(i) + e(ij) 9 Var{e(ij)} Sigma(i)^2 10 11 M o d e l A3: Yij = Mu(i) + e(ij) 12 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 13 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 14 w e r e s p e c i f i e d b y t h e u s e r 15 16 M o d e l R: Yi = M u + e(i) 17 V a r { e ( i ) } = S i g m a ^ 2 18 19 20 L i k e l i h o o d s o f I n t e r e s t 21 22 Model Log(likelihood) # Param's AIC 23 A1 -181.340979 8 378.681959 24 A2 -175.820265 14 3 7 9 . 6 4 0 5 2 9 25 A3 -181.238690 9 380.477380 26 fitted -189.800288 5 389.600575 27 R -212.367055 2 428.734109 28 29 30 E x p l a n a t i o n o f T e s t s 31 32 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 33 (A2 vs. R) 34 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 35 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 36 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 37 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 38 39 T e s t s o f I n t e r e s t 40 41 Test -2*log(Likelihood Ratio) Test df p-value 42 43 T e s t 1 73.0936 12 <.0001 44 T e s t 2 11.0414 6 0.0871 45 T e s t 3 10.8369 5 0.05471 46 T e s t 4 17.1232 4 0.001829 47 48 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 49 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 50 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 51 52 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 53 m o d e l a p p e a r s t o b e a p p r o p r i a t e 54 55 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t t o c o n s i d e r a 56 d i f f e r e n t v a r i a n c e m o d e l 57 58 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 59 m o d e l 60 61 62 B e n c h m a r k D o s e C o m p u t a t i o n 63 64 S p e c i f i e d e f f e c t = 1 65 66 R i s k T y p e = Estimated standard deviations from the control mean 67 68 C o n f i d e n c e l e v e l = 0.95 69 70 BMD = 8.63239 This document is a draftfor review purposes only and does not constitute Agency policy. E-250 DRAFT--DO NOT CITE OR QUOTE 1 2 BMDL 1.49823 3 4 5 E .2 .4 9 .3 . Figurefor Selected Model: Hill Hill Model with 0.95 Confidence Level Mean Response 6 7 8 9 E .2.49.4. Outputfor Additional Model Presented: Hill, Unrestricted 10 W h ite et al., 1986: C H 50 11 12 13 14 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 15 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 7 1 _ W h i t e _ 1 9 8 6 _ C H 5 0 _ H i l l _ U _ 1 . ( d ) 16 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 7 1 _ W h i t e _ 1 9 8 6 _ C H 5 0 _ H i l l _ U _ 1 . p l t 17 M o n F e b 08 1 3 : 3 5 : 5 7 2 0 1 0 18 19 20 [ i n s e r t s t u d y n o t e s ] 21 22 23 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 24 25 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 26 27 28 D e p e n d e n t v a r i a b l e = M e a n This document is a draftfor review purposes only and does not constitute Agency policy. E-251 DRAFT--DO NOT CITE OR QUOTE 1 Independent variable = Dose 2 Power pa r a m e t e r is not restricted 3 The v ariance is to be mode l e d as Var(i) = exp(lalpha + rho * ln(mean(i))) 4 5 Total number of dose groups = 7 6 Total number of records with missing values = 0 7 Maximum number of iterations = 250 8 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 9 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 10 11 12 13 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 14 lalpha 5.60999 15 rho 0 16 intercept 91 17 -74 18 0 . 1 1 8 0 3 6 19 1 . 0 9 4 20 21 22 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 23 24 lalpha rho intercept v n k 25 26 l a l p h a 1 -1 0.16 0.19 -0.4 -0.014 27 CO LO g CO oo Lf) CO o g CO '-- 1 g oo CO Lf) o '-- 1 o 28 rh o -1 1 -0.16 0.4 0.011 29 30 i n t e r c e p t 0.16 -0.16 1 0.15 -0.58 0.015 31 32 v 0 . 1 9 0.15 1 -0.02 33 34 n -0.4 0.4 -0.02 1 35 36 k -0.014 0.011 0.015 1 37 38 39 40 P a r a m e t e r E s t i m a t e s 41 42 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 43 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 44 lalpha 6.54093 2.08879 2.44698 10.6349 45 rho -0.245847 0.541645 -1.30745 0.815757 46 ntercept 89.6302 5.59428 78.6656 100.595 47 v -628.486 727.973 -2055.29 798.315 48 n 0.246409 0.058636 0.131484 0.361333 49 k 493877 2.74838e+006 -4.89284e+006 5.88059e+006 50 51 52 53 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 54 55 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 56 57 58 0 8 91 8 9 . 6 1 4 . 1 1 5 . 1 0.256 59 1 . 0 9 4 8 54 65.2 8.49 15.8 -2.01 60 4 . 0 8 5 8 63 56.3 11.3 16 1.17 61 7 . 1 4 8 56 51.7 25.5 16.2 0.746 62 2 6 . 8 1 8 41 38.3 17 16.8 0.453 63 4 8 . 7 2 8 32 30.9 17 17.3 0.175 64 9 0 . 5 6 8 17 22.3 17 18 -0.831 65 66 67 68 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-252 DRAFT--DO NOT CITE OR QUOTE 1 Model A1: Yij Mu(i) + e(ij) 2 Var{e(ij)} Sigma^2 3 4 Model A2: Yij Mu(i) + e(ij) 5 Var{e(ij)} Sigma(i)^2 6 7 Model A3: Yij = Mu(i) + e(ij) 8 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 9 Model A3 uses any fixed variance parameters that 10 w e r e s p e c i f i e d b y t h e u s e r 11 12 M o d e l R: Yi = M u + e(i) 13 V a r { e ( i ) } = S i g m a ^ 2 14 15 16 L i k e l i h o o d s o f I n t e r e s t 17 18 Model Log(likelihood) # Param's AIC 19 A1 -181.340979 8 378.681959 20 A2 -175.820265 14 3 7 9 . 6 4 0 5 2 9 21 A3 -181.238690 9 380.477380 22 fitted -184.759769 6 381.519538 23 R -212.367055 2 428.734109 24 25 26 E x p l a n a t i o n o f T e s t s 27 28 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 29 (A2 vs. R) 30 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 31 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 32 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 33 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 34 35 T e s t s o f I n t e r e s t 36 37 Test -2*log(Likelihood Ratio) Test df p-value 38 39 T e s t 1 73.0936 12 <.0001 40 T e s t 2 11.0414 6 0.0871 41 T e s t 3 10.8369 5 0.05471 42 T e s t 4 7.04216 3 0.07057 43 44 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 45 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 46 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 47 48 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 49 m o d e l a p p e a r s t o b e a p p r o p r i a t e 50 51 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t t o c o n s i d e r a 52 d i f f e r e n t v a r i a n c e m o d e l 53 54 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 55 m o d e l 56 57 58 B e n c h m a r k D o s e C o m p u t a t i o n 59 60 S p e c i f i e d e f f e c t : 1 61 62 R i s k T y p e Estimated standard deviations from the control mean 63 64 C o n f i d e n c e l e v e l : 0.95 65 66 BMD 0.148074 67 68 BMDL = 0.00435112 69 This document is a draftfor review purposes only and does not constitute Agency policy. E-253 DRAFT--DO NOT CITE OR QUOTE 1 E.2.49.5. F igure f o r A d d itio n a l M o d el Presented: H ill, U nrestricted Hill Model with 0.95 Confidence Level Mean Response 2 13:35 02/08 2010 3 4 5 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-254 DRAFT--DO NOT CITE OR QUOTE 1 E.3. A D M IN IST E R E D D O SE BM D S R E SU L T S 2 E.3.1. A m in et al., 2000: 0.25% Saccharin C onsum ed, Fem ale 3 E .3 .1 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes linear b 1 0.358 179.702 8.816E+01 5.890E+01 polynom ial, 2degree 1 0.358 179.702 8.816E+01 5.890E+01 power 1 0.358 179.702 8 .816E + 01 5 .890E + 01 p ow er b oun d hit (p ow er = 1) power, unrestricted c 0 N /A 180.858 7.530E+01 2.537E+01 unrestricted (power = 0.605) a N on-con stant variance m odel selected (p = 0 .0 0 0 5 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 4 5 6 E .3 .1 .2 . Outputfor Selected Model: Linear 7 A m in et al., 2000: 0.25% Saccharin Consum ed, Fem ale 8 9 10 11 P o l y n o m i a l M o d e l . ( V e r s i o n : 2 . 1 3 ; D a t e : 0 4 / 0 8 / 2 0 0 8 ) 12 I n p u t D a t a F i l e : C : \ 1 \ 1 _ A m i n _ 2 0 0 0 _ 2 5 _ S C _ L i n e a r _ 1 . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 1 _ A m i n _ 2 0 0 0 _ 2 5 _ S C _ L i n e a r _ 1 . p l t 14 ~ T u e _ F e b 1 6 ~ 1 7 : 2 2 : 1 6 2 0 1 0 15 16 17 18 19 20 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 21 22 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 23 24 25 D e p e n d e n t v a r i a b l e = M e a n 26 I n d e p e n d e n t v a r i a b l e = D o s e 27 S i g n s o f t h e p o l y n o m i a l c o e f f i c i e n t s a r e n o t r e s t r i c t e d 28 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 29 30 T o t a l n u m b e r o f d o s e g r o u p s = 3 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 33 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 34 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 35 36 37 38 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 39 lalpha = 5.29482 40 rho = 0 41 beta_0 = 30.8266 42 beta_1 = -0.204134 43 This document is a draftfor review purposes only and does not constitute Agency policy. E-255 DRAFT--DO NOT CITE OR QUOTE 1 2 Asymptotic Correlation Matrix of Parameter Estimates 3 4 lalpha rho beta_0 beta_1 5 6 lalpha 1 -0.99 -0.016 0.03 7 8 rho -0.99 1 0.013 -0.026 9 10 b e t a _ 0 - 0 . 0 1 6 0.013 1 -0.94 11 12 b e t a _ 1 0.03 -0.026 -0.94 1 13 14 15 16 P a r a m e t e r E s t i m a t e s 17 18 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 19 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 20 lalpha -2.55843 1.66185 -5.8156 0.698746 21 rho 2.42056 0.545617 1.35117 3.48995 22 beta_0 30.3968 4.03582 22.4868 38.3069 23 beta 1 -0.196699 0.0443352 -0.283594 -0.109803 24 25 26 27 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 28 29 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 30 31 32 0 10 31.7 30.4 20.6 17.3 0.233 33 25 10 24.6 25.5 12 14 -0.2 34 100 10 10.7 10.7 5.33 4.92 -0.0204 35 36 37 38 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 39 40 41 M o d e l A1: Yij Mu(i) + e(ij) 42 Var{e(ij)} Sigma^2 43 44 M o d e l A2: Yij Mu(i) + e(ij) 45 Var{e(ij)} Sigma(i)^2 46 47 M o d e l A3: Yij = Mu(i) + e(ij) 48 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 49 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 50 w e r e s p e c i f i e d b y t h e u s e r 51 52 M o d e l R: Yi = M u + e(i) 53 V a r { e ( i ) } = S i g m a ^ 2 54 55 56 L i k e l i h o o d s o f I n t e r e s t 57 58 Model Log(likelihood) # Param's AIC 59 A1 -92.841935 4 193.683870 60 A2 -85.255316 6 182.510632 61 A3 -85.429148 5 180.858295 62 fitted -85.851107 4 179.702213 63 R -98.136607 2 200.273213 64 65 66 E x p l a n a t i o n o f T e s t s 67 68 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 69 (A2 vs. R) 70 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) This document is a draftfor review purposes only and does not constitute Agency policy. E-256 DRAFT--DO NOT CITE OR QUOTE 1 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 2 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 3 (Note: When rho=0 the results of Test 3 and Test 2 will be the same.) 4 5 Tests of Interest 6 7 Test -2*log(Likelihood Ratio) Test df p-value 8 9 Test 1 25.7626 4 <.0001 10 T e s t 2 15.1732 2 0.0005072 11 T e s t 3 0.347663 1 0.5554 12 T e s t 4 0.843918 1 0.3583 13 14 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 15 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 16 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 17 18 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 19 m o d e l a p p e a r s t o b e a p p r o p r i a t e 20 21 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 22 t o b e a p p r o p r i a t e h e r e 23 24 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 25 t o a d e q u a t e l y d e s c r i b e t h e d a t a 26 27 28 B e n c h m a r k D o s e C o m p u t a t i o n 29 30 S p e c i f i e d e f f e c t = 1 31 32 R i s k T y p e = Estimated standard deviations from the control mean 33 34 C o n f i d e n c e l e v e l = 0.95 35 36 BMD = 88.1623 37 38 39 BMDL = 58.9029 40 This document is a draftfor review purposes only and does not constitute Agency policy. E-257 DRAFT--DO NOT CITE OR QUOTE 1 E.3.1.3. F igure f o r Selected M odel: L in ea r Linear Model with 0.95 Confidence Level 2 17:22 02/16 2010 3 4 5 E .3.1.4. Outputfor Additional Model Presented: Power, Unrestricted 6 A m in et al., 2000: 0.25% Saccharin Consum ed, Fem ale 7 8 9 10 11 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 12 I n p u t D a t a Fil e : C : \ 1 \ 1 A m i n 2 0 0 0 25 S C P w r U 1 .(d) 13 G n u p l o t P l o t t i n g Fil e : C : \ 1 \ 1 A m i n 2 0 0 0 25 SC P w r U 1 . p l t 14 T u e F e b 16 17 :22 :17 2 0 1 0 15 16 17 18 19 20 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 21 22 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 23 24 25 D e p e n d e n t v a r i a b l e = M e a n 26 I n d e p e n d e n t v a r i a b l e = D o s e 27 T h e p o w e r is n o t r e s t r i c t e d 28 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) This document is a draftfor review purposes only and does not constitute Agency policy. E-258 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 3 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 lalpha = 5.29482 12 rho = 0 13 control = 31.6727 14 slope = -0.567889 15 power = 0.783745 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 lalpha rho control slope power 21 22 l a l p h a 1 -0.99 0.34 -0.14 -0.061 23 24 rho -0.99 1 -0.42 0.15 0.068 25 Lf) o Lf) o 26 c o n t r o l 0.34 -0.42 1 -0.67 27 28 slope -0.14 0.15 -0.67 1 0.99 29 30 power -0.061 0.068 0.99 1 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 37 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 38 lalpha -2.48291 2.08669 -6.57274 1.60693 39 rho 2.38455 0.692047 1.02817 3.74094 40 c o n t r o l 32.99 5.40754 22.3914 43.5886 41 slope -1.36469 2.01258 -5.30927 2.5799 42 power 0.605364 0.288476 0.0399625 1.17077 43 44 45 46 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 47 48 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 49 50 51 0 10 31.7 33 20.6 18.7 -0.223 52 25 10 24.6 23.4 12 12.4 0.302 CO o o 53 100 10 10.7 10.8 5.33 4.94 54 55 W a r n i n g : L i k e l i h o o d f o r f i t t e d m o d e l l a r g e r t h a n t h e L i k e l i h o o d f o r m o d e l A3. 56 57 58 59 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 60 61 62 M o d e l A1: Yij Mu(i) + e(ij) 63 Var{e(ij)} Sigma^2 64 65 M o d e l A2: Yij Mu(i) + e(ij) 66 Var{e(ij)} Sigma(i)^2 67 68 M o d e l A3: Yij = Mu(i) + e(ij) 69 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 70 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t This document is a draftfor review purposes only and does not constitute Agency policy. E-259 DRAFT--DO NOT CITE OR QUOTE 1 were specified by the user 2 3 M o d e l R: Yi = M u + e(i) 4 Var{e(i)} = Sigma^2 5 6 7 Likelihoods of Interest 8 9 Model Log(likelihood) # Param's AIC 10 A1 -92.841935 4 193.683870 11 A2 -85.255316 6 182.510632 12 A3 -85.429148 5 180.858295 13 fitted -85.429148 5 180.858295 14 R -98.136607 2 200.273213 15 16 17 E x p l a n a t i o n o f T e s t s 18 19 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 20 (A2 vs. R) 21 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 22 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 23 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 24 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 25 26 T e s t s o f I n t e r e s t 27 28 Test -2*log(Likelihood Ratio) Test df p-value 29 30 T e s t 1 25.7626 4 <.0001 31 T e s t 2 15.1732 2 0.0005072 32 T e s t 3 0.347663 1 0.5554 33 T e s t 4 -8.2423e-013 0 NA 34 35 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 36 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 37 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 38 39 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 40 m o d e l a p p e a r s t o b e a p p r o p r i a t e 41 42 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 43 t o b e a p p r o p r i a t e h e r e 44 45 N A - D e g r e e s o f f r e e d o m f o r T e s t 4 a r e l e s s t h a n o r e q u a l t o 0. T h e C h i - S q u a r e 46 t e s t f o r f i t is n o t v a l i d 47 48 49 B e n c h m a r k D o s e C o m p u t a t i o n 50 51 S p e c i f i e d e f f e c t = 1 52 53 R i s k T y p e = Estimated standard deviations from the control mean 54 55 C o n f i d e n c e l e v e l = 0.95 56 57 B M D = 7 5 . 2 9 9 4 58 59 60 B M D L = 2 5 . 3 7 1 7 61 This document is a draftfor review purposes only and does not constitute Agency policy. E-260 DRAFT--DO NOT CITE OR QUOTE 1 E.3.1.5. F igure f o r A d d itio n a l M o d el Presented: Power, Unrestricted Power Model with 0.95 Confidence Level Mean Response 2 17:22 02/16 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-261 DRAFT--DO NOT CITE OR QUOTE 1 E.3.2. Amin et al., 2000: 0.25% Saccharin Preference Ratio, Female 2 E.3.2.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes linear b 1 0.002 228.094 1.264E+02 6.128E+01 polynom ial, 2degree 1 0.002 228.094 1.264E+02 6.128E+01 power 1 0 .0 0 2 2 2 8 .0 9 4 1.264E + 02 6 .1 2 8 E + 0 1 p o w er b oun d hit (p ow er = 1) a N on-con stant variance m odel selected (p = 0 .0 1 3 5 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3 .2 .2 . Outputfor Selected Model: Linear 6 A m in et al., 2000: 0.25% Saccharin Preference Ratio, Fem ale 7 8 9 10 P o l y n o m i a l M o d e l . ( V e r s i o n : 2 . 1 3 ; D a t e : 0 4 / 0 8 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 2 _ A m i n _ 2 0 0 0 _ 2 5 _ S P _ L i n e a r _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 2 _ A m i n _ 2 0 0 0 _ 2 5 _ S P _ L i n e a r _ 1 . p l t 13 ~ T u e _ F e b 1 6 ~ 1 7 : 2 2 : 4 4 2 0 1 0 14 15 16 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 S i g n s o f t h e p o l y n o m i a l c o e f f i c i e n t s a r e n o t r e s t r i c t e d 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 28 29 T o t a l n u m b e r o f d o s e g r o u p s = 3 30 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 31 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 32 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 33 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 34 35 36 37 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 38 lalpha = 6.34368 39 rho = 0 40 beta_0 = 74.2008 41 beta 1 = -0.219781 42 43 44 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 45 46 lalpha rho beta_0 beta_1 47 O K> CO 48 l a l p h a 1 -1 0.2 This document is a draftfor review purposes only and does not constitute Agency policy. E-262 DRAFT--DO NOT CITE OR QUOTE 1 g '-- 1 o g '-- 1 o 2 rho -1 1 0.28 3 4 beta_0 0.2 1 -0.76 5 6 beta_1 o K> CO 0.28 -0.76 1 7 8 9 10 P a r a m e t e r E s t i m a t e s 11 12 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 13 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 14 lalpha 0.338774 9.23768 -17.7667 18.4443 15 rho 1.43998 2.21674 -2.90476 5.78472 16 beta_0 73.6633 6.6623 60.6054 86.7211 17 beta 1 -0.207175 0.101074 -0.405276 -0.00907442 18 19 20 21 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 22 23 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 24 25 26 0 10 82.1 73.7 13.3 26.2 1.02 27 25 10 58.1 68.5 33.9 24.8 -1.32 28 100 10 54.9 52.9 19.5 20.6 0.295 29 30 31 32 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 33 34 35 M o d e l A1: Yij Mu(i) + e(ij) 36 Var{e(ij)} Sigma^2 37 38 M o d e l A2: Yij Mu(i) + e(ij) 39 Var{e(ij)} Sigma(i)^2 40 41 M o d e l A3: Yij = Mu(i) + e(ij) 42 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 43 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 44 w e r e s p e c i f i e d b y t h e u s e r 45 46 M o d e l R: Yi = M u + e(i) 47 V a r { e ( i ) } = S i g m a ^ 2 48 49 50 L i k e l i h o o d s o f I n t e r e s t 51 52 Model Log(likelihood) # Param's AIC 53 A1 -108.574798 4 225.149597 54 A2 -104.269377 6 220.538754 55 A3 -105.147952 5 220.295903 56 fitted -110.046917 4 228.093834 57 R -112.382522 2 228.765045 58 59 60 E x p l a n a t i o n o f T e s t s 61 62 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 63 (A2 vs. R) 64 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 65 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 66 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 67 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 68 69 T e s t s o f I n t e r e s t 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-263 DRAFT--DO NOT CITE OR QUOTE 1 Test -2*log(Likelihood Ratio) Test df p-value 2 3 Test 1 16.2263 4 0.00273 4 Test 2 8.61084 2 0.0135 5 Test 3 1.75715 1 0.185 6 Test 4 9.79793 1 0.001747 7 8 The p - v a l u e for Test 1 is less t h a n .05. T h e r e a p p e a r s to be a 9 difference between response and/or variances among the dose levels 10 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 11 12 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 13 m o d e l a p p e a r s t o b e a p p r o p r i a t e 14 15 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 16 t o b e a p p r o p r i a t e h e r e 17 18 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 19 m o d e l 20 21 22 B e n c h m a r k D o s e C o m p u t a t i o n 23 24 S p e c i f i e d e f f e c t = 1 25 26 R i s k T y p e = Estimated standard deviations from the control mean 27 28 C o n f i d e n c e l e v e l = 0.95 29 30 BMD = 126.365 31 32 33 BMDL = 61.2812 34 This document is a draftfor review purposes only and does not constitute Agency policy. E-264 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E.3.2.3. F igure f o r Selected M odel: L in ea r Linear Model with 0.95 Confidence Level 2 17:22 02/16 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-265 DRAFT--DO NOT CITE OR QUOTE 1 E.3.3. Amin et al., 2000: 0.50% Saccharin Consumed, Female 2 E.3.3.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes linear b 1 0.031 159.737 9.874E+01 6.417E+01 polynom ial, 2degree 1 0.031 159.737 9.874E+01 6.417E+01 power 1 0.031 159.737 9 .874E + 01 6.4 1 7 E + 0 1 p ow er b oun d hit (p ow er = 1) power, unrestricted c 0 N /A 157.060 5.610E+01 6.781E +00 unrestricted (power = 0.325) a N on-con stant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .3 .2 . Outputfor Selected Model: Linear 6 A m in et al., 2000: 0.50% Saccharin Consum ed, Fem ale 7 8 9 10 P o l y n o m i a l M o d e l . ( V e r s i o n : 2 . 1 3 ; D a t e : 0 4 / 0 8 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 3 _ A m i n _ 2 0 0 0 _ 5 0 _ S C _ L i n e a r _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 3 _ A m i n _ 2 0 0 0 _ 5 0 _ S C _ L i n e a r _ 1 . p l t 13 ~ T u e _ F e b 1 6 ~ 1 7 : 2 3 : 1 4 2 0 1 0 14 15 16 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 S i g n s o f t h e p o l y n o m i a l c o e f f i c i e n t s a r e n o t r e s t r i c t e d 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 28 29 T o t a l n u m b e r o f d o s e g r o u p s = 3 30 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 31 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 32 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 33 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 34 35 36 37 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 38 lalpha = 4.68512 39 rho = 0 40 beta_0 = 19.3484 41 beta_1 = -0.158141 42 43 44 Asymptotic Correlation Matrix of Parameter Estimates This document is a draftfor review purposes only and does not constitute Agency policy. E-266 DRAFT--DO NOT CITE OR QUOTE 1 2 lalpha rho beta 0 beta 1 3 4 lalpha 1 -0.97 0.018 -0.0021 5 6 rho -0.97 1 -0.027 0.014 7 Lf) o Lf) o 8 beta 0 0.018 -0.027 1 9 10 beta 1 -0.0021 0.014 1 11 12 13 14 P a r a m e t e r E s t i m a t e s 15 16 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 17 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 18 lalpha -0.997428 0.992786 -2.94325 0.948397 19 rho 2.13634 0.404989 1.34257 2.9301 20 beta 0 18.1144 3.10302 12.0326 24.1962 21 beta 1 -0.135736 0.0331501 -0.200709 -0.0707631 22 23 24 25 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 26 27 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 28 29 30 0 10 22.4 18.1 16 13.4 1 31 25 10 11.4 14.7 7.66 10.7 -0.983 32 100 10 4.54 4.54 3.33 3.06 -0.00393 33 34 35 36 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 37 38 39 M o d e l A1: Yij Mu(i) + e(ij) 40 Var{e(ij)} Sigma^2 41 42 M o d e l A2: Yij Mu(i) + e(ij) 43 Var{e(ij)} Sigma(i)^2 44 45 M o d e l A3: Yij = Mu(i) + e(ij) 46 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 47 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 48 w e r e s p e c i f i e d b y t h e u s e r 49 50 M o d e l R: Yi = M u + e(i) 51 V a r { e ( i ) } = S i g m a ^ 2 52 53 54 L i k e l i h o o d s o f I n t e r e s t 55 56 Model Log(likelihood) # Param's AIC 57 A1 -83.696404 4 175.392808 58 A2 -73.511830 6 159.023660 59 A3 -73.530233 5 157.060467 60 fitted -75.868688 4 159.737377 61 R -90.294746 2 184.589492 62 63 64 E x p l a n a t i o n o f T e s t s 65 66 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 67 (A2 vs. R) 68 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 69 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 70 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) This document is a draftfor review purposes only and does not constitute Agency policy. E-267 DRAFT--DO NOT CITE OR QUOTE 1 (Note: When rho=0 the results of Test 3 and Test 2 will be the same.) 2 3 Tests of Interest 4 5 Test -2*log(Likelihood Ratio) Test df p-value 6 7 Test 1 33.5658 4 <.0001 8 Test 2 20.3691 2 <.0001 9 Test 3 0.0368066 1 0.8479 10 T e s t 4 4.67691 1 0.03057 11 12 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 13 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 14 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 15 16 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 17 m o d e l a p p e a r s t o b e a p p r o p r i a t e 18 19 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 20 t o b e a p p r o p r i a t e h e r e 21 22 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 23 m o d e l 24 25 26 B e n c h m a r k D o s e C o m p u t a t i o n 27 28 S p e c i f i e d e f f e c t = 1 29 30 R i s k T y p e = Estimated standard deviations from the control mean 31 32 C o n f i d e n c e l e v e l = 0.95 33 34 BMD = 98.7409 35 36 37 BMDL = 64.169 38 This document is a draftfor review purposes only and does not constitute Agency policy. E-268 DRAFT--DO NOT CITE OR QUOTE 1 E.3.3.3. F igure f o r Selected M odel: L in ea r Linear Model with 0.95 Confidence Level 2 17:23 02/16 2010 3 4 5 E .3.3.4. Outputfor Additional Model Presented: Power, Unrestricted 6 A m in et al., 2000: 0.50% Saccharin Consum ed, Fem ale 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 3 _ A m i n _ 2 0 0 0 _ 5 0 _ S C _ P w r _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 3 _ A m i n _ 2 0 0 0 _ 5 0 _ S C _ P w r _ U _ 1 . p l t 13 ~ T u e _ F e b ~ 1 6 1 7 : 2 3 : 1 5 2 0 1 0 14 15 16 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 T h e p o w e r is n o t r e s t r i c t e d 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-269 DRAFT--DO NOT CITE OR QUOTE 1 Total number of dose groups = 3 2 Total number of records with missing values = 0 3 Maximum number of iterations = 250 4 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 5 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 6 7 8 9 Default Initial Parameter Values 10 lalpha = 4.68512 11 rho = 0 12 control = 22.3564 13 slope = -3.55874 14 power = 0.349799 15 16 17 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 18 19 lalpha rho control slope power 20 21 l a l p h a 1 -0.96 0.34 -0.26 -0.15 22 23 rho -0.96 1 -0.47 0.3 0.15 24 25 c o n t r o l 0.34 -0.47 1 -0.73 -0.52 26 27 slope -0.26 0.3 -0.73 1 0.96 28 29 power -0.15 0.15 -0.52 0.96 1 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 36 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 37 lalpha -0.708629 1.298 -3.25267 1.83541 38 rho 1.96142 0.529653 0.923323 2.99953 39 control 22.6293 4.48416 13.8405 31.4181 40 slope -4.03215 3.21302 -10.3296 2.26526 41 power 0.325414 0.138761 0.053447 0.597381 42 43 44 45 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 46 47 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 48 49 50 0 10 22.4 22.6 16 15 -0.0577 51 25 10 11.4 11.1 7.66 7.46 0.105 52 100 10 4.54 4.58 3.33 3.12 -0.0475 53 54 W a r n i n g : L i k e l i h o o d f o r f i t t e d m o d e l l a r g e r t h a n t h e L i k e l i h o o d f o r m o d e l A3. 55 56 57 58 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 59 60 61 M o d e l A1: Yij Mu(i) + e(ij) 62 Var{e(ij)} Sigma^2 63 64 M o d e l A2: Yij Mu(i) + e(ij) 65 Var{e(ij)} Sigma(i)^2 66 67 M o d e l A3: Yij = Mu(i) + e(ij) 68 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 69 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 70 w e r e s p e c i f i e d b y t h e u s e r This document is a draftfor review purposes only and does not constitute Agency policy. E-270 DRAFT--DO NOT CITE OR QUOTE 1 2 M o d e l R: Yi = M u + e(i) 3 Var{e(i)} = Sigma^2 4 5 6 Likelihoods of Interest 7 8 Model Log(likelihood) # Param's AIC 9 A1 -83.696404 4 175.392808 10 A2 -73.511830 6 159.023660 11 A3 -73.530233 5 157.060467 12 fitted -73.530233 5 157.060467 13 R -90.294746 2 184.589492 14 15 16 E x p l a n a t i o n o f T e s t s 17 18 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 19 (A2 vs. R) 20 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 21 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 22 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 23 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 24 25 T e s t s o f I n t e r e s t 26 27 Test -2*log(Likelihood Ratio) Test df p-value 28 29 T e s t 1 33.5658 4 <.0001 30 T e s t 2 20.3691 2 <.0001 31 T e s t 3 0.0368066 1 0.8479 32 T e s t 4 -2.84217e-014 0 NA 33 34 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 35 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 36 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 37 38 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 39 m o d e l a p p e a r s t o b e a p p r o p r i a t e 40 41 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 42 t o b e a p p r o p r i a t e h e r e 43 44 N A - D e g r e e s o f f r e e d o m f o r T e s t 4 a r e l e s s t h a n o r e q u a l t o 0. T h e C h i - S q u a r e 45 t e s t f o r f i t is n o t v a l i d 46 47 48 B e n c h m a r k D o s e C o m p u t a t i o n 49 50 S p e c i f i e d e f f e c t = 1 51 52 R i s k T y p e = Estimated standard deviations from the control mean 53 54 C o n f i d e n c e l e v e l = 0.95 55 56 B M D = 5 6 . 0 9 6 7 57 58 59 B M D L = 6 . 7 8 1 1 2 60 This document is a draftfor review purposes only and does not constitute Agency policy. E-271 DRAFT--DO NOT CITE OR QUOTE 1 E.3.3.5. F igure f o r A d d itio n a l M o d el Presented: Power, Unrestricted Power Model with 0.95 Confidence Level Mean Response 2 17:23 02/16 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-272 DRAFT--DO NOT CITE OR QUOTE 1 E.3.4. Amin et al., 2000: 0.50% Saccharin Preference Ratio, Female 2 E.3.4.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes linear b 1 0.088 234.936 8.278E+01 5.100E+01 polynom ial, 2degree 1 0.088 234.936 8.278E+01 5.100E+01 power 1 0 .0 8 8 2 3 4 .9 3 6 8 .278E + 01 5 .100E + 01 p ow er b oun d hit (p ow er = 1) power, unrestricted c 0 N /A 234.020 1.817E+01 1.000E-13 unrestricted (power = 0.232) a C onstant variance m odel selected (p = 0 .5 5 9 3 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .4 .2 . Outputfor Selected Model: Linear 6 A m in et al., 2000: 0.50% Saccharin Preference Ratio, Fem ale 7 8 9 10 P o l y n o m i a l M o d e l . ( V e r s i o n : 2 . 1 3 ; D a t e : 0 4 / 0 8 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 4 _ A m i n _ 2 0 0 0 _ 5 0 _ S P _ L i n e a r C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 4 _ A m i n _ 2 0 0 0 _ 5 0 _ S P _ L i n e a r C V _ 1 . p l t 13 ~ T u e _ F e b 16 1 7 : 2 3 : 4 3 2 0 1 0 14 15 16 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 S i g n s o f t h e p o l y n o m i a l c o e f f i c i e n t s a r e n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t 29 30 T o t a l n u m b e r o f d o s e g r o u p s = 3 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 33 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 34 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 35 36 37 38 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 39 alpha = 764.602 40 rho = 0 Specified 41 beta_0 = 64.1858 42 beta_1 = -0.332668 43 44 This document is a draftfor review purposes only and does not constitute Agency policy. E-273 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 ( *** The model parameter(s) -rho 4 have been estimated at a boundary point, or have been specified by the user 5 and do not appear in the correlation matrix ) 6 7 alpha beta 0 beta 1 8 9 alpha 1 2e-008 1.4e-009 10 11 beta 0 2e-008 1 -0.7 12 13 beta 1 1.4e-009 -0.7 1 14 15 16 17 P a r a m e t e r E s t i m a t e s 18 19 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 20 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 21 alpha 758.396 195.817 374.602 1142.19 22 beta 0 64.1858 7.04184 50.3841 77.9876 23 beta 1 -0.332668 0.118327 -0.564584 -0.100752 24 25 26 27 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 28 29 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 30 31 32 0 10 72.7 64.2 24.6 27.5 0.981 33 25 10 44.5 55.9 32.9 27.5 -1.31 34 100 10 33.8 30.9 24.6 27.5 0.327 35 36 37 38 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 39 40 41 M o d e l A1: Yij Mu(i) + e(ij) 42 Var{e(ij)} Sigma^2 43 44 M o d e l A2: Yij Mu(i) + e(ij) 45 Var{e(ij)} Sigma(i)^2 46 47 M o d e l A3: Yij = Mu(i) + e(ij) 48 V a r { e ( i j ) } = S i g m a ^ 2 49 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 50 w e r e s p e c i f i e d b y t h e u s e r 51 52 M o d e l R: Yi = M u + e(i) 53 V a r { e ( i ) } = S i g m a ^ 2 54 55 56 L i k e l i h o o d s o f I n t e r e s t 57 58 Model Log(likelihood) # Param's AIC 59 A1 -113.009921 4 234.019841 60 A2 -112.428886 6 236.857773 61 A3 -113.009921 4 234.019841 62 fitted -114.468091 3 234.936183 63 R -117.976057 2 239.952114 64 65 66 E x p l a n a t i o n o f T e s t s 67 68 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 69 (A2 vs. R) 70 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) This document is a draftfor review purposes only and does not constitute Agency policy. E-274 DRAFT--DO NOT CITE OR QUOTE 1 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 2 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 3 (Note: When rho=0 the results of Test 3 and Test 2 will be the same.) 4 5 Tests of Interest 6 7 Test -2*log(Likelihood Ratio) Test df p-value 8 9 Test 1 11.0943 4 0.02552 10 T e s t 2 1.16207 2 0.5593 11 T e s t 3 1.16207 2 0.5593 12 T e s t 4 2.91634 1 0.08769 13 14 p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e 15 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 16 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 17 18 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 19 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 20 21 22 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 23 t o b e a p p r o p r i a t e h e r e 24 25 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 26 m o d e l 27 28 29 B e n c h m a r k D o s e C o m p u t a t i o n 30 31 S p e c i f i e d e f f e c t = 1 32 33 R i s k T y p e = Estimated standard deviations from the control mean 34 35 C o n f i d e n c e l e v e l = 0.95 36 37 BMD = 82.7823 38 39 40 BMDL = 50.9971 41 This document is a draftfor review purposes only and does not constitute Agency policy. E-275 DRAFT--DO NOT CITE OR QUOTE 1 E.3.4.3. F igure f o r Selected M odel: L in ea r Linear Model with 0.95 Confidence Level 2 17:23 02/16 2010 3 dose 4 5 E .3.4.4. Outputfor Additional Model Presented: Power, Unrestricted 6 A m in et al., 2000: 0.50% Saccharin Preference Ratio, Fem ale 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 4 _ A m i n _ 2 0 0 0 _ 5 0 _ S P _ P w r C V _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 4 _ A m i n _ 2 0 0 0 _ 5 0 _ S P _ P w r C V _ U _ 1 . p l t 13 ~ T u e _ F e b 16 1 7 : 2 3 : 4 4 2 0 1 0 14 15 16 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 T h e p o w e r is n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t This document is a draftfor review purposes only and does not constitute Agency policy. E-276 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 3 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha = 764.602 12 rho = 0 Specified 13 control = 72.7273 14 slope = -13.387 15 power = 0.231973 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 21 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 22 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 23 24 alpha control slope power 25 26 a l p h a 1 -1.3e-008 5.9e-009 2.5e-009 27 28 control -1.3e-008 1 -0.4 -0.22 29 30 slope 5.9e-009 -0.4 1 0.97 31 32 power 2.5e-009 -0.22 0.97 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 39 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 40 alpha 688.142 177.677 339.9 1036.38 41 control 72.7273 8.29543 56.4686 88.986 42 slope -13.387 15.9957 -44.738 17.9639 43 power 0.231973 0.268067 -0.293429 0.757376 44 45 46 47 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 48 49 D o s e N Obs Mean Est Mean ; Std Dev Est Std Dev Scaled Res 50 51 52 0 10 72.7 72.7 24.6 26.2 5.16e-008 53 25 10 44.5 44.5 32.9 26.2 -1.27e-008 54 100 10 33.8 33.8 24.6 26.2 -2e-008 55 56 D e g r e e s o f f r e e d o m f o r T e s t A 3 v s f i t t e d < 0 57 58 59 60 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 61 62 63 M o d e l A1: Yij = M u ( i ) + e( 64 V a r j e i ;ij)} = S i g m a ^ 2 65 66 M o d e l A2: Yij = M u ( i ) + e( 67 V a r j e i :ij)} = S i g m a ( i ) ^ 2 68 69 M o d e l A3: Yij = M u ( i ) + e( 70 V a r j e i :ij)} = S i g m a ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-277 DRAFT--DO NOT CITE OR QUOTE 1 Model A3 uses any fixed variance parameters that 2 were specified by the user 3 4 M o d e l R: Yi = M u + e(i) 5 Var{e(i)} = Sigma^2 6 7 8 Likelihoods of Interest 9 10 Model Log(likelihood) # Param's AIC 11 A1 -113.009921 4 234.019841 12 A2 -112.428886 6 236.857773 13 A3 -113.009921 4 234.019841 14 fitted -113.009921 4 234.019841 15 R -117.976057 2 239.952114 16 17 18 E x p l a n a t i o n o f T e s t s 19 20 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 21 (A2 vs. R) 22 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 23 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 24 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 25 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 26 27 T e s t s o f I n t e r e s t 28 29 Test -2*log(Likelihood Ratio) Test df p-vali 30 31 T e s t 1 11.0943 4 0.02552 32 T e s t 2 1.16207 2 0.5593 33 T e s t 3 1.16207 2 0.5593 34 T e s t 4 00 NA 35 36 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 37 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 38 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 39 40 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 41 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 42 43 44 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 45 t o b e a p p r o p r i a t e h e r e 46 47 N A - D e g r e e s o f f r e e d o m f o r T e s t 4 a r e l e s s t h a n o r e q u a l t o 0. T h e C h i - S q u a r e 48 t e s t f o r f i t is n o t v a l i d 49 50 51 B e n c h m a r k D o s e C o m p u t a t i o n 52 53 S p e c i f i e d e f f e c t = 1 54 55 R i s k T y p e = Estimated standard deviations from the control mean 56 57 C o n f i d e n c e l e v e l = 0.95 58 59 B M D = 1 8 . 1 7 3 2 60 61 62 B M D L = 1 e - 0 1 3 63 This document is a draftfor review purposes only and does not constitute Agency policy. E-278 DRAFT--DO NOT CITE OR QUOTE 1 E.3.4.5. F igure f o r A d d itio n a l M o d el Presented: Power, Unrestricted Power Model with 0.95 Confidence Level Mean Response 2 17:23 02/16 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-279 DRAFT--DO NOT CITE OR QUOTE 1 E.3.5. Bell et al., 2007a: B alano-P reputial Separation, Postnatal D ay 49 2 E .3 .5 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes gamma 2 0.369 1 13.514 7 .3 3 2 E + 0 0 4 .6 8 7 E + 0 0 p o w er b oun d hit (p ow er = 1) logistic negative intercept (intercept = 2 0.237 114.853 1.501E+01 1.137E+01 -2.07) log-logistic a 2 0.456 112.952 5.209E+00 2.870E+00 slope bound hit (slope = 1) log-probit 2 0.178 1 15.488 1.428E +01 9 .1 3 8 E + 0 0 slop e b ound hit (slop e = 1) multistage, 3degree probit 2 0.369 113.514 7.332E+00 4.687E+00 final B = 0 negative intercept (intercept = 2 0.248 114.723 1.399E+01 1.061E+01 -1.23) W eibull 2 0.369 1 13.514 7 .3 3 2 E + 0 0 4 .6 8 7 E + 0 0 p o w er b oun d hit (p ow er = 1) gamma, unrestricted log-logistic, unrestricted b log-probit, unrestricted W eibull, unrestricted 1 0.566 113.746 1.894E+00 7.609E -02 unrestricted (pow er = 0.506) 1 0.484 113.908 2.127E+00 1.363E-01 unrestricted (slope = 0.67) 1 0.439 114.021 2.179E+00 1.671E-01 unrestricted (slope = 0.389) 1 0.534 113.802 2.007E +00 1.075E-01 unrestricted (pow er = 0.574) aBest-fitting m odel, BM D S output presented in this appendix b Alternate m odel, BM D S output also presented in this appendix 3 E .3.5.2. 4 E .3 .5 .3 . Outputfor Selected Model: Log-Logistic 5 B ell et al., 2007a: Balano-Preputial Separation, Postnatal D ay 49 6 7 8 9 Logistic Model. (Version: 2.12; Date: 05/16/2008) 10 I n p u t D a t a F i l e : C : \ 1 \ 5 _ B e l l _ 2 0 0 7 _ B P S _ L o g L o g i s t i c _ 1 . ( d ) 11 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 5 _ B e l l _ 2 0 0 7 _ B P S _ L o g L o g i s t i c _ 1 . p l t 12 ~ T u e F e b 16 1 7 7 2 4 : 1 0 2 0 1 0 13 14 15 0 16 17 18 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 19 20 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 21 22 23 D e p e n d e n t v a r i a b l e = D i c h E f f 24 I n d e p e n d e n t v a r i a b l e = D o s e 25 S l o p e p a r a m e t e r is r e s t r i c t e d as s l o p e > = 1 This document is a draftfor review purposes only and does not constitute Agency policy. E-280 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of observations = 4 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 11 12 13 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 14 background = 0.0333333 15 intercept = -3.75371 16 slope = 1 17 18 19 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 20 21 ( * * * T h e m o d e l p a r a m e t e r ( s ) - s l o p e 22 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 23 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 24 25 background intercept 26 27 b a c k g r o u n d 1 -0.58 28 29 i n t e r c e p t -0.58 1 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 36 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf 37 b a c k g r o u n d 0 . 0 6 3 5 2 5 1 * * * 38 intercept -3.84765 * * * 39 s l o p e 1 40 41 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 42 43 44 45 A n a l y s i s o f D e v i a n c e T a b l e 46 47 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 48 Full model -53.7077 4 49 F i t t e d m o d e l -54.476 2 1.53661 2 0.4638 50 R e d u c e d m o d e l -63.9797 1 20.544 3 0.0001309 51 52 AIC: 112.952 53 54 55 G o o d n e s s o f F i t 56 57 Dose Est. Prob. Expected Observed Size Residual 58 59 0.0000 0.0635 1.906 1.000 30 -0.678 60 2.4000 0.1091 3.274 5.000 30 1.011 61 8.0000 0.2000 6.001 6.000 30 -0.000 62 46.0000 0.5273 15.819 15.000 30 -0.300 63 64 C h i ^ 2 = 1 . 5 7 d.f. = 2 P-value = 0.4559 65 66 67 B e n c h m a r k D o s e C o m p u t a t i o n 68 69 S p e c i f i e d e f f e c t = 0.1 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-281 DRAFT--DO NOT CITE OR QUOTE 1 Risk Type 2 3 Confidence level 4 5 BMD 6 7 BMDL 8 9 Extra risk 0.95 5.20918 2.86991 10 E .3 .5 .4 . Figurefor Selected Model: Log-Logistic Log-Logistic Model with 0.95 Confidence Level Fraction Affected 11 17:24 02/16 2010 12 13 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-282 DRAFT--DO NOT CITE OR QUOTE 1 E .3.5.5. Outputfor Additional Model Presented: Log-Logistic, Unrestricted 2 B ell et al., 2007a: Balano-Preputial Separation, Postnatal D ay 49 3 4 5 6 Logistic Model. (Version: 2.12; Date: 05/16/2008) 7 Input Data File: C:\1\5_Bell_2007_BPS_LogLogistic_U_1.(d) 8 Gnuplot Plotting File: C:\1\5_Bell_2007_BPS_LogLogistic_U_1.plt 9 ~ Tue Feb 16 17721:10 2010 10 11 12 0 13 14 15 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 16 17 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 18 19 20 D e p e n d e n t v a r i a b l e = D i c h E f f 21 I n d e p e n d e n t v a r i a b l e = D o s e 22 S l o p e p a r a m e t e r is n o t r e s t r i c t e d 23 24 T o t a l n u m b e r o f o b s e r v a t i o n s = 4 25 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 26 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 27 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 28 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 29 30 31 32 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 33 34 35 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 36 background = 0.0333333 37 intercept = -2.54947 38 slope = 0.615936 39 40 41 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 42 43 background intercept slope 44 45 b a c k g r o u n d 1 -0.49 0.35 46 CO Oi o CO Oi o 47 i n t e r c e p t -0.49 1 48 49 s l o p e 0.35 1 50 51 52 53 P a r a m e t e r E s t i m a t e s 54 55 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 56 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 57 b a c k g r o u n d 0 . 0 3 5 4 7 1 4 58 intercept -2.70296 59 slope 0.670238 60 61 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 62 63 64 65 A n a l y s i s o f D e v i a n c e T a b l e 66 67 Model Log(likelihood) # Param's Deviance Test d.f. P-value 68 Full model -53.7077 4 This document is a draftfor review purposes only and does not constitute Agency policy. E-283 DRAFT--DO NOT CITE OR QUOTE 1 Fitted model -53.9541 3 0.492844 1 0.4827 2 Reduced model -63.9797 1 20.544 3 0.0001309 3 4 AIC: 113.908 5 6 7 Goodness of Fit 8 Scaled 9 Dose Est. Prob. Expected Observed Size Residual 10 11 0.0000 0.0355 1.064 1.000 30 -0.063 12 2.4000 0.1392 4.176 5.000 30 0.435 13 8.0000 0.2405 7.216 6.000 30 -0.520 14 46.0000 0.4848 14.544 15.000 30 0.167 15 16 C h i ^ 2 = 0 . 4 9 d.f. = 1 P-value = 0.4836 17 18 19 B e n c h m a r k D o s e C o m p u t a t i o n 20 21 S p e c i f i e d e f f e c t = 0.1 22 23 R i s k T y p e = Extra risk 24 25 C o n f i d e n c e l e v e l = 0.95 26 27 BMD = 2.12667 28 29 BMDL = 0.13633 30 31 This document is a draftfor review purposes only and does not constitute Agency policy. E-284 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 E.3.5.6. F igure f o r A d d itio n a l M o d el Presented: Log-Logistic, U nrestricted Log-Logistic Model with 0.95 Confidence Level 2 17:24 02/16 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-285 DRAFT--DO NOT CITE OR QUOTE 1 E.3.6. Cantoni et al., 1981: Urinary Coproporhyrins, 3 Months 2 E.3.6.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M 2) 2 0.002 33.792 1.101E+02 5.318E+01 Notes exponential (M 3) exponential (M4) b exponential (M 5) 2 0.002 33.792 1 .101E + 02 5.3 1 8 E + 0 1 p ow er hit bound (d = 1) 1 0.341 23.881 3.741E-01 1.253E-01 1 0.341 23.881 3 .741E -01 1.253E -01 p ow er hit bound (d = 1) H ill 1 0.535 23.359 3.273E-01 error n lo w er b oun d hit (n = 1) linear polynom ial, 3degree power power, unrestricted c H ill, unrestricted 2 0.002 33.301 7.734E+01 1.975E+01 2 0.002 33.301 7.734E+01 1.975E+01 2 0 .0 0 2 33.301 7 .7 3 4 E + 0 1 1.975E +01 p ow er b oun d hit (p ow er = 1) 1 0.665 23.162 4.637E-03 8.796E-08 unrestricted (power = 0.22) 0 N /A 24.974 7.264E -02 1.656E-04 unrestricted (n = 0.48) a N on-constant variance m odel selected (p = 0 .0039) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3.6.2. Outputfor Selected Model: Exponential (M4) 6 Cantoni et al., 1981: Urinary Coproporhyrins, 3 M onths 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 6 _ C a n t o n i _ 1 9 8 1 _ U r i C o p r o _ E x p _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 T u e F e b 16 1 7 : 2 4 : 3 9 2 0 1 0 14 15 16 F i g u r e l - U r i n a r y C o p r o p o r p h y r i n 3 m o n t h s 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-l) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-l) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; This document is a draftfor review purposes only and does not constitute Agency policy. E-286 DRAFT--DO NOT CITE OR QUOTE 1 sign = -1 for d e c r e a s i n g trend. 2 3 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 4 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 5 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 6 7 8 Dependent variable = Mean 9 Independent variable = Dose 10 D a t a a r e a s s u m e d t o b e d i s t r i b u t e d : n o r m a l l y 11 V a r i a n c e M o d e l : e x p ( l n a l p h a + r h o * l n ( Y [ d o s e ] ) ) 12 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 13 14 T o t a l n u m b e r o f d o s e g r o u p s = 4 15 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 16 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 17 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 18 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 19 20 M L E s o l u t i o n p r o v i d e d : E x a c t 21 22 23 I n i t i a l P a r a m e t e r V a l u e s 24 25 Variable Model 4 26 27 lnalpha -1.50063 28 rho 2.60979 29 a 0 . 7 0 4 3 0 3 30 b 0 . 0 2 0 5 9 2 7 31 c 4 . 4 7 2 6 8 32 d 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 Variable Model 4 39 40 lnalpha -1.74154 41 rho 2.66803 42 a 0 . 7 5 5 9 8 2 43 b 0 . 3 7 1 5 44 c 3 . 9 3 8 4 5 45 d 1 46 47 48 T a b l e o f S t a t s F r o m I n p u t D a t a 49 50 Dose N Obs Mean Obs Std 51 52 04 0.7414 0.3475 53 1.43 4 1.807 0.8341 54 14.3 4 2.734 1.506 55 143 4 3 2.6 56 57 58 E s t i m a t e d V a l u e s o f I n t e r e s t 59 60 Dose Est Mean Est Std Scaled Residual 61 62 0 0.756 0.2882 -0.1014 63 1.43 1.671 0.8307 0.3265 64 14.3 2.966 1.786 -0.2607 65 143 2.977 1.794 0.02532 66 67 68 69 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d : 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-287 DRAFT--DO NOT CITE OR QUOTE 1 Model A1: Yij = Mu(i) + e(ij) 2 V a r { e i ;ij)} = S i g m a ^ 2 3 4 Model A2: Yij = Mu(i) + e(ij) 5 V a r { e i :ij)} = S i g m a ( i ) ^ 2 6 7 Model A3: Yij = Mu(i) + e(ij) 8 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 9 10 M o d e l R: Yij = M u + e(i) 11 V a r { e i ij)} = S i g m a ^ 2 12 13 14 L i k e l i h o o d s o f I n t e r e s t 15 16 Model Log(likelihood) DF AIC 17 18 A1 -12.90166 5 35.80333 19 A2 -6.203643 8 28.40729 20 A3 -6.487204 6 24.97441 21 R -15.73713 2 35.47427 22 4 -6.940389 5 23.88078 23 24 25 Additive constant for all log-likelihoods = -14.7. This constant added to the 26 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 27 d e p e n d o n t h e m o d e l p a r a m e t e r s . 28 29 30 E x p l a n a t i o n o f T e s t s 31 32 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 33 T e s t 2: A r e H o m o g e n e o u s ? (A2 vs. A1) 34 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 35 36 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 37 38 39 T e s t s o f I n t e r e s t 40 41 T e s t -2*log(Likelihood Ratio) D. F. p-value 42 43 T e s t 1 19.07 6 0.004052 44 T e s t 2 13.4 3 0.003854 45 T e s t 3 0.5671 2 0.7531 46 T e s t 6a 0.9064 1 0.3411 47 48 49 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 50 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 51 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 52 53 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 54 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e . 55 56 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 57 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 58 59 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 60 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 61 62 63 B e n c h m a r k D o s e C o m p u t a t i o n s : 64 65 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 66 67 R i s k T y p e = E s t i m a t e d s t a n d a r d d e v i a t i o n s f r o m c o n t r o l 68 69 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-288 DRAFT--DO NOT CITE OR QUOTE 1 BMD = 0.374114 2 3 BMDL = 0.125287 4 5 6 E .3.6.3. Figurefor Selected Model: Exponential (M4) Exponential_beta Model 4 with 0.95 Confidence Level 7 8 9 10 E .3.6.4. Outputfor Additional Model Presented: Power, Unrestricted 11 Cantoni et al., 1981: Urinary Coproporhyrins, 3 M onths 12 13 14 15 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 16 I n p u t D a t a F i l e : C : \ 1 \ 6 _ C a n t o n i _ 1 9 8 1 _ U r i C o p r o _ P w r _ U _ 1 . ( d ) 17 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 6 _ C a n t o n i _ 1 9 8 1 _ U r i C o p r o _ P w r _ U _ 1 . p l t 18 T u e F e b 16 1 7 7 2 4 : 4 1 2 0 1 0 19 20 21 F i g u r e 1 - U r i n a r y C o p r o p o r p h y r i n 3 m o n t h s 22 23 24 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 25 26 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 27 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-289 DRAFT--DO NOT CITE OR QUOTE 1 Dependent variable = Mean 2 Independent variable = Dose 3 The power is not restricted 4 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) * rho) 5 6 Total number of dose groups = 4 7 Total number of records with missing values = 0 8 Maximum number of iterations = 250 9 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 10 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 11 12 13 14 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 15 lalpha = 0.90039 16 rho = 0 17 control = 0.741372 18 slope = 1.00533 19 power = 0.163111 20 21 22 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 23 24 lalpha rho control slope power 25 26 l a l p h a 1 -0.62 -0.53 -0.038 0.027 27 28 rho -0.62 1 0.43 -0.24 -0.16 29 30 c o n t r o l CO Lf) o 0.43 1 -0.3 0.09 31 32 slope -0.038 -0.24 -0.3 1 -0.72 33 34 p o w e r 0 . 0 2 7 - 0 . 1 6 0.09 -0.72 1 35 36 37 38 P a r a m e t e r E s t i m a t e s 39 40 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 41 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 42 lalpha -1.78404 0.61698 -2.9933 -0.57478 43 rho 2.6428 0.74449 1.18363 4.10197 44 control 0.757242 0.139966 0.482915 1.03157 45 slope 0.927009 0.325923 0.288212 1.56581 46 power 0.220276 0.0964599 0.031218 0.409334 47 48 49 50 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 51 52 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 53 54 55 04 0.741 0.757 0.348 0.284 -0.112 56 1 . 4 3 4 1.81 1.76 0.834 0.865 0.108 57 1 4 . 3 4 2.73 2.42 1.51 1.32 0.471 58 143 4 3 3.52 2.6 2.16 -0.483 59 60 61 62 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 63 64 65 M o d e l A1: Yij Mu(i) + e(ij) 66 Var{e(ij)} Sigma^2 67 68 M o d e l A2: Yij Mu(i) + e(ij) 69 Var{e(ij)} Sigma(i)^2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-290 DRAFT--DO NOT CITE OR QUOTE 1 Model A3: Yij = Mu(i) + e(ij) 2 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 3 Model A3 uses any fixed variance parameters that 4 were specified by the user 5 6 M o d e l R: Yi = M u + e(i) 7 Var{e(i)} = Sigma^2 8 9 10 L i k e l i h o o d s o f I n t e r e s t 11 12 Model Log(likelihood) # Param's AIC 13 A1 -12.901663 5 35.803325 14 A2 -6.203643 8 28.407287 15 A3 -6.487204 6 24.974409 16 fitted -6.580755 5 23.161510 17 R -15.737135 2 35.474269 18 19 20 E x p l a n a t i o n o f T e s t s 21 22 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 23 (A2 vs. R) 24 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 25 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 26 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 27 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 28 29 T e s t s o f I n t e r e s t 30 31 Test -2*log(Likelihood Ratio) Test df p-value 32 33 T e s t 1 19.067 6 0.004052 34 T e s t 2 13.396 3 0.003854 35 T e s t 3 0.567122 2 0.7531 36 T e s t 4 0.187101 1 0.6653 37 38 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 39 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 40 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 41 42 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 43 m o d e l a p p e a r s t o b e a p p r o p r i a t e 44 45 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 46 t o b e a p p r o p r i a t e h e r e 47 48 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 49 t o a d e q u a t e l y d e s c r i b e t h e d a t a 50 51 52 B e n c h m a r k D o s e C o m p u t a t i o n 53 54 S p e c i f i e d e f f e c t = 1 55 56 R i s k T y p e = Estimated standard deviations from the control mean 57 58 C o n f i d e n c e l e v e l = 0.95 59 60 B M D = 0 . 0 0 4 6 3 7 4 6 61 62 63 B M D L = 8 . 7 9 6 3 4 e - 0 0 8 64 65 This document is a draftfor review purposes only and does not constitute Agency policy. E-291 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E.3.6.5. F igure f o r A d d itio n a l M o d el Presented: Power, Unrestricted Power Model with 0.95 Confidence Level 2 17:24 02/16 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-292 DRAFT--DO NOT CITE OR QUOTE 1 E.3.7. C antoni et al., 1981: U rinary Porphyrins 2 E .3 .7 .1 . Summary Table o f BMDS Modeling Results Model a exponential (M2) b Degrees of Freedom x2p Value AIC 2 <0.0001 58.753 BMD BMDL (ng/kg-d) (ng/kg-d) 1.223E+01 9.037E+00 Notes exponential (M 3) 2 < 0 .0 0 0 1 58.753 1.223E +01 9 .0 3 7 E + 0 0 p o w er hit bound (d = 1) exponential (M 4) 1 <0.0001 63.138 2.227E-01 1.137E-01 exponential (M 5) 1 < 0 .0 0 0 1 6 3 .1 3 8 2 .2 2 7 E -0 1 1.137E -01 p o w er hit bound (d = 1) Hill 0 N /A 62.356 9.363E+00 4.664E+00 linear 2 <0.0001 62.487 7.732E-01 2.816E-01 polynom ial, 3degree 1 <0.0001 10.000 error error power 2 < 0 .0 0 0 1 6 2 .4 8 7 7 .7 3 2 E -0 1 2.8 1 6 E -0 1 p o w er b oun d hit (p ow er = 1) power, unrestricted 1 <0.0001 59.914 1.025E-01 2.389E -02 unrestricted (power = 0.746) a N on-con stant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3.7.2. Outputfor Selected Model: Exponential (M2) 6 Cantoni et al., 1981: Urinary Porphyrins 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 7 _ C a n t o n i _ 1 9 8 1 _ U r i P o r _ E x p _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 T u e F e b 16 1 7 : 2 5 : 1 4 2 0 1 0 14 15 16 T a b l e 1, d o s e c o n v e r t e d t o n g p e r k g p e r d a y 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 30 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 31 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. This document is a draftfor review purposes only and does not constitute Agency policy. E-293 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Dependent variable = Mean 4 Independent variable = Dose 5 Data are assumed to be distributed: normally 6 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 7 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) * rho) 8 9 Total number of dose groups = 4 10 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 11 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 12 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 13 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 15 M L E s o l u t i o n p r o v i d e d : E x a c t 16 17 18 I n i t i a l P a r a m e t e r V a l u e s 19 20 Variable Model 2 21 22 lnalpha -3.57509 23 rho 2.23456 24 a 3 . 8 3 1 4 1 25 b 0 . 0 2 7 7 8 2 2 26 c 0 27 d 1 28 29 30 31 P a r a m e t e r E s t i m a t e s 32 33 Variable Model 2 34 35 lnalpha -1.55886 36 rho 1.77962 37 a 4 . 1 7 2 6 8 38 b 0 . 0 2 7 0 4 1 5 39 c 0 40 d 1 41 42 43 T a b l e o f S t a t s F r o m I n p u t D a t a 44 45 Dose N Obs Mean Obs Std Dev 46 47 04 2.27 0.49 48 1.43 4 5.55 0.85 49 14.3 3 7.62 1.79 50 143 3 196.9 63.14 51 52 53 E s t i m a t e d V a l u e s o f I n t e r e s t 54 55 Dose Est Mean Est Std Scaled Residual 56 57 0 4.173 1.635 -2.327 58 1.43 4.337 1.692 1.433 59 14.3 6.143 2.307 1.109 60 143 199.4 51.04 -0.08645 61 62 63 64 O t h e r m o d e l s ; f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d 65 66 M o d e l A1: Yij = M u ( i ) + e( ij) 67 V a r { e ( i j ) } = S i g m a ^ 2 68 69 M o d e l A2: Yij = M u ( i ) + e( ij) 70 V a r { e ( i j ) } = S i g m a ( i ) ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-294 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A3: Yij Mu(i) + e(ij) 3 Var{e(ij)} e x p ( l a l p h a + log(mean(i)) * rho) 4 5 M o d e l R: Yij M u + e(i) 6 Var{e(ij)} Sigma^2 7 8 9 Likelihoods of Interest 10 11 Model Log(likelihood) DF AIC 12 13 A1 -51.42175 5 112.8435 14 A2 -15.31211 8 46.62422 15 A3 -15.66963 6 43.33925 16 R -68.75058 2 141.5012 17 2 -25.37651 4 58.75302 18 19 20 Additive constant for all log-likelihoods = -12.87. This constant added to the 21 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 22 d e p e n d o n t h e m o d e l p a r a m e t e r s . 23 24 25 E x p l a n a t i o n o f T e s t s 26 27 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 28 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 29 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 30 T e s t 4: D o e s M o d e l 2 f i t t h e d a t a ? (A3 vs. 2) 31 32 33 T e s t s o f I n t e r e s t 34 35 T e s t -2*log(Likelihood Ratio) D. F. p-value 36 37 T e s t 1 106.9 6 < 0.0001 38 T e s t 2 72.22 3 < 0.0001 39 T e s t 3 0.715 2 0.6994 40 T e s t 4 19.41 2 < 0.0001 41 42 43 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 44 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 45 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 46 47 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 48 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e . 49 50 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 51 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 52 53 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. M o d e l 2 m a y n o t a d e q u a t e l y 54 d e s c r i b e t h e d a t a ; y o u m a y w a n t t o c o n s i d e r a n o t h e r m o d e l . 55 56 57 B e n c h m a r k D o s e C o m p u t a t i o n s : 58 59 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 60 61 R i s k T y p e = E s t i m a t e d s t a n d a r d d e v i a t i o n s f r o m c o n t r o l 62 63 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 64 65 BMD = 12.2272 66 67 BMDL = 9.03732 This document is a draftfor review purposes only and does not constitute Agency policy. E-295 DRAFT--DO NOT CITE OR QUOTE 1 E.3.7.3. F igure f o r Selected M odel: E x p o n en tia l (M2) Exponential_beta Model 2 with 0.95 Confidence Level Mean Response 2 17:25 02/16 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-296 DRAFT--DO NOT CITE OR QUOTE 1 E .3.8. C rofton et al., 2005: Serum , T4 2 E .3 .8 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M 2) 8 <0.0001 518.241 2.136E+03 1.157E+03 Notes exponential (M 3) exponential (M4) b exponential (M 5) 8 < 0 .0 0 0 1 51 8 .2 4 1 2 .1 3 6 E + 0 3 1.157E +03 p o w er hit b oun d (d = 1) 7 0.957 476.204 5.633E+01 3.006E+01 7 0.957 4 7 6 .2 0 4 5.6 3 3 E + 0 1 3.0 0 6 E + 0 1 p o w er hit b oun d (d = 1) H ill 6 0.973 477.434 5.564E+01 2.590E+01 linear 8 <0.0001 523.518 4.246E+03 3.086E+03 polynom ial, 8degree 8 <0.0001 523.518 4.246E+03 3.086E+03 power 8 < 0 .0 0 0 1 5 2 3 .5 1 8 4 .2 4 6 E + 0 3 3.0 8 6 E + 0 3 p o w er b ound hit (p ow er = 1) power, unrestricted 7 0.030 489.670 2.179E+01 2.271E +00 unrestricted (power = 0.217) a C onstant variance m odel selected (p = 0 .7 6 4 7 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3.8.2. Outputfor Selected Model: Exponential (M4) 6 Crofton et al., 2005: Serum, T4 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 8 _ C r o f t o n _ 2 0 0 5 _ T 4 _ E x p C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 T u e F e b 16 1 7 : 2 6 : 0 1 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 30 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 31 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. This document is a draftfor review purposes only and does not constitute Agency policy. E-297 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Dependent variable = Mean 4 Independent variable = Dose 5 Data are assumed to be distributed: normally 6 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 7 r h o is s e t t o 0. 8 A constant v a r i a n c e model is fit. 9 10 T o t a l n u m b e r o f d o s e g r o u p s = 10 11 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 12 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 13 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 16 M L E s o l u t i o n p r o v i d e d : E x a c t 17 18 19 I n i t i a l P a r a m e t e r V a l u e s 20 21 Variable Model 4 22 23 lnalpha 5.47437 24 rho(S) 0 25 a 1 0 4 . 9 9 9 26 b 0 . 0 0 0 3 7 1 6 9 4 27 c 0 . 4 4 5 7 6 4 28 d 1 29 30 (S) = S p e c i f i e d 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 Variable Model 4 37 38 lnalpha 5.50283 39 rho 0 40 a 9 9 . 7 7 6 41 b 0 . 0 0 7 2 8 3 8 7 42 c 0 . 5 3 3 5 1 6 43 d 1 44 45 46 T a b l e o f S t a t s F r o m I n p u t D a t a 47 48 Dose N Obs Mean Obs Std 49 50 0 14 100 15.44 51 0.1 6 96.27 14.98 52 3 12 98.57 18.11 53 10 6 99.76 19.04 54 30 6 93.32 12.11 55 100 6 70.94 12.74 56 300 6 62.52 14.75 57 1000 6 52.68 22.73 58 3000 6 54.66 19.71 59 1e+004 4 49.15 11.15 60 61 62 E s t i m a t e d V a l u e s o f I n t e r e s t 63 64 Dose Est Mean Est Std Scaled Residual 65 66 0 99. 78 15.66 0.05325 67 0.1 99. 74 15.66 -0.5434 68 3 98. 77 15.66 -0.04357 69 10 96. 51 15.66 0.5085 70 30 90. 64 15.66 0.4195 This document is a draftfor review purposes only and does not constitute Agency policy. E-298 DRAFT--DO NOT CITE OR QUOTE 1 100 75.7 15.66 -0.744 2 300 58.47 15.66 0.6334 3 1000 53.26 15.66 -0.09133 4 3000 53.23 15.66 0.2237 5 1e+004 53.23 15.66 -0.5218 6 7 8 9 Other models for which likelihoods are calculated: 10 11 M o d e l A1: Yij = M u ( i ) + e(i j ) 12 V a r { e (ij)} = S i g m a ^ 2 13 14 M o d e l A2: Yij = M u ( i ) + e(i j ) 15 V a r { e (ij)} = S i g m a ( i ) ^ 2 16 17 M o d e l A3: Yij = M u ( i ) + e(i j ) 18 V a r { e (ij)} = e x p ( l a l p h a + l o g (mean(i)) rho) 19 20 M o d e l R: Yij = M u + e(i) 21 V a r { e (ij)} = S i g m a ^ 2 22 23 24 L i k e l i h o o d s o f I n t e r e s t 25 26 Model Log(likelihood) DF AIC 27 28 A1 -233.0774 11 48 8 . 1 5 4 9 29 A2 -230.2028 20 500.4056 30 A3 -233.0774 11 48 8 . 1 5 4 9 31 R -268.4038 2 540.8076 32 4 -234.1019 4 476.2038 33 34 35 Additive constant for all log-likelihoods = -66.16. This constant added to the 36 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 37 d e p e n d o n t h e m o d e l p a r a m e t e r s . 38 39 40 E x p l a n a t i o n o f T e s t s 41 42 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 43 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 44 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 45 46 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 47 48 49 T e s t s o f I n t e r e s t 50 51 T e s t -2*log(Likelihood Ratio) D. F. p-value 52 53 T e s t 1 76.4 18 < 0.0001 54 T e s t 2 5.749 9 0.7647 55 T e s t 3 5.749 9 0.7647 56 T e s t 6a 2.049 7 0.9571 57 58 59 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 60 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 61 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 62 63 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 64 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e . 65 66 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 67 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 68 69 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 70 t o a d e q u a t e l y d e s c r i b e t h e d a t a . This document is a draftfor review purposes only and does not constitute Agency policy. E-299 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Benchmark Dose Computations: 4 5 Specified Effect = 1.000000 6 7 Risk Type = Estimated standard deviations from control 8 9 Confidence Level = 0.950000 10 11 BMD = 56.3321 12 13 BMDL = 30.0635 14 15 16 E .3 .8 .3 . Figurefor Selected Model: Exponential (M4) Exponential_beta Model 4 with 0.95 Confidence Level Mean Response 17 18 This document is a draftfor review purposes only and does not constitute Agency policy. E-300 DRAFT--DO NOT CITE OR QUOTE 1 E .3.9. F ranc et al., 2001: S-D R ats, R elative L iver W eight 2 E .3 .9 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD (ng/kg-d) BMDL (ng/kg-d) Notes H ill 1 0 .7 9 7 2 3 6 .3 7 1 1.826E +01 5 .4 6 3 E + 0 0 n lo w er b ound hit (n = 1) exponential (M 2) 2 0.935 234.440 2.262E+01 1.757E+01 exponential (M 3) 2 0.9 3 5 2 3 4 .4 4 0 2 .2 6 2 E + 0 1 1.757E +01 p o w er hit boun d (d = 1) exponential (M 4) 1 0.797 236.371 1.827E+01 6.112E+00 exponential (M 5) 1 0 .7 9 7 2 3 6 .3 7 1 1.827E +01 6 .1 1 2 E + 0 0 p o w er hit boun d (d = 1) linear polynom ial, 3degree power b 2 2 2 0.967 234.372 1.861E+01 1.339E+01 0.967 234.372 1.861E+01 1.339E+01 0 .9 6 7 2 3 4 .3 7 2 1.861E +01 1.339E +01 p o w er bound hit (p ow er = 1) H ill, unrestricted 0 N /A 238.366 1.726E+01 2.022E+00 unrestricted (n = 0.965) power, unrestricted c 1 0.805 236.365 1.725E+01 2.003E +00 unrestricted (power = 0.962) a C onstant variance m odel selected (p = 0.1 0 7 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .9 .2 . Outputfor Selected Model: Power 6 Franc et al., 2001: S-D Rats, R elative L iver W eight 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 8 8 _ F r a n c _ 2 0 0 1 _ S D _ R e l L i v W t _ P o w e r C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 8 8 _ F r a n c _ 2 0 0 1 _ S D _ R e l L i v W t _ P o w e r C V _ 1 . p l t 13 F r i A p r 16 1 6 : 2 8 : 4 5 2 0 1 0 14 15 16 F i g u r e 5, S D r a t s , r e l a t i v e l i v e r w e i g h t 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 This document is a draftfor review purposes only and does not constitute Agency policy. E-301 DRAFT--DO NOT CITE OR QUOTE 1 The power is r estricted to be greater than or equal to 1 2 A constant v ariance model is fit 3 4 Total number of dose groups = 4 5 Total number of records with missing values = 0 6 Maximum number of iterations = 250 7 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 8 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 9 10 11 12 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 13 alpha = 527.447 14 rho = 0 Specified 15 control = 100 16 slope = 1.15946 17 power = 0.839423 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( *** The model parameter(s) -rho -power 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 24 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 25 26 alpha control slope 27 28 a l p h a 1 1.3e-012 -6.2e-013 29 30 control 1.3e-012 1 -0.67 31 32 slope -6.2e-013 -0.67 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 39 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 40 alpha 462.485 115.621 235.872 689.099 41 control 101.047 5.10511 91.0415 111.053 42 slope 0.542984 0.0973507 0.352181 0.733788 43 p o w e r 1 NA 44 45 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 46 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 47 h a s n o s t a n d a r d e r r o r . 48 49 50 51 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 52 53 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Re 54 55 56 0 8 1 0 0 101 14 21.5 - 0 . 1 3 8 57 10 8 108 106 16.9 21.5 0.208 58 30 8 117 117 25.9 21.5 -0.0702 59 100 8 155 155 30.9 21.5 0.000298 60 61 62 63 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 64 65 66 M o d e l A1: Yij = Mu(i) + e(ij) 67 V a r { e i ;i j ) } = S i g m a ^ 2 68 69 M o d e l A2: Yij = Mu(i) + e(ij) 70 V a r { e i :i j ) } = S i g m a ( i ) ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-302 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A3: Yij = Mu(i) + e(ij) 3 Var{e(ij)} = Sigma^2 4 Model A3 uses any fixed variance parameters that 5 were specified by the user 6 7 M o d e l R: Yi = M u + e(i) 8 Var{e(i)} = Sigma^2 9 10 11 L i k e l i h o o d s o f I n t e r e s t 12 13 Model Log(likelihood) # Param's AIC 14 A1 -114.152281 5 238.304562 15 A2 -111.103649 8 238.207299 16 A3 -114.152281 5 238.304562 17 fitted -114.185827 3 234.371654 18 R -125.052064 2 254.104127 19 20 21 E x p l a n a t i o n o f T e s t s 22 23 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 24 (A2 vs. R) 25 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 26 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 27 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 28 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 29 30 T e s t s o f I n t e r e s t 31 32 Test -2*log(Likelihood Ratio) Test df p-val 33 34 T e s t 1 27.8968 6 <.0001 35 T e s t 2 6.09726 3 0.107 36 T e s t 3 6.09726 3 0.107 37 T e s t 4 0.0670927 2 0.967 38 39 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 40 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 41 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 42 43 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 44 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 45 46 47 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 48 t o b e a p p r o p r i a t e h e r e 49 50 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 51 t o a d e q u a t e l y d e s c r i b e t h e d a t a 52 53 54 B e n c h m a r k D o s e C o m p u t a t i o n 55 56 S p e c i f i e d e f f e c t = 0.1 57 58 R i s k T y p e = Relative risk 59 60 C o n f i d e n c e l e v e l = 0.95 61 62 B M D = 1 8 . 6 0 9 6 63 64 65 B M D L = 1 3 . 3 8 7 9 66 This document is a draftfor review purposes only and does not constitute Agency policy. E-303 DRAFT--DO NOT CITE OR QUOTE 1 E.3.9.3. F igure f o r Selected M odel: Pow er Power Model with 0.95 Confidence Level 2 3 4 5 E .3.9.4. Outputfor Additional Model Presented: Power, Unrestricted 6 Franc et al., 2001: S-D Rats, R elative L iver W eight 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 8 8 _ F r a n c _ 2 0 0 1 _ S D _ R e l L i v W t _ P o w e r C V _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 8 8 _ F r a n c _ 2 0 0 1 _ S D _ R e l L i v W t _ P o w e r C V _ U _ 1 . p l t 13 F r i A p r 16 1 6 : 2 8 : 4 6 2 0 1 0 14 15 16 F i g u r e 5, S D r a t s , r e l a t i v e l i v e r w e i g h t 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 T h e p o w e r is n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t 29 This document is a draftfor review purposes only and does not constitute Agency policy. E-304 DRAFT--DO NOT CITE OR QUOTE 1 Total number of dose groups = 4 2 Total number of records with missing values = 0 3 Maximum number of iterations = 250 4 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 5 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 6 7 8 9 Default Initial Parameter Values 10 alpha = 527.447 11 rho = 0S p e c i f i e d 12 control = 100 13 slope = 1.15946 14 power = 0.839423 15 16 17 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 18 19 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 20 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 21 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 22 23 alpha control slope power 24 25 a l p h a 1 1e-009 -6.2e-010 4.7e-010 26 27 c o n t r o l 1e-009 1 -0.74 0.71 28 29 slope -6.2e-010 -0.74 1 -1 30 31 power 4.7e-010 0.71 -1 1 32 33 34 35 P a r a m e t e r E s t i m a t e s 36 37 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 38 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 39 alpha 462.394 115.598 235.825 688.963 40 control 100.636 7.29156 86.3448 114.927 41 slope 0.650456 1.43713 -2.16627 3.46718 42 power 0.961853 0.465182 0.0501134 1.87359 43 44 45 46 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 47 48 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled R 49 50 51 0 8 1 0 0 101 14 21.5 -0.0836 52 10 8 108 107 16.9 21.5 0.192 53 30 8 117 118 25.9 21.5 -0.128 54 100 8 155 155 30.9 21.5 0.0192 55 56 57 58 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 59 60 61 M o d e l A1: Yij = Mu(i) + e(ij) 62 V a r { e i ;i j ) } = S i g m a ^ 2 63 64 M o d e l A2: Yij = Mu(i) + e(ij) 65 V a r { e i :i j ) } = S i g m a ( i ) ^2 66 67 M o d e l A3: Yij = Mu(i) + e(ij) 68 V a r { e i :i j ) } = S i g m a ^ 2 69 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 70 w e r e s p e c i f i e d b y t h e u s e r This document is a draftfor review purposes only and does not constitute Agency policy. E-305 DRAFT--DO NOT CITE OR QUOTE 1 2 Model R: Yi = M u + e(i) 3 Var{e(i)} = Sigma^2 4 5 6 Likelihoods of Interest 7 8 Model Log(likelihood) # Param's AIC 9 A1 -114.152281 5 238.304562 10 A2 -111.103649 8 238.207299 11 A3 -114.152281 5 238.304562 12 fitted -114.182670 4 236.365340 13 R -125.052064 2 254.104127 14 15 16 E x p l a n a t i o n o f T e s t s 17 18 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 19 (A2 vs. R) 20 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 21 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 22 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 23 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 24 25 T e s t s o f I n t e r e s t 26 27 Test -2*log(Likelihood Ratio) Test df p-val 28 29 T e s t 1 27.8968 6 <.0001 30 T e s t 2 6.09726 3 0.107 31 T e s t 3 6.09726 3 0.107 32 T e s t 4 0.0607785 1 0.8053 33 34 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 35 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 36 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 37 38 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 39 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 40 41 42 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 43 t o b e a p p r o p r i a t e h e r e 44 45 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 46 t o a d e q u a t e l y d e s c r i b e t h e d a t a 47 48 49 B e n c h m a r k D o s e C o m p u t a t i o n 50 51 S p e c i f i e d e f f e c t = 0.1 52 53 R i s k T y p e = Relative risk 54 55 C o n f i d e n c e l e v e l = 0.95 56 57 B M D = 1 7 . 2 4 6 9 58 59 60 B M D L = 2 . 0 0 3 3 6 61 This document is a draftfor review purposes only and does not constitute Agency policy. E-306 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E.3.9.5. F igure f o r A d d itio n a l M o d el Presented: Power, U nrestricted Power Model with 0.95 Confidence Level 2 16:28 04/16 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-307 DRAFT--DO NOT CITE OR QUOTE 1 E .3.10. F ranc et al., 2001: L-E R ats, R elative L iver W eight 2 E .3 .1 0 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2pValue AIC BMD (ng/kg-d) BMDL (ng/kg-d) Notes exponential (M 2) 2 0.245 210.148 5.143E+01 3.188E+01 exponential (M 3) 2 0.245 2 1 0 .1 4 8 5.1 4 3 E + 0 1 3.1 8 8 E + 0 1 p ow er hit b ound (d = 1) exponential (M 4) 1 0.607 209.599 1.476E+01 3.702E+00 exponential (M 5) 1 Hill b 1 0.607 0.703 2 0 9 .5 9 9 1.476E +01 3 .7 0 2 E + 0 0 p ow er hit b ound (d = 1) 2 0 9 .4 8 0 1.321E +01 1 .591E + 00 n lo w er b oun d hit (n = 1) linear 2 0.273 209.933 4.753E+01 2.788E+01 polynom ial, 3degree 1 <0.0001 10.000 1.505E+01 error power 2 0.273 2 0 9 .9 3 3 4 .7 5 3 E + 0 1 2 .7 8 8 E + 0 1 p ow er b oun d hit (p ow er = 1) H ill, unrestricted c0 N /A 211.341 1.163E+01 9.756E-01 unrestricted (n = 0.418) power, unrestricted 1 0.940 209.340 1.155E+01 1.513E-02 unrestricted (power = 0.394) a N on-constant variance m odel selected (p = 0 .0632) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .1 0 .2 . Outputfor Selected Model: Hill 6 Franc et al., 2001: L-E Rats, R elative Liver W eight 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 8 9 _ F r a n c _ 2 0 0 1 _ L E _ R e l L i v W t _ H i l l _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 8 9 _ F r a n c _ 2 0 0 1 _ L E _ R e l L i v W t _ H i l l _ 1 . p l t 13 F r i A p r 16 1 6 : 2 9 : 2 0 2 0 1 0 14 15 16 F i g u r e 5, L - E r a t s , r e l a t i v e l i v e r w e i g h t 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 P o w e r p a r a m e t e r r e s t r i c t e d t o b e g r e a t e r t h a n 1 This document is a draftfor review purposes only and does not constitute Agency policy. E-308 DRAFT--DO NOT CITE OR QUOTE 1 The variance is to be modeled as Var(i) = exp(lalpha + rho ln(mean(i))) 2 3 Total number of dose groups = 4 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 12 lalpha 5.41581 13 rho 0 14 intercept 100 15 2 2 . 2 2 5 16 0 . 3 2 9 5 2 6 17 4 0 . 8 4 0 3 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( * * * T h e m o d e l p a r a m e t e r ( s ) -n 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 24 a n d d o n o t a p p e a r in t h e c o r r e l a t i o n m a t r i x ) 25 26 lalpha rho intercept v k 27 CO <-- 1 o CO CO o 28 l a l p h a 1 -1 0.38 0.2 29 30 rh o -1 1 0.17 -0.38 -0.2 31 32 i n t e r c e p t -0.18 0.17 1 -0.13 0.39 33 34 v 0 . 3 8 -0.13 1 0.77 35 36 k 0.2 -0.2 0.39 0.77 1 37 38 39 40 P a r a m e t e r E s t i m a t e s 41 42 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 43 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 44 lalpha -15.3958 17.0376 -48.7889 17.9973 45 rho 4.38043 3.61867 -2.71204 11.4729 46 i n t e r c e p t 99.5667 3.7178 92.28 106.853 47 v 28.8965 12.6477 4.10739 53.6856 48 n 1 N A 49 k 25.1273 30.138 -33.9421 84.1966 50 51 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 52 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 53 h a s n o s t a n d a r d e r r o r . 54 55 56 57 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 58 59 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 60 61 62 0 8 1 0 0 99.6 10 10.8 0.114 63 10 8 106 108 17.9 12.8 -0.329 64 30 8 117 115 8.97 14.9 0.288 65 100 8 122 123 19.9 17 - 0 . 0 7 2 3 66 67 68 69 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-309 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A1: Yij Mu(i) + e(ij) 3 Var{e(ij)} Sigma^2 4 5 Model A2: Yij Mu(i) + e(ij) 6 Var{e(ij)} Sigma(i)^2 7 8 Model A3: Yij = Mu(i) + e(ij) 9 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 10 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 11 w e r e s p e c i f i e d b y t h e u s e r 12 13 M o d e l R: Yi = M u + e(i) 14 V a r { e ( i ) } = S i g m a ^ 2 15 16 17 L i k e l i h o o d s o f I n t e r e s t 18 19 Model Log(likelihood) # Param's AIC 20 A1 -100. 516456 5 211.032912 21 A2 -96. 870820 8 209.741641 22 A3 -99. 666984 6 211.333969 23 fitted -99. 739888 5 209.479776 24 R -105. 717087 2 215.434174 25 26 27 E x p l a n a t i o n o f T e s t s 28 29 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 30 (A2 vs. R) 31 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 32 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 33 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 34 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 35 36 T e s t s o f I n t e r e s t 37 38 Test -2*log(Likelihood Ratio) Test df p-value 39 40 T e s t 1 17.6925 6 0.007048 41 T e s t 2 7.29127 3 0.06317 42 T e s t 3 5.59233 2 0.06104 43 T e s t 4 0.145807 1 0.7026 44 45 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 46 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 47 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 48 49 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 50 m o d e l a p p e a r s t o b e a p p r o p r i a t e 51 52 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t t o c o n s i d e r a 53 d i f f e r e n t v a r i a n c e m o d e l 54 55 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 56 t o a d e q u a t e l y d e s c r i b e t h e d a t a 57 58 59 B e n c h m a r k D o s e C o m p u t a t i o n 60 61 S p e c i f i e d e f f e c t = 0.1 62 63 R i s k T y p e = Relative risk 64 65 C o n f i d e n c e l e v e l = 0.95 66 67 BMD = 13.2094 68 69 BMDL = 1.59127 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-310 DRAFT--DO NOT CITE OR QUOTE 1 E .3 .1 0 .3 . Figurefor Selected Model: Hill Hill Model with 0.95 Confidence Level 2 3 4 5 E .3.10.4. Outputfor Additional Model Presented: Hill, Unrestricted 6 Franc et al., 2001: L-E Rats, R elative Liver W eight 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 8 9 _ F r a n c _ 2 0 0 1 _ L E _ R e l L i v W t _ H i l l _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 8 9 _ F r a n c _ 2 0 0 1 _ L E _ R e l L i v W t _ H i l l _ U _ 1 . p l t 13 _ F r i A p r 16 1 6 : 2 9 7 2 7 2 0 1 0 14 15 16 F i g u r e 5, L - E r a t s , r e l a t i v e l i v e r w e i g h t 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 P o w e r p a r a m e t e r is n o t r e s t r i c t e d 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + r h o * l n ( m e a n ( i ) ) ) 28 29 T o t a l n u m b e r o f d o s e g r o u p s = 4 This document is a draftfor review purposes only and does not constitute Agency policy. E-311 DRAFT--DO NOT CITE OR QUOTE 1 Total number of records with missing values = 0 2 Maximum number of iterations = 250 3 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 4 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 5 6 7 8 Default Initial Parameter Values 9 lalpha 5.41581 10 rho 0 11 intercept 100 12 2 2 . 2 2 5 13 0 . 3 2 9 5 2 6 14 4 0 . 8 4 0 3 15 16 17 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 18 19 lalpha rho intercept v n k 20 21 l a l p h a 1 -1 -0.21 -0.099 0.23 -0.13 22 23 rh o -1 1 0.21 0.099 -0.23 0.13 24 25 i n t e r c e p t -0.21 0.21 1 0.023 0.14 0.011 26 CO o CO o 27 v -0.099 0.099 0.023 1 1 28 o CO CO o CO CO 29 n 0.23 -0.23 0.14 1 30 31 k -0.13 0.13 0.011 1 1 32 33 34 35 P a r a m e t e r E s t i m a t e s 36 37 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 38 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 39 lalpha -18.8355 18.0637 -54.2397 16.5688 40 rho 5.1098 3.83743 -2.41144 12.631 41 i n t e r c e p t 99.526 3.53402 92.5994 106.453 42 v 286.422 4487.2 -8508.33 9081.17 43 n 0.418159 0.457476 -0.478477 1.31479 44 k 32981.9 1.52481e+006 -2.95559e+006 3. 0 2 1 5 5 e + 0 0 6 45 46 47 48 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 49 50 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 51 52 53 0 8 1 0 0 99.5 10 10.3 0.13 54 10 8 106 109 17.9 13 -0.563 55 30 8 117 114 8.97 14.6 0.529 56 100 8 122 123 19.9 17.7 -0.0942 57 58 D e g r e e s o f f r e e d o m f o r T e s t A 3 v s f i t t e d < = 0 59 60 61 62 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 63 64 65 M o d e l A1: Yij Mu(i) + e(ij) 66 Var{e(ij)} Sigma^2 67 68 M o d e l A2: Yij Mu(i) + e(ij) 69 Var{e(ij)} Sigma(i)^2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-312 DRAFT--DO NOT CITE OR QUOTE 1 Model A3: Yij = Mu(i) + e(ij) 2 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 3 Model A3 uses any fixed variance parameters that 4 were specified by the user 5 6 M o d e l R: Yi = M u + e(i) 7 Var{e(i)} = Sigma^2 8 9 10 L i k e l i h o o d s o f I n t e r e s t 11 12 Model Log(likelihood) # Param's AIC 13 A1 -100.516456 5 211.032912 14 A2 -96.870820 8 209.741641 15 A3 -99.666984 6 211.333969 16 fitted -99.670736 6 211.341472 17 R -105.717087 2 215.434174 18 19 20 E x p l a n a t i o n o f T e s t s 21 22 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 23 (A2 vs. R) 24 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 25 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 26 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 27 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 28 29 T e s t s o f I n t e r e s t 30 31 Test -2*log(Likelihood Ratio) Test df p-value 32 33 T e s t 1 17.6925 6 0.007048 34 T e s t 2 7.29127 3 0.06317 35 T e s t 3 5.59233 2 0.06104 36 T e s t 4 0.00750301 0 NA 37 38 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 39 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 40 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 41 42 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 43 m o d e l a p p e a r s t o b e a p p r o p r i a t e 44 45 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t t o c o n s i d e r a 46 d i f f e r e n t v a r i a n c e m o d e l 47 48 N A D e g r e e s o f f r e e d o m f o r T e s t 4 a r e l e s s t h a n o r e q u a l t o 0. T h e C h i - S q u a r e 49 t e s t f o r f i t is n o t v a l i d 50 51 52 B e n c h m a r k D o s e C o m p u t a t i o n 53 54 S p e c i f i e d e f f e c t = 0.1 55 56 R i s k T y p e = Relative risk 57 58 C o n f i d e n c e l e v e l 0.95 59 60 BMD 11.6342 61 62 BMDL 0.975601 63 This document is a draftfor review purposes only and does not constitute Agency policy. E-313 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E.3.10.5. F igure f o r A d d itio n a l M o d el Presented: H ill, U nrestricted Hill Model with 0.95 Confidence Level 2 16:29 04/16 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-314 DRAFT--DO NOT CITE OR QUOTE 1 E.3.11. Franc et al., 2001: S-D Rats, Relative Thymus Weight 2 E.3.11.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD (ng/kg-d) BMDL (ng/kg-d) Notes exponential (M 2) 2 0.551 285.890 6.730E+00 3.627E+00 exponential (M 3) 1 exponential (M4) b 1 exponential (M 5) 0 <0.0001 303.995 3.858E+02 6.615E-01 0.972 286.698 3.559E+00 1.714E+00 N /A 288.696 3.796E+00 1.714E+00 H ill 0 N /A 288.696 4.299E+00 9.311E-01 linear 2 0.252 287.456 1.330E+01 1.062E+01 polynom ial, 3degree c 2 0.252 287.456 1.330E+01 1.062E+01 power 2 0.252 2 8 7 .4 5 6 1.330E +01 1.062E +01 p o w er b ound hit (p ow er = 1) power, unrestricted 1 0.510 287.131 5.049E-01 4.411E -04 unrestricted (power = 0.388) a N on-constant variance m odel selected (p = 0 .0320) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .1 1 .2 . Outputfor Selected Model: Exponential (M4) 6 Franc et al., 2001: S-D Rats, R elative Thym us W eight 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 9 1 _ F r a n c _ 2 0 0 1 _ S D _ R e l T h y W t _ E x p _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 F r i A p r 16 1 6 : 3 0 : 0 7 2 0 1 0 14 15 16 F i g u r e 5, S D r a t s , r e l a t i v e t h y m u s w e i g h t 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. This document is a draftfor review purposes only and does not constitute Agency policy. E-315 DRAFT--DO NOT CITE OR QUOTE 1 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 2 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 3 4 5 Dependent variable = Mean 6 Independent variable = Dose 7 Data are assumed to be distributed: normally 8 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 9 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) * rho) 10 11 T o t a l n u m b e r o f d o s e g r o u p s = 4 12 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 13 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 14 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 16 17 M L E s o l u t i o n p r o v i d e d : E x a c t 18 19 20 I n i t i a l P a r a m e t e r V a l u e s 21 22 Variable Model 4 23 24 lnalpha 3.35464 25 rho 1.08199 26 a 1 0 5 27 b 0 . 0 4 2 4 3 6 1 28 c 0 . 2 0 6 7 2 6 29 d 1 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 Variable Model 4 36 37 lnalpha 2.54324 38 rho 1.25901 39 a 1 0 8 . 9 0 4 40 b 0 . 0 3 7 9 3 4 3 41 c 0 . 2 0 8 1 4 6 42 d 1 43 44 45 T a b l e o f S t a t s F r o m I n p u t D a t a 46 47 ise N Obs Mean Obs Std 48 49 08 100 83 .2 50 10 8 91.17 47. 97 51 30 8 51.41 43. 48 52 100 8 22.79 29. 98 53 54 55 E s t i m a t e d V a l u e s o f I n t e r e s t 56 57 Dose Est Mean Est Std Scaled Residual 58 59 0 108.9 68.33 -0.3686 60 10 81.68 57.01 0.4706 61 30 50.3 42.02 0.0748 62 100 24.61 26.79 -0.192 63 64 65 66 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d 67 68 M o d e l A1: Yij = M u ( i ) + e(i j ) 69 V a r { e ( i j ) } = S i g m a ^ 2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-316 DRAFT--DO NOT CITE OR QUOTE 1 Model A2: Yij Mu(i) + e(ij) 2 Var{e(ij)} Sigma(i)^2 3 4 Model A3: Yij Mu(i) + e(ij) 5 Var{e(ij)} e x p ( l a l p h a + log(mean(i)) * rho) 6 7 M o d e l R: Yij M u + e(i) 8 Var{e(ij)} Sigma^2 9 10 11 L i k e l i h o o d s o f I n t e r e s t 12 13 Model Log(likelihood) DF AIC 14 15 A1 -141.9834 5 293.9669 16 A2 -137.5818 8 291.1637 17 A3 -138.3482 6 288.6964 18 R -146.9973 2 297.9946 19 4 -138.3488 5 286.6976 20 21 22 Additive constant for all log-likelihoods = -29.41. This constant added to the 23 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 24 d e p e n d o n t h e m o d e l p a r a m e t e r s . 25 26 27 E x p l a n a t i o n o f T e s t s 28 29 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 30 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 31 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 32 33 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 34 35 36 T e s t s o f I n t e r e s t 37 38 T e s t -2*log(Likelihood Ratio) D. F. p-value 39 40 T e s t 1 18.83 6 0.004459 41 T e s t 2 8.803 3 0.03203 42 T e s t 3 1.533 2 0.4647 43 T e s t 6a 0.001216 1 0.9722 44 45 46 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 47 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 48 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 49 50 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 51 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e . 52 53 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 54 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 55 56 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 57 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 58 59 60 B e n c h m a r k D o s e C o m p u t a t i o n s : 61 62 S p e c i f i e d E f f e c t = 0 . 1 0 0 0 0 0 63 64 R i s k T y p e = R e l a t i v e d e v i a t i o n 65 66 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 67 68 BMD = 3.55883 69 70 BMDL = 1.71399 This document is a draftfor review purposes only and does not constitute Agency policy. E-317 DRAFT--DO NOT CITE OR QUOTE 1 E .3 .1 1 .3 . Figurefor Selected Model: Exponential (M4) Exponential_beta Model 4 with 0.95 Confidence Level Mean Response 2 3 4 5 E .3.11.4. Outputfor Additional Model Presented: Polynomial, 3-Degree 6 Franc et al., 2001: S-D Rats, R elative Thym us W eight 7 8 9 10 P o l y n o m i a l M o d e l . ( V e r s i o n : 2 . 1 3 ; D a t e : 0 4 / 0 8 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 9 1 _ F r a n c _ 2 0 0 1 _ S D _ R e l T h y W t _ P o l y _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 9 1 _ F r a n c _ 2 0 0 1 _ S D _ R e l T h y W t _ P o l y _ 1 . p l t 13 _ F r i A p r 16 1 6 : 3 0 7 1 1 2 0 1 0 14 15 16 F i g u r e 5, S D r a t s , r e l a t i v e t h y m u s w e i g h t 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 T h e p o l y n o m i a l c o e f f i c i e n t s a r e r e s t r i c t e d t o b e n e g a t i v e 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 28 29 T o t a l n u m b e r o f d o s e g r o u p s = 4 This document is a draftfor review purposes only and does not constitute Agency policy. E-318 DRAFT--DO NOT CITE OR QUOTE 1 Total number of records with missing values = 0 2 Maximum number of iterations = 250 3 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 4 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 5 6 7 8 Default Initial Parameter Values 9 lalpha = 8.0075 10 rho = 0 11 beta_0 = 100 12 beta_1 = -0.352259 13 beta_2 = -0.0585481 14 beta 3 = 0 15 16 17 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 18 19 ( *** The model parameter(s) -beta 2 -beta 3 20 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 21 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 22 23 lalpha rho beta 0 beta 1 24 25 l a l p h a 1 -0.99 0.031 -0.016 26 27 rho -0.99 1 -0.034 0.022 28 'tr CO o CO o 29 b e t a 0 0.031 -0.034 1 30 31 beta 1 -0.016 0.022 1 32 33 34 35 P a r a m e t e r E s t i m a t e s 36 37 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 38 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 39 lalpha 2.92328 1.7394 -0.485884 6.33243 40 rho 1.18295 0.423359 0.353177 2.01271 41 beta 0 89.841 13.7418 62.9076 116.774 42 beta 1 -0.675682 0.175538 -1.01973 -0.331634 43 b e t a 2 0 NA 44 b e t a 3 0 NA 45 46 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 47 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 48 h a s n o s t a n d a r d e r r o r . 49 50 51 52 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 53 54 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 55 56 57 0 8 1 0 0 89.8 83.2 61.7 0.466 58 10 8 91.2 83.1 48 58.9 0.388 59 30 8 51.4 69.6 43.5 53 -0.968 60 100 8 22.8 22.3 30 27 0 . 0 5 4 3 61 62 63 64 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 65 66 67 M o d e l A1: Yij = Mu(i) + e(ij) 68 V a r { e ( i j ) } = S i g m a ^ 2 69 70 M o d e l A2: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-319 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma(i)^2 2 3 Model A3: Yij = Mu(i) + e(ij) 4 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 5 Model A3 uses any fixed variance parameters that 6 were specified by the user 7 8 M o d e l R: Yi = M u + e(i) 9 Var{e(i)} = Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) # Param's AIC 15 A1 -141.983433 5 293.966865 16 A2 -137.581833 8 291.163667 17 A3 -138.348184 6 288.696368 18 fitted -139.728204 4 287.456407 19 R -146.997301 2 297.994602 20 21 22 E x p l a n a t i o n o f T e s t s 23 24 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 25 (A2 vs. R) 26 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 27 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 28 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 29 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 30 31 T e s t s o f I n t e r e s t 32 33 Test -2*log(Likelihood Ratio) Test df p-value 34 35 T e s t 1 18.8309 6 0.004459 36 T e s t 2 8.8032 3 0.03203 37 T e s t 3 1.5327 2 0.4647 38 T e s t 4 2.76004 2 0.2516 39 40 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 41 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 42 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 43 44 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 45 m o d e l a p p e a r s t o b e a p p r o p r i a t e 46 47 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 48 t o b e a p p r o p r i a t e h e r e 49 50 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 51 t o a d e q u a t e l y d e s c r i b e t h e d a t a 52 53 54 B e n c h m a r k D o s e C o m p u t a t i o n 55 56 S p e c i f i e d e f f e c t = 0.1 57 58 R i s k T y p e = Relative risk 59 60 C o n f i d e n c e l e v e l = 0.95 61 62 BMD = 13.2963 63 64 65 BMDL = 10.6163 66 This document is a draftfor review purposes only and does not constitute Agency policy. E-320 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E.3.11.5. F igure f o r A d d itio n a l M o d el Presented: Polynom ial, 3-D egree Polynomial Model with 0.95 Confidence Level 2 16:30 04/16 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-321 DRAFT--DO NOT CITE OR QUOTE 1 E.3.12. Franc et al., 2001: L-E Rats, Relative Thymus Weight 2 E.3.12.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD (ng/kg-d) BMDL (ng/kg-d) Notes exponential (M 2) 2 0.394 301.666 6.406E+00 2.122E+00 exponential (M 3) 2 exponential (M4) b 1 exponential (M 5) 0 0 .3 9 4 3 0 1 .6 6 6 6 .4 0 6 E + 0 0 2 .1 2 2 E + 0 0 p o w er hit bound (d = 1) 0.317 302.808 3.520E+00 1.067E+00 N /A 303.805 1.280E+01 1.450E+00 H ill 0 N /A 303.805 1.195E+01 9.965E-01 linear 2 0.236 302.690 1.429E+01 9.087E +00 polynom ial, 3degree 2 0.236 302.690 1.429E+01 9.087E +00 power 2 0 .2 3 6 3 0 2 .6 9 0 1.429E +01 9 .0 8 7 E + 0 0 p o w er b oun d hit (p ow er = 1) power, unrestricted 1 0.175 303.643 1.297E+00 2.703E -08 unrestricted (pow er = 0.454) a C onstant variance m odel selected (p = 0 .5 0 6 3 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3 .1 2 .2 . Outputfor Selected Model: Exponential (M4) 6 Franc et al., 2001: L-E Rats, R elative Thym us W eight 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 9 2 _ F r a n c _ 2 0 0 1 _ L E _ R e l T h y W t _ E x p C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 F r i A p r 16 1 6 : 3 0 : 5 8 2 0 1 0 14 15 16 F i g u r e 5, L - E r a t s , r e l a t i v e t h y m u s w e i g h t 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 30 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 31 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. This document is a draftfor review purposes only and does not constitute Agency policy. E-322 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Dependent variable = Mean 4 Independent variable = Dose 5 Data are assumed to be distributed: normally 6 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 7 r h o is s e t t o 0. 8 A constant v a r i a n c e model is fit. 9 10 T o t a l n u m b e r o f d o s e g r o u p s = 4 11 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 12 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 13 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 16 M L E s o l u t i o n p r o v i d e d : E x a c t 17 18 19 I n i t i a l P a r a m e t e r V a l u e s 20 21 Variable Model 4 22 23 lnalpha 8.1814 24 rho(S) 0 25 a 1 0 5 26 b 0 . 0 4 1 3 9 4 5 27 c 0 . 3 1 7 3 28 d 1 29 30 (S) = S p e c i f i e d 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 Variable Model 4 37 38 lnalpha 8.21275 39 rho 0 40 a 1 0 6 . 5 7 41 b 0 . 0 4 2 5 9 6 7 42 c 0 . 2 8 1 8 9 43 d 1 44 45 46 T a b l e o f S t a t s F r o m I n p u t D a t a 47 48 ise N Obs Mean Obs Std 49 50 08 100 54. 72 51 10 8 95.41 70. 46 52 30 8 38.69 47. 97 53 100 8 34.98 77. 96 54 55 56 E s t i m a t e d V a l u e s o f I n t e r e s t 57 58 Dose Est Mean Est Std Scaled Residual 59 60 0 106.6 60.73 -0.306 61 10 80.03 60.73 0.7164 62 30 51.36 60.73 -0.5902 63 100 31.12 60.73 0.1798 64 65 66 67 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d 68 69 M o d e l A1: Yij = M u ( i ) + e(i j ) 70 V a r { e ( i j ) } = S i g m a ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-323 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A2: Yij Mu(i) + e(ij) 3 Var{e(ij)} Sigma(i)^2 4 5 Model A3: Yij Mu(i) + e(ij) 6 Var{e(ij)} e x p ( l a l p h a + log(mean(i)) * rho) 7 8 M o d e l R: Yij M u + e(i) 9 Var{e(ij)} Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) DF AIC 15 16 A1 -146.9024 5 303.8049 17 A2 -145.7361 8 307.4723 18 A3 -146.9024 5 303.8049 19 R -150.6049 2 305.2098 20 4 -147.404 4 302.8079 21 22 23 Additive constant for all log-likelihoods = -29.41. This constant added to the 24 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 25 d e p e n d o n t h e m o d e l p a r a m e t e r s . 26 27 28 E x p l a n a t i o n o f T e s t s 29 30 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 31 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 32 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 33 34 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 35 36 37 T e s t s o f I n t e r e s t 38 39 T e s t -2*log(Likelihood Ratio) D. F. p-value 40 41 T e s t 1 9.738 6 0.1362 42 T e s t 2 2.333 3 0.5063 43 T e s t 3 2.333 3 0.5063 44 T e s t 6a 1.003 1 0.3166 45 46 47 T h e p - v a l u e f o r T e s t 1 is g r e a t e r t h a n .05. T h e r e m a y n o t b e a 48 d i f f e n c e b e t w e e n r e s p o n s e s a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 49 M o d e l l i n g t h e d a t a w i t h a d o s e / r e s p o n s e c u r v e m a y n o t b e a p p r o p r i a t e . 50 51 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 52 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e . 53 54 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 55 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 56 57 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 58 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 59 60 61 B e n c h m a r k D o s e C o m p u t a t i o n s : 62 63 S p e c i f i e d E f f e c t = 0 . 1 0 0 0 0 0 64 65 R i s k T y p e = R e l a t i v e d e v i a t i o n 66 67 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 68 69 BMD = 3.52038 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-324 DRAFT--DO NOT CITE OR QUOTE 1 BMDL = 1.06729 2 E .3 .1 2 .3 . Figurefor Selected Model: Exponential (M4) Exponential_beta Model 4 with 0.95 Confidence Level Mean Response 3 16:30 04/16 2010 4 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-325 DRAFT--DO NOT CITE OR QUOTE 1 E.3.13. Franc et al., 2001: H/W Rats, Relative Thymus Weight 2 E.3.13.1. S u m m a ry Table o f B M D S M odeling R esults Model a exponential (M 2) c Degrees of Freedom x2p Value 2 0.682 AIC 261.694 BMD (ng/kg-d) BMDL (ng/kg-d) Notes 1.366E+01 8.014E +00 exponential (M 3) 2 exponential (M4) b 1 exponential (M 5) 0 0 .6 8 2 2 6 1 .6 9 4 1.366E +01 8 .0 1 4 E + 0 0 p o w er hit bound (d = 1) 0.512 263.358 8.820E+00 3.219E+00 N /A 264.927 1.776E+01 3.500E +00 H ill 0 N /A 264.927 1.701E+01 2.729E +00 linear 2 0.543 262.148 1.919E+01 1.373E+01 polynom ial, 3degree 2 0.543 262.148 1.919E+01 1.373E+01 power 2 0.543 2 6 2 .1 4 8 1.919E +01 1.373E +01 p o w er b oun d hit (p ow er = 1) power, unrestricted 1 0.381 263.694 8.127E+00 1.406E-01 unrestricted (pow er = 0.665) a C onstant variance m odel selected (p = 0 .4 3 3 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .1 3 .2 . Outputfor Selected Model: Exponential (M2) 6 Franc et al., 2001: H /W Rats, R elative Thym us W eight 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 9 3 _ F r a n c _ 2 0 0 1 _ H W _ R e l T h y W t _ E x p C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 F r i A p r 16 1 6 : 3 1 : 4 0 2 0 1 0 14 15 16 F i g u r e 5, H / W r a t s , r e l a t i v e t h y m u s w e i g h t 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. This document is a draftfor review purposes only and does not constitute Agency policy. E-326 DRAFT--DO NOT CITE OR QUOTE 1 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 2 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 3 4 5 Dependent variable = Mean 6 Independent variable = Dose 7 Data are assumed to be distributed: normally 8 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 9 r h o is s e t t o 0. 10 A c o n s t a n t v a r i a n c e m o d e l is fit. 11 12 T o t a l n u m b e r o f d o s e g r o u p s = 4 13 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 14 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 15 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 16 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 17 18 M L E s o l u t i o n p r o v i d e d : E x a c t 19 20 21 I n i t i a l P a r a m e t e r V a l u e s 22 23 Variable Model 2 24 25 lnalpha 6.96647 26 rho(S) 0 27 a 5 9 . 5 0 8 4 28 b 0 . 0 0 7 1 5 4 5 8 29 c 0 30 d 1 31 32 (S) = S p e c i f i e d 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 Variable Model 2 39 40 lnalpha 6.99043 41 rho 0 42 a 9 9 . 7 7 6 1 43 b 0 . 0 0 7 7 1 3 4 1 44 c 0 45 d 1 46 47 48 T a b l e o f S t a t s F r o m I n p u t D a t a 49 50 Dose N Obs Mean Obs Std Dev 51 52 08 100 35.98 53 10 8 97.53 32.98 54 30 8 71.02 23.99 55 100 8 49.29 43.48 56 57 58 E s t i m a t e d V a l u e s o f I n t e r e s t 59 60 Dose Est Mean Est Std Scaled Residual 61 62 0 99.78 32.96 0.01921 63 10 92.37 32.96 0.4426 64 30 79.16 32.96 -0.6986 65 100 46.14 32.96 0.271 66 67 68 69 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d : 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-327 DRAFT--DO NOT CITE OR QUOTE 1 Model A1: Yij = Mu(i) + e(ij) 2 V a r { e i ;ij)} = S i g m a ^ 2 3 4 Model A2: Yij = Mu(i) + e(ij) 5 V a r { e i :ij)} = S i g m a ( i ) ^ 2 6 7 Model A3: Yij = Mu(i) + e(ij) 8 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 9 10 M o d e l R: Yij = M u + e(i) 11 V a r { e i ij)} = S i g m a ^ 2 12 13 14 L i k e l i h o o d s o f I n t e r e s t 15 16 Model Log(likelihood) DF AIC 17 18 A1 -127.4636 5 264.9271 19 A2 -126.0925 8 268.185 20 A3 -127.4636 5 264.9271 21 R -132.935 2 269.87 22 2 -127.8469 3 261.6939 23 24 25 Additive constant for all log-likelihoods = -29.41. This constant added to the 26 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 27 d e p e n d o n t h e m o d e l p a r a m e t e r s . 28 29 30 E x p l a n a t i o n o f T e s t s 31 32 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 33 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 34 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 35 T e s t 4: D o e s M o d e l 2 f i t t h e d a t a ? (A3 vs. 2) 36 37 38 T e s t s o f I n t e r e s t 39 40 T e s t -2*log(Likelihood Ratio) D. F. p-value 41 42 T e s t 1 13.69 6 0.03336 43 T e s t 2 2.742 3 0.4331 44 T e s t 3 2.742 3 0.4331 45 T e s t 4 0.7668 2 0.6815 46 47 48 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 49 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 50 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 51 52 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 53 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e . 54 55 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 56 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 57 58 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. M o d e l 2 s e e m s 59 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 60 61 62 B e n c h m a r k D o s e C o m p u t a t i o n s : 63 64 S p e c i f i e d E f f e c t = 0 . 1 0 0 0 0 0 65 66 R i s k T y p e = R e l a t i v e d e v i a t i o n 67 68 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 69 70 BMD = 13.6594 This document is a draftfor review purposes only and does not constitute Agency policy. E-328 DRAFT--DO NOT CITE OR QUOTE 1 2 BMDL = 8.01373 3 E .3 .1 3 .3 . Figurefor Selected Model: Exponential (M2) Exponential_beta Model 2 with 0.95 Confidence Level Mean Response 4 5 6 7 E .3.13.4. Outputfor Additional Model Presented: Exponential (M4) 8 Franc et al., 2001: H /W Rats, R elative Thym us W eight 9 10 11 12 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 13 I n p u t D a t a F i l e : C : \ 1 \ 9 3 _ F r a n c _ 2 0 0 1 _ H W _ R e l T h y W t _ E x p C V _ 1 . ( d ) 14 G n u p l o t P l o t t i n g Fil e : 15 F r i A p r 16 1 6 : 3 1 : 4 0 2 0 1 0 16 17 18 F i g u r e 5, H / W r a t s , r e l a t i v e t h y m u s w e i g h t 19 20 21 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 22 M o d e l 2: Y dose] = a * exp{sign * b * dose} 23 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 24 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 25 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 26 27 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 28 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 29 s i g n = -1 f o r d e c r e a s i n g t r e n d . This document is a draftfor review purposes only and does not constitute Agency policy. E-329 DRAFT--DO NOT CITE OR QUOTE 1 2 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 3 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 4 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 5 6 7 Dependent variable = Mean 8 Independent variable = Dose 9 Data are assumed to be distributed: normally 10 V a r i a n c e M o d e l : e x p ( l n a l p h a + r h o * l n ( Y [ d o s e ] ) ) 11 r h o is s e t t o 0. 12 A c o n s t a n t v a r i a n c e m o d e l is fit. 13 14 T o t a l n u m b e r o f d o s e g r o u p s = 4 15 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 16 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 17 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 18 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 19 20 M L E s o l u t i o n p r o v i d e d : E x a c t 21 22 23 I n i t i a l P a r a m e t e r V a l u e s 24 25 Variable Model 4 26 27 lnalpha 6.96647 28 rho(S) 0 29 a 1 0 5 30 b 0 . 0 3 1 6 9 31 c 0 . 4 4 7 1 0 5 32 d 1 33 34 (S) = S p e c i f i e d 35 36 37 38 P a r a m e t e r E s t i m a t e s 39 40 Variable Model 4 41 42 lnalpha 6.97993 43 rho 0 44 a 1 0 3 . 0 9 1 45 b 0 . 0 2 0 4 8 46 c 0 . 3 9 4 9 0 4 47 d 1 48 49 50 T a b l e o f S t a t s F r o m I n p u t D a t a 51 52 Dose N Obs Mean Obs Std Dev 53 54 08 100 35.98 55 10 8 97.53 32.98 56 30 8 71.02 23.99 57 100 8 49.29 43.48 58 59 60 E s t i m a t e d V a l u e s o f I n t e r e s t 61 62 Dose Est Mean Est Std Scaled Residu; 63 64 0 103.1 32.78 -0.2667 65 10 91.54 32.78 0.5166 66 30 74.46 32.78 -0.2961 67 100 48.76 32.78 0.04621 68 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-330 DRAFT--DO NOT CITE OR QUOTE 1 Other models for which likelihoods are calculated: 2 3 Model A1: Yij = Mu(i) + e(ij) 4 V a r { e i ;ij)} = S i g m a ^ 2 5 6 Model A2: Yij = Mu(i) + e(ij) 7 V a r { e i :ij)} = S i g m a ( i ) ^ 2 8 9 Model A3: Yij = Mu(i) + e(ij) 10 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 11 12 M o d e l R: Yij = M u + e(i) 13 V a r { e i ij)} = S i g m a ^ 2 14 15 16 L i k e l i h o o d s o f I n t e r e s t 17 18 Model Log(likelihood) DF AIC 19 20 A1 -127.4636 5 264.9271 21 A2 -126.0925 8 268.185 22 A3 -127.4636 5 264.9271 23 R -132.935 2 269.87 24 4 -127.6789 4 263.3577 25 26 27 Additive constant for all log-likelihoods = -29.41. This constant added to the 28 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 29 d e p e n d o n t h e m o d e l p a r a m e t e r s . 30 31 32 E x p l a n a t i o n o f T e s t s 33 34 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 35 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 36 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 37 38 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 39 40 41 T e s t s o f I n t e r e s t 42 43 T e s t -2*log(Likelihood Ratio) D. F. p-value 44 45 T e s t 1 13.69 6 0.03336 46 T e s t 2 2.742 3 0.4331 47 T e s t 3 2.742 3 0.4331 48 T e s t 6a 0.4306 1 0.5117 49 50 51 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 52 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 53 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 54 55 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 56 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e . 57 58 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 59 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 60 61 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 62 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 63 64 65 B e n c h m a r k D o s e C o m p u t a t i o n s : 66 67 S p e c i f i e d E f f e c t = 0 . 1 0 0 0 0 0 68 69 R i s k T y p e = R e l a t i v e d e v i a t i o n 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-331 DRAFT--DO NOT CITE OR QUOTE 1 Confidence Level = 0.950000 2 3 BMD = 8.82023 4 5 BMDL = 3.21928 6 E .3.13.5. Figurefor Additional Model Presented: Exponential (M4) Exponential_beta Model 4 with 0.95 Confidence Level Mean Response 7 This document is a draftfor review purposes only and does not constitute Agency policy. E-332 DRAFT--DO NOT CITE OR QUOTE 1 E .3.14. H ojo et al., 2002: D R L R einforce P er M inute 2 E .3 .1 4 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of X2 pFreedom Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) H ill 0 N /A 6.465 2.060E+01 1.713E-05 Notes linear b polynom ial, 3degree power 2 0.008 9.552 2.677E+02 1.100E+02 2 0.008 9.552 2.677E+02 1.100E+02 2 0 .0 0 8 9.5 5 2 2 .6 7 7 E + 0 2 1.100E + 02 p o w er bound hit (p ow er = 1) power, unrestricted 1 0.025 6.780 2.187E +00 4.612E -08 unrestricted (power = 0.089) exponential (M2) 2 0.006 9.894 3.043E+02 1.505E+02 exponential (M3) 2 0 .0 0 6 9.8 9 4 3 .0 4 3 E + 0 2 1.505E + 02 p o w er hit boun d (d = 1) exponential (M 4) c 1 0.062 5.241 1.734E+01 3.827E-02 exponential (M5) 0 N /A 6.465 2.140E+01 1.240E-05 a Constant variance m odel selected (p = 0.4 3 2 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .1 4 .2 . Outputfor Selected Model: Linear 6 H ojo et al., 2002: D R L R einforce Per M inute 7 8 9 10 P o l y n o m i a l M o d e l . ( V e r s i o n : 2 . 1 3 ; D a t e : 0 4 / 0 8 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 2 0 _ H o j o _ 2 0 0 2 _ D R L r e i n _ L i n e a r C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 2 0 _ H o j o _ 2 0 0 2 _ D R L r e i n _ L i n e a r C V _ 1 . p l t 13 ~ T u e F e b 16 1 7 : 2 9 : 4 2 2 0 1 0 14 15 16 T a b l e 5 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 S i g n s o f t h e p o l y n o m i a l c o e f f i c i e n t s a r e n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t 29 30 T o t a l n u m b e r o f d o s e g r o u p s = 4 This document is a draftfor review purposes only and does not constitute Agency policy. E-333 DRAFT--DO NOT CITE OR QUOTE 1 Total number of records with missing values = 0 2 Maximum number of iterations = 250 3 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 4 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 5 6 7 8 Default Initial Parameter Values 9 alpha = 0.337763 10 rho = 0S p e c i f i e d 11 beta_0 = -0.404 12 beta_1 = 0.00249615 13 14 15 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 16 17 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 18 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r 19 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 20 21 alpha beta_0 beta_1 22 23 a l p h a 1 -1.4e-008 2.2e-008 24 25 beta 0 -1.4e-008 1 -0.69 26 27 beta 1 2.2e-008 -0.69 1 28 29 30 31 P a r a m e t e r E s t i m a t e s 32 33 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 34 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 35 alpha 0.435671 0.134451 0.172152 0.69919 36 beta_0 -0.372098 0.198702 -0.761547 0.017352 37 beta 1 0.00246548 0.00211361 -0.00167711 0.00660807 38 39 40 41 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 42 43 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 44 45 46 0 5 -0.814 -0.372 0.448 0.66 -1.5 47 20 5 - 0 . 3 6 4 -0.323 0.821 0.66 -0.14 48 60 6 0.374 -0.224 0.54 0.66 2.22 49 1 8 0 5 - 0 . 1 6 3 0.0717 0.443 0.66 -0.795 50 51 52 53 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 54 55 56 M o d e l A1: Yij = Mu(i) + e(ij) 57 V a r { e i ;i j ) } = S i g m a ^ 2 58 59 M o d e l A2: Yij = Mu(i) + e(ij) 60 V a r { e i :i j ) } = S i g m a ( i ) ^2 61 62 M o d e l A3: Yij = Mu(i) + e(ij) 63 V a r { e i :i j ) } = S i g m a ^ 2 64 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 65 w e r e s p e c i f i e d b y t h e u s e r 66 67 M o d e l R: Yi = M u + e(i) 68 V a r { e ( i ) } = S i g m a ^ 2 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-334 DRAFT--DO NOT CITE OR QUOTE 1 Likelihoods of Interest 2 3 Model Log(likelihood) # Param's AIC 4 A1 3.115550 5 3.768900 5 A2 4.489557 8 7.020886 6 A3 3.115550 5 3.768900 7 fitted -1.775882 3 9.551763 8 R -2.435087 2 8.870174 9 10 11 E x p l a n a t i o n o f T e s t s 12 13 T e s t 1: D o r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 14 ( A 2 vs. R) 15 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? ( A 1 vs A2) 16 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 17 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 18 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 19 20 T e s t s o f I n t e r e s t 21 22 Test -2*log(Likelihood Ratio) Test df p-value 23 24 T e s t 1 13.8493 6 0.03137 25 T e s t 2 2.74801 3 0.4321 26 T e s t 3 2.74801 3 0.4321 27 T e s t 4 9.78286 2 0.007511 28 29 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 30 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 31 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 32 33 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 34 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 35 36 37 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 38 t o b e a p p r o p r i a t e h e r e 39 40 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 41 m o d e l 42 43 44 B e n c h m a r k D o s e C o m p u t a t i o n 45 46 S p e c i f i e d e f f e c t = 1 47 48 R i s k T y p e = Estimated standard deviations from the control mean 49 50 C o n f i d e n c e l e v e l = 0.95 51 52 BMD = 267.718 53 54 55 BMDL = 110.032 56 This document is a draftfor review purposes only and does not constitute Agency policy. E-335 DRAFT--DO NOT CITE OR QUOTE 1 E.3.14.3. F igure f o r Selected M odel: L in ea r Linear Model with 0.95 Confidence Level Mean Response 2 17:29 02/16 2010 3 dose 4 5 E .3.14.4. Outputfor Additional Model Presented: Exponential (M4) 6 H ojo et al., 2002: D R L R einforce Per M inute 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a Fil e : C : \ 1 \ 2 1 H o j o 2 0 0 2 D R L r e i n E x p C V 1.( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 T u e F e b 16 1 7 : 3 0 : 14 15 16 T a b l e 5, v a l u e s a d j u s t e d b y a c o n s t a n t t o a l l o w e x p o n e n t i a l m o d e l 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-336 DRAFT--DO NOT CITE OR QUOTE 1 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 2 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 3 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 4 5 6 Dependent variable = Mean 7 Independent variable = Dose 8 Data are assumed to be distributed: normally 9 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 10 r h o is s e t t o 0. 11 A c o n s t a n t v a r i a n c e m o d e l is fit. 12 13 T o t a l n u m b e r o f d o s e g r o u p s = 4 14 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 15 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 16 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 17 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 18 19 M L E s o l u t i o n p r o v i d e d : E x a c t 20 21 22 I n i t i a l P a r a m e t e r V a l u e s 23 24 Variable Model 4 25 26 lnalpha -1.29672 27 rho(S) 0 28 a 0 . 0 8 1 7 29 b 0 . 0 0 8 8 0 8 6 7 30 c 1 6 . 3 7 3 3 31 d 1 32 33 (S) = S p e c i f i e d 34 35 36 37 P a r a m e t e r E s t i m a t e s 38 39 Variable Model 4 40 41 lnalpha -1.13136 42 rho 0 43 a 0 . 0 5 4 2 8 6 8 44 b 0 . 0 5 2 5 0 1 6 45 c 1 8 . 5 0 7 2 46 d 1 47 48 49 T a b l e o f S t a t s F r o m I n p u t D a t a 50 51 Dose N Obs Mean Obs Std Dev 52 53 05 0.086 0.448 54 20 5 0.536 0.821 55 60 6 1.274 0.54 56 180 5 0.737 0.443 57 58 59 E s t i m a t e d V a l u e s o f I n t e r e s t 60 61 Dose Est Mean Est Std Scaled Residu 62 63 0 0.05429 0.568 0.1249 64 20 0.6721 0.568 -0.5359 65 60 0.964 0.568 1.337 66 180 1.005 0.568 -1.054 67 68 69 70 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d : This document is a draftfor review purposes only and does not constitute Agency policy. E-337 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A1: Yij = Mu(i) + e(ij) 3 V a r { e i ;ij)} = S i g m a ^ 2 4 5 Model A2: Yij = Mu(i) + e(ij) 6 V a r { e i :ij)} = S i g m a ( i ) ^ 2 7 8 Model A3: Yij = Mu(i) + e(ij) 9 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 10 11 M o d e l R: Yij = M u + e(i) 12 V a r { e i ij)} = S i g m a ^ 2 13 14 15 L i k e l i h o o d s o f I n t e r e s t 16 17 Model Log(likelihood) DF AIC 18 19 A1 3.11555 5 3.7689 20 A2 4.489557 8 7.020886 21 A3 3.11555 5 3.7689 22 R -2.435087 2 8.870174 23 4 1.379312 4 5.241376 24 25 26 Additive constant for all log-likelihoods = -19.3. This constant added to the 27 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 28 d e p e n d o n t h e m o d e l p a r a m e t e r s . 29 30 31 E x p l a n a t i o n o f T e s t s 32 33 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 34 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 35 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 36 37 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 38 39 40 T e s t s o f I n t e r e s t 41 42 T e s t -2*log(Likelihood Ratio) D. F. p-value 43 44 T e s t 1 13.85 6 0.03137 45 T e s t 2 2.748 3 0.4321 46 T e s t 3 2.748 3 0.4321 47 T e s t 6a 3.472 1 0.0624 48 49 50 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 51 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 52 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 53 54 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 55 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e . 56 57 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 58 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 59 60 T h e p - v a l u e f o r T e s t 6a is l e s s t h a n .1. M o d e l 4 m a y n o t a d e q u a t e l y 61 d e s c r i b e t h e d a t a ; y o u m a y w a n t t o c o n s i d e r a n o t h e r m o d e l . 62 63 64 B e n c h m a r k D o s e C o m p u t a t i o n s : 65 66 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 67 68 R i s k T y p e = E s t i m a t e d s t a n d a r d d e v i a t i o n s f r o m c o n t r o l 69 70 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 This document is a draftfor review purposes only and does not constitute Agency policy. E-338 DRAFT--DO NOT CITE OR QUOTE 1 2 BMD 17.3391 3 4 BMDL = 0.0382689 5 6 7 E .3.14.5. Figurefor Additional Model Presented: Exponential (M4) Exponential_beta Model 4 with 0.95 Confidence Level Mean Response 8 17:30 02/16 2010 9 This document is a draftfor review purposes only and does not constitute Agency policy. E-339 DRAFT--DO NOT CITE OR QUOTE 1 E.3.15. H ojo et al., 2002: D R L R esponse Per M inute 2 E .3 .1 5 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) H ill 0 N /A 126.353 1.646E+01 1.800E-13 Notes linear polynom ial, 3degree power power, unrestricted exponential (M 2) 2 0.004 132.825 2.067E+02 9.757E+01 2 0.004 132.825 2.067E+02 9.757E+01 2 0.004 132.825 2 .0 6 7 E + 0 2 9 .7 5 7 E + 0 1 p o w er b oun d hit (p ow er = 1) 2 0.741 122.455 1.800E +04 error unrestricted (power = 0) 2 0.568 122.985 6.184E +00 error exponential (M 3) exponential (M4) b exponential (M 5) 2 0.568 122.985 6.184E +00 error p o w er hit b ound (d = 1) 1 0.479 124.356 4.775E+00 2.704E-01 0 N /A 126.353 1.118E+01 2.127E-01 a C onstant variance m odel selected (p = 0 .3 0 0 4 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3 .1 5 .2 . Outputfor Selected Model: Exponential (M4) 6 H ojo et al., 2002: D R L R esponse Per M inute 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 2 3 _ H o j o _ 2 0 0 2 _ D R L r e s p _ E x p C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 T u e F e b 16 1 7 : 3 1 : 2 4 2 0 1 0 14 15 16 T a b l e 5, v a l u e s a d j u s t e d b y a c o n s t a n t t o a l l o w e x p o n e n t i a l m o d e l 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 30 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 31 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. This document is a draftfor review purposes only and does not constitute Agency policy. E-340 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Dependent variable = Mean 4 Independent variable = Dose 5 Data are assumed to be distributed: normally 6 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 7 r h o is s e t t o 0. 8 A constant v a r i a n c e model is fit. 9 10 T o t a l n u m b e r o f d o s e g r o u p s = 4 11 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 12 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 13 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 16 M L E s o l u t i o n p r o v i d e d : E x a c t 17 18 19 I n i t i a l P a r a m e t e r V a l u e s 20 21 Variable Model 4 22 23 lnalpha 4.51689 24 rho(S) 0 25 a 2 4 . 6 3 6 2 26 b 0 . 0 2 1 2 6 7 9 27 c 0 . 0 1 8 4 7 8 5 28 d 1 29 30 (S) = S p e c i f i e d 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 Variable Model 4 37 38 lnalpha 4.54075 39 rho 0 40 a 2 3 . 4 6 5 41 b 0 . 1 2 8 5 9 42 c 0 . 1 0 0 6 1 5 43 d 1 44 45 46 T a b l e o f S t a t s F r o m I n p u t D a t a 47 48 Dose N Obs Mean Obs Std Dev 49 50 05 23.46 7.986 51 20 5 4.013 10.96 52 60 6 0.478 7.194 53 180 5 4.594 15.23 54 55 56 E s t i m a t e d V a l u e s o f I n t e r e s t 57 58 Dose Est Mean Est Std Scaled Residu 59 60 0 23.47 9. 6 8 3 -0.0004677 61 20 3.973 9. 6 8 3 0.009182 62 60 2.37 9. 6 8 3 -0.4787 63 180 2.361 9. 6 8 3 0.5157 64 65 66 67 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d : 68 69 M o d e l A1: Yij = M u ( i ) + e (ij) 70 V a r { e ( i j ) } = S i g m a ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-341 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A2: Yij Mu(i) + e(ij) 3 Var{e(ij)} Sigma(i)^2 4 5 Model A3: Yij Mu(i) + e(ij) 6 Var{e(ij)} e x p ( l a l p h a + log(mean(i)) * rho) 7 8 M o d e l R: Yij M u + e(i) 9 Var{e(ij)} Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) DF AIC 15 16 A1 -57.92733 5 125.8547 17 A2 -56.09669 8 128.1934 18 A3 -57.92733 5 125.8547 19 R -64.49611 2 132.9922 20 4 -58.17787 4 124.3557 21 22 23 Additive constant for all log-likelihoods = -19.3. This constant added to the 24 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 25 d e p e n d o n t h e m o d e l p a r a m e t e r s . 26 27 28 E x p l a n a t i o n o f T e s t s 29 30 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 31 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 32 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 33 34 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 35 36 37 T e s t s o f I n t e r e s t 38 39 T e s t -2*log(Likelihood Ratio) D. F. p-value 40 41 T e s t 1 16.8 6 0.01005 42 T e s t 2 3.661 3 0.3004 43 T e s t 3 3.661 3 0.3004 44 T e s t 6a 0.5011 1 0.479 45 46 47 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 48 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 49 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 50 51 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 52 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e . 53 54 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 55 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 56 57 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 58 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 59 60 61 B e n c h m a r k D o s e C o m p u t a t i o n s : 62 63 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 64 65 R i s k T y p e = E s t i m a t e d s t a n d a r d d e v i a t i o n s f r o m c o n t r o l 66 67 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 68 69 BMD = 4.77493 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-342 DRAFT--DO NOT CITE OR QUOTE 1 BMDL 0.270447 2 3 4 E .3 .1 5 .3 . Figurefor Selected Model: Exponential (M4) Exponential_beta Model 4 with 0.95 Confidence Level Mean Response 5 17:31 02/16 2010 6 This document is a draftfor review purposes only and does not constitute Agency policy. E-343 DRAFT--DO NOT CITE OR QUOTE 1 E .3.16. K attainen et al., 2001: 3rd M olar E ruption, Fem ale 2 E .3 .1 6 .1 . Summary Table o f BMDS Modeling Results Model logistic Degrees of Freedom x2p Value AIC 3 0.292 89.060 BMD BMDL (ng/kg-d) (ng/kg-d) Notes negative intercept (intercept = 1.941E+02 1.390E+02 -1.508) log-logistic a 3 0.923 85.535 4.763E+01 2.481E+01 slope bound hit (slope = 1) log-probit 3 0 .3 9 0 88.231 1 .574E + 02 9.5 1 2 E + 0 1 slop e bound hit (slop e = 1) probit multistage, 4degree log-logistic, unrestricted b log-probit, unrestricted negative intercept (intercept = 3 0.306 88.919 1.858E+02 1.370E+02 -0.927) 3 0.641 86.798 8.677E+01 5.520E+01 final h = 0 2 0.952 87.157 2.599E+01 1.730E+00 unrestricted (slope = 0.794) 2 0.941 87.179 2.813E+01 2.334E+00 unrestricted (slope = 0.478) aBest-fitting m odel, BM D S output presented in this appendix b Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .1 6 .2 . Outputfor Selected Model: Log-Logistic 6 Kattainen et al., 2001: 3rd M olar Eruption, Fem ale 7 8 9 10 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 2 4 _ K a t t _ 2 0 0 1 _ E r u p _ L o g L o g i s t i c _ B M R 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 2 4 _ K a t t _ 2 0 0 1 _ E r u p _ L o g L o g i s t i c _ B M R 1 . p l t 13 ~ T u e _ F e b 16 1 7 : 3 1 : 5 2 2 0 1 0 14 15 16 F i g u r e 2 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 S l o p e p a r a m e t e r is r e s t r i c t e d as s l o p e > = 1 27 28 T o t a l n u m b e r o f o b s e r v a t i o n s = 5 29 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 30 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 31 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 32 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 33 34 35 36 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 37 This document is a draftfor review purposes only and does not constitute Agency policy. E-344 DRAFT--DO NOT CITE OR QUOTE 1 2 Default Initial Parameter Values 3 background = 0.0625 4 intercept = -6.063 5 slope = 1 6 7 8 Asymptotic Correlation Matrix of Parameter Estimates 9 10 ( * * * T h e m o d e l p a r a m e t e r ( s ) - s l o p e 11 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 12 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 13 14 background intercept 15 16 b a c k g r o u n d 1 -0.56 17 18 i n t e r c e p t -0.56 1 19 20 21 22 P a r a m e t e r E s t i m a t e s 23 24 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 25 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 26 b a c k g r o u n d 0 . 0 8 4 6 7 8 5 27 intercept -6.06063 28 s l o p e 1 29 30 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 31 32 33 34 A n a l y s i s o f D e v i a n c e T a b l e 35 36 Model Log(likelihood) # Param's Deviance Test d.f. P-value 37 Full model -40.5286 5 38 F i t t e d m o d e l -40.7674 2 0.477533 3 0.9238 39 R e d u c e d m o d e l -50.7341 1 20.411 4 0.0004142 40 41 AIC: 85.5347 42 43 44 G o o d n e s s o f F i t 45 S c a l e d 46 Dose Est. Prob. Expected Observed Size Residual 47 48 0.0000 0.0847 1.355 1. 0 0 0 16 -0.319 49 30.0000 0.1445 2.457 3. 0 0 0 17 0.374 50 100.0000 0.2578 3.867 4. 0 0 0 15 0.078 51 300.0000 0.4615 5.538 6. 0 0 0 12 0.267 52 1 0 0 0 . 0 0 0 0 0.7254 13.782 13. 000 19 -0.402 53 54 C h i ^ 2 = 0 . 4 8 d.f. = 3 P-value = 0.9231 55 56 57 B e n c h m a r k D o s e C o m p u t a t i o n 58 59 S p e c i f i e d e f f e c t 0.1 60 61 R i s k T y p e : Extra risk 62 63 C o n f i d e n c e l e v e l 0.95 64 65 BMD 47.6274 66 67 BMDL = 24.8121 68 69 This document is a draftfor review purposes only and does not constitute Agency policy. E-345 DRAFT--DO NOT CITE OR QUOTE 1 E.3.16.3. F igure f o r Selected M odel: Log-Logistic Log-Logistic Model with 0.95 Confidence Level 2 3 4 5 E .3.16.4. Outputfor Additional Model Presented: Log-Logistic, Unrestricted 6 Kattainen et al., 2001: 3rd M olar Eruption, Fem ale 7 8 9 10 11 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a F i l e : C : \ 1 \ 2 4 _ K a t t _ 2 0 0 1 _ E r u p _ L o g L o g i s t i c _ U _ B M R 1 . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 2 4 _ K a t t _ 2 0 0 1 _ E r u p _ L o g L o g i s t i c _ U _ B M R 1 . p l t 14 ~ T u e _ F e b 16 1 7 : 3 1 : 5 3 2 0 1 0 15 16 17 F i g u r e 2 18 19 20 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 21 22 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 23 24 25 D e p e n d e n t v a r i a b l e = D i c h E f f 26 I n d e p e n d e n t v a r i a b l e = D o s e 27 S l o p e p a r a m e t e r is n o t r e s t r i c t e d 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-346 DRAFT--DO NOT CITE OR QUOTE 1 Total number of observations = 5 2 Total number of records with missing values = 0 3 Maximum number of iterations = 250 4 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 5 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 6 7 8 9 User has chosen the log transformed model 10 11 12 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 13 background = 0.0625 14 intercept = -4.71231 15 slope = 0.782659 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 background intercept slope 21 CO 'tr o CO 'tr o 22 ) a c k g r o u n d 1 0.39 23 24 i n t e r c e p t 1 -0.98 25 26 s l o p e 0.39 -0.98 1 27 28 29 30 P a r a m e t e r E s t i m a t e s 31 32 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 33 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 34 b a c k g r o u n d 0 . 0 6 3 3 2 1 7 35 intercept -4.78282 36 slope 0.793723 37 38 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 39 40 41 42 A n a l y s i s o f D e v i a n c e T a b l e 43 44 Model Log(likelihood) # Param's Deviance Test d.f. P-value 45 Full model -40.5286 5 46 F i t t e d m o d e l -40.5783 3 0.0994416 2 0.9515 47 R e d u c e d m o d e l -50.7341 1 20.411 4 0.0004142 48 49 AIC: 87.1566 50 51 52 G o o d n e s s o f F i t 53 S c a l e d 54 Dose Est. Prob. Expected Observed Size Residual 55 56 0.0000 0.0633 1.013 1. 0 0 0 16 -0.013 57 30.0000 0.1670 2.840 3. 0 0 0 17 0.104 58 100.0000 0.2924 4.387 4. 0 0 0 15 -0.219 59 300.0000 0.4721 5.666 6. 0 0 0 12 0.193 60 1 0 0 0 . 0 0 0 0 0.6892 13.095 13. 000 19 -0.047 61 62 C h i ^ 2 = 0 . 1 0 d.f. = 2 P-value = 0.9518 63 64 65 B e n c h m a r k D o s e C o m p u t a t i o n 66 67 S p e c i f i e d e f f e c t 0.1 68 69 R i s k T y p e = Extra risk 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-347 DRAFT--DO NOT CITE OR QUOTE 1 Confidence level 2 3 BMD = 4 5 BMDL = 6 7 0.95 25.986 1.73001 8 E .3.16.5. Figurefor Additional Model Presented: Log-Logistic, Unrestricted Log-Logistic Model with 0.95 Confidence Level Fraction Affected 9 17:31 02/16 2010 10 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-348 DRAFT--DO NOT CITE OR QUOTE 1 E .3.17. K attainen et al., 2001: 3rd M olar L ength, F em ale 2 E .3 .1 7 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of x2p Freedom Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M2) 3 <0.0001 -122.954 4.027E+02 2.366E+02 Notes exponential (M3) 3 < 0 .0 0 0 1 -1 2 2 .9 5 4 4 .0 2 7 E + 0 2 2 .3 6 6 E + 0 2 p o w er hit boun d (d = 1) exponential (M4) 2 <0.0001 -80.747 error error exponential (M5) Hill b 1 <0.0001 -78.747 error error 2 0.013 -151.152 4.052E+00 2.144E+00 n lower bound hit (n = 1) linear polynom ial, 4degree power 3 <0.0001 -122.325 4.659E+02 2.963E+02 3 <0.0001 -122.325 4.659E+02 2.963E+02 3 < 0 .0 0 0 1 -1 2 2 .3 2 5 4 .6 5 9 E + 0 2 2 .9 6 3 E + 0 2 p o w er b ound hit (p ow er = 1) Hill, unrestricted c 1 0.087 -154.939 1.913E-02 1.928E-04 unrestricted (n = 0.197) power, unrestricted 2 0.250 -157.093 9.098E-03 9.097E-03 unrestricted (power = 0.169) a N on-constant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .1 7 .2 . Outputfor Selected Model: Hill 6 Kattainen et al., 2001: 3rd M olar Length, Fem ale 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 2 5 _ K a t t _ 2 0 0 1 _ L e n g t h _ H i l l _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 2 5 _ K a t t _ 2 0 0 1 _ L e n g t h _ H i l l _ 1 . p l t 13 ~ T u e F e b 1 6 ~ 1 7 : 3 2 : 2 1 2 0 1 0 14 15 16 F i g u r e 3 f e m a l e o n l y 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 P o w e r p a r a m e t e r r e s t r i c t e d t o b e g r e a t e r t h a n 1 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + r h o * l n ( m e a n ( i ) ) ) This document is a draftfor review purposes only and does not constitute Agency policy. E-349 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 5 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 lalpha -2.37155 12 rho 0 13 intercept 1.85591 14 - 0 . 5 0 7 8 7 4 15 0 . 8 2 6 2 0 4 16 2 7 . 3 3 0 5 17 18 19 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 20 21 ( * * * T h e m o d e l p a r a m e t e r ( s ) -n 22 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 23 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 24 25 lalpha rho intercept v k 26 27 l a l p h a 1 -0.98 -0.16 0.84 -0.37 28 29 rho -0.98 1 0.2 -0.79 0.39 30 31 i n t e r c e p t -0.16 0.2 1 -0.31 -0.11 32 CO o CO o 33 v 0.84 -0.79 -0.31 1 34 35 k -0.37 0.39 -0.11 1 36 37 38 39 P a r a m e t e r E s t i m a t e s 40 41 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 42 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 43 lalpha 3.34561 1.40443 0.592981 6.09824 44 rho -14.3325 2.62129 -19.4701 -9.19484 45 i n t e r c e p t 1.8548 0.0159017 1.82364 1.88597 46 v -0.441166 0.058852 -0.556513 -0.325818 47 n 1 N A 48 k 24.0343 7.84495 8.65852 39.4101 49 50 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 51 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 52 h a s n o s t a n d a r d e r r o r . 53 54 55 56 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 57 58 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled R 59 60 61 0 16 1.86 1.85 0.0661 0.0637 0.0692 62 30 17 1.58 1.61 0.185 0.176 -0.768 63 100 15 1.6 1.5 0.265 0.293 1.28 64 300 12 1.5 1.45 0.221 0.378 0.527 65 1000 19 1.35 1.42 0.515 0.423 -0.783 66 67 68 69 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-350 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A1: Yij Mu(i) + e(ij) 3 Var{e(ij)} Sigma^2 4 5 Model A2: Yij Mu(i) + e(ij) 6 Var{e(ij)} Sigma(i)^2 7 8 Model A3: Yij = Mu(i) + e(ij) 9 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 10 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 11 w e r e s p e c i f i e d b y t h e u s e r 12 13 M o d e l R: Yi = M u + e(i) 14 V a r { e ( i ) } = S i g m a ^ 2 15 16 17 L i k e l i h o o d s o f I n t e r e s t 18 19 Model Log(likelihood) # Param's AIC 20 A1 56.758717 6 -101.517434 21 A 2 8 5 . 8 5 6 4 5 0 10 - 1 5 1 . 7 1 2 9 0 1 22 A3 84.934314 7 -155.868628 23 fitted 80.575940 5 -151.151880 24 R 45.373551 2 -86.747101 25 26 27 E x p l a n a t i o n o f T e s t s 28 29 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 30 (A2 vs. R) 31 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 32 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 33 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 34 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 35 36 T e s t s o f I n t e r e s t 37 38 Test -2*log(Likelihood Ratio) Test df p-val 39 40 T e s t 1 80.9658 8 <.0001 41 T e s t 2 58.1955 4 <.0001 42 T e s t 3 1.84427 3 0.6053 43 T e s t 4 8.71675 2 0.0128 44 45 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 46 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 47 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 48 49 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 50 m o d e l a p p e a r s t o b e a p p r o p r i a t e 51 52 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 53 t o b e a p p r o p r i a t e h e r e 54 55 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 56 m o d e l 57 58 59 B e n c h m a r k D o s e C o m p u t a t i o n 60 61 S p e c i f i e d e f f e c t = 1 62 63 R i s k T y p e = Estimated standard deviations from the control mean 64 65 C o n f i d e n c e l e v e l = 0.95 66 67 BMD = 4.05231 68 69 BMDL = 2.14357 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-351 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E.3.17.3. F igure f o r Selected M odel: H ill Hill Model with 0.95 Confidence Level 2 3 4 5 E .3.17.4. Outputfor Additional Model Presented: Hill, Unrestricted 6 Kattainen et al., 2001: 3rd M olar Length, Fem ale 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 2 5 _ K a t t _ 2 0 0 1 _ L e n g t h _ H i l l _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 2 5 _ K a t t _ 2 0 0 1 _ L e n g t h _ H i l l _ U _ 1 . p l t 13 ~ T u e F e b 1 6 ~ 1 7 : 3 2 : 2 1 2 0 1 0 14 15 16 F i g u r e 3 f e m a l e o n l y 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 P o w e r p a r a m e t e r is n o t r e s t r i c t e d 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + r h o * l n ( m e a n ( i ) ) ) 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-352 DRAFT--DO NOT CITE OR QUOTE 1 Total number of dose groups = 5 2 Total number of records with missing values = 0 3 Maximum number of iterations = 250 4 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 5 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 6 7 8 9 Default Initial Parameter Values 10 lalpha -2.37155 11 rho 0 12 intercept 1.85591 13 - 0 . 5 0 7 8 7 4 14 0 . 8 2 6 2 0 4 15 2 7 . 3 3 0 5 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 lalpha rho intercept v n k 21 22 l a l p h a 1 -0.98 -0.18 0.18 -0.28 -0.011 23 24 rho -0.98 1 0.22 -0.18 0.29 0.011 25 26 i n t e r c e p t -0.18 0.22 1 -0.025 -0.059 0.0019 27 28 v 0.18 -0.18 -0.025 1 0.51 -0.96 29 30 n -0.28 0.29 -0.059 0.51 1 -0.71 31 32 k -0.011 0.011 0.0019 -0.96 -0.71 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 39 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 40 lalpha 3.21882 1.4221 0.431563 6.00607 41 rho -14.0862 2.68292 -19.3446 -8.82777 42 ntercept 1.85564 0.0160224 1.82424 1.88704 43 v -2.48572 2.89658 -8.16291 3.19148 44 n 0.196925 0.0499318 0.0990606 0.29479 45 k 1.92967e+006 1.60869e+007 -2.96e+007 3.34593e+007 46 47 48 49 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 50 51 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 52 53 54 0 16 1.86 1.86 0.0661 0.0643 0.0164 55 30 17 1.58 1.6 0.185 0.18 -0.598 56 100 15 1.6 1.54 0.265 0.234 0.857 57 300 12 1.5 1.48 0.221 0.316 0.259 58 1000 19 1.35 1.4 0.515 0.471 -0.466 59 60 61 62 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 63 64 65 M o d e l A1: Yij Mu(i) + e(ij) 66 Var{e(ij)} Sigma^2 67 68 M o d e l A2: Yij Mu(i) + e(ij) 69 Var{e(ij)} Sigma(i)^2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-353 DRAFT--DO NOT CITE OR QUOTE 1 Model A3: Yij = Mu(i) + e(ij) 2 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 3 Model A3 uses any fixed variance parameters that 4 were specified by the user 5 6 M o d e l R: Yi = M u + e(i) 7 Var{e(i)} = Sigma^2 8 9 10 L i k e l i h o o d s o f I n t e r e s t 11 12 Model Log(likelihood) # Param's AIC 13 A1 56.758717 6 -101.517434 14 A 2 8 5 . 8 5 6 4 5 0 10 - 1 5 1 . 7 1 2 9 0 1 15 A3 84.934314 7 -155.868628 16 fitted 83.469680 6 -154.939361 17 R 45.373551 2 -86.747101 18 19 20 E x p l a n a t i o n o f T e s t s 21 22 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 23 (A2 vs. R) 24 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 25 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 26 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 27 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 28 29 T e s t s o f I n t e r e s t 30 31 Test -2*log(Likelihood Ratio) Test df p-value 32 33 T e s t 1 80.9658 8 <.0001 34 T e s t 2 58.1955 4 <.0001 35 T e s t 3 1.84427 3 0.6053 36 T e s t 4 2.92927 1 0.08699 37 38 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 39 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 40 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 41 42 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 43 m o d e l a p p e a r s t o b e a p p r o p r i a t e 44 45 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 46 t o b e a p p r o p r i a t e h e r e 47 48 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 49 m o d e l 50 51 52 B e n c h m a r k D o s e C o m p u t a t i o n 53 54 S p e c i f i e d e f f e c t = 1 55 56 R i s k T y p e = Estimated standard deviations from the control mean 57 58 C o n f i d e n c e l e v e l = 0.95 59 60 BMD = 0.0191282 61 62 BMDL = 0.0001928 63 64 This document is a draftfor review purposes only and does not constitute Agency policy. E-354 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E.3.17.5. F igure f o r A d d itio n a l M o d el Presented: H ill, U nrestricted Hill Model with 0.95 Confidence Level 2 17:32 02/16 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-355 DRAFT--DO NOT CITE OR QUOTE 1 E.3.18. Keller et al., 2007: Missing Mandibular Molars, CBA J 2 E.3.18.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2pValue AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes gamma 1 0.105 52.490 7.293E+01 2.027E+01 logistic negative intercept (intercept = 2 0.320 50.095 7.168E+01 5.142E+01 -3.372) log-logistic 1 0.105 52.524 9.278E+01 5.273E+01 log-probit multistage, 1degree a multistage, 2degree multistage, 3degree probit W eibull 1 0.105 52.524 8.849E+01 5.297E+01 3 0.276 49.409 2.778E+01 1.884E+01 1 0.126 51.515 4.619E+01 2.214E+01 1 0.141 51.222 4.253E+01 2.212E+01 negative intercept (intercept = 2 0.325 50.032 6.848E+01 4.775E+01 -1.851) 1 0.108 52.216 6.079E+01 2.078E+01 aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3.18.2. Outputfor Selected Model: Multistage, 1-Degree 6 K eller et al., 2007: M issing M andibular M olars, C B A J 7 8 9 10 M u l t i s t a g e M o d e l . ( V e r s i o n : 3.0; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 2 6 _ K e l l e r _ 2 0 0 7 _ M o l a r s _ M u l t i 1 _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 2 6 _ K e l l e r _ 2 0 0 7 _ M o l a r s _ M u l t i 1 _ 1 . p l t 13 T u e F e b _ 16 1 7 : 3 2 : 5 6 2 0 1 0 14 15 16 T a b l e 1 u s i n g m a n d i b u l a r m o l a r s o n l y 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) * [ 1 - E X P ( 22 - b e t a i * d o s e ^ i ) ] 23 24 T h e p a r a m e t e r b e t a s a r e r e s t r i c t e d t o b e p o s i t i v e 25 26 27 D e p e n d e n t v a r i a b l e = D i c h E f f 28 I n d e p e n d e n t v a r i a b l e = D o s e 29 30 T o t a l n u m b e r o f o b s e r v a t i o n s = 4 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 T o t a l n u m b e r o f p a r a m e t e r s i n m o d e l = 2 33 T o t a l n u m b e r o f s p e c i f i e d p a r a m e t e r s = 0 This document is a draftfor review purposes only and does not constitute Agency policy. E-356 DRAFT--DO NOT CITE OR QUOTE 1 Degree of polynomial = 1 2 3 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 Background = 0 12 B e t a ( 1 ) = 1 . 0 2 9 0 9 e + 0 1 7 13 14 15 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 16 17 ( * * * T h e m o d e l p a r a m e t e r ( s ) - B a c k g r o u n d 18 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 19 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 20 21 B e t a ( 1 ) 22 23 B e t a ( 1 ) 1 24 25 26 27 P a r a m e t e r E s t i m a t e s 28 29 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 30 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf 31 B a c k g r o u n d 0* * * 32 Beta(1) 0.00379264 * * * 33 34 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 35 36 37 38 A n a l y s i s o f D e v i a n c e T a b l e 39 40 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 41 Full model -21.5798 4 42 F i t t e d m o d e l -23.7044 1 4.24924 3 0.2358 43 R e d u c e d m o d e l -71.326 1 99.4926 3 <.0001 44 45 AIC: 49.4088 46 47 48 G o o d n e s s o f F i t 49 S c a l e d 50 Dose Est. Prob. Expected Observed Size Residu. 51 52 0.0000 0.0000 0.000 0.000 29 0.000 53 10.0000 0.0372 0.856 2.000 23 1.260 54 100.0000 0.3156 9.153 6.000 29 -1.260 55 1 0 0 0 . 0 0 0 0 0.9775 29.324 30.000 30 0.832 56 57 C h i ^ 2 = 3 . 8 7 d.f. = 3 P-value = 0.2762 58 59 60 B e n c h m a r k D o s e C o m p u t a t i o n 61 62 S p e c i f i e d e f f e c t = 0.1 63 64 R i s k T y p e = Extra risk 65 66 C o n f i d e n c e l e v e l = 0.95 67 68 BMD = 27.7803 69 70 BMDL = 18.8447 This document is a draftfor review purposes only and does not constitute Agency policy. E-357 DRAFT--DO NOT CITE OR QUOTE 1 2 BMDU 41.7256 3 4 T a k e n t ogether, (18.8447, 41.7256) is a 90 5 interval for the BMD 6 7 % two-sided confidence 8 E .3.18.3. Figurefor Selected Model: Multistage, 1-Degree Multistage Model with 0.95 Confidence Level Fraction Affected 9 17:32 02/16 2010 10 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-358 DRAFT--DO NOT CITE OR QUOTE 1 E.3.19. Kociba et al., 1978: Urinary Coproporphyrin, Females 2 E.3.19.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M 2) 2 <0.0001 84.006 7.054E+01 4.341E+01 Notes exponential (M 3) exponential (M4) b exponential (M 5) 2 < 0 .0 0 0 1 8 4 .0 0 6 7 .0 5 4 E + 0 1 4 .3 4 1 E + 0 1 p ow er hit b ound (d = 1) 1 0.040 70.556 1.625E+00 7.300E-01 0 N /A 69.092 3.128E+00 1.024E+00 H ill 0 N /A 69.047 6.677E +00 error linear 2 <0.0001 83.713 6.195E+01 3.112E+01 polynom ial, 3degree 2 <0.0001 83.713 6.195E+01 3.112E+01 power 2 < 0 .0 0 0 1 83.713 6 .1 9 5 E + 0 1 3.1 1 2 E + 0 1 p ow er b oun d hit (p ow er = 1) power, unrestricted 1 0.001 78.260 7.808E-01 1.693E-08 unrestricted (power = 0.306) a N on-constant variance m odel selected (p = 0 .0298) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3 .1 9 .2 . Outputfor Selected Model: Exponential (M4) 6 K ociba et al., 1978: Urinary Coproporphyrin, Fem ales 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a Fil e : C : \ 1 \ 2 9 K o c i b a 1 9 7 8 C o p r o E x p 1.( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 T u e F e b 16 1 7 : 3 4 : 14 15 16 T a b l e 2 - U r i n a r y C o p r o p o r p h y r i n 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 30 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 31 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. This document is a draftfor review purposes only and does not constitute Agency policy. E-359 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Dependent variable = Mean 4 Independent variable = Dose 5 Data are assumed to be distributed: normally 6 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 7 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) * rho) 8 9 Total number of dose groups = 4 10 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 11 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 12 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 13 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 15 M L E s o l u t i o n p r o v i d e d : E x a c t 16 17 18 I n i t i a l P a r a m e t e r V a l u e s 19 20 Variable Model 4 21 22 lnalpha -5.58269 23 rho 2.98472 24 a 8 .17 25 b 0 . 0 2 5 9 4 6 9 26 c 2 . 2 3 6 2 3 27 d 1 28 29 30 31 P a r a m e t e r E s t i m a t e s 32 33 Variable Model 4 34 35 lnalpha -4.94473 36 rho 2.76088 37 a 8 . 9 3 0 3 9 38 b 0 . 1 3 6 5 5 4 39 c 1 . 9 7 5 3 40 d 1 41 42 43 T a b l e o f S t a t s F r o m I n p u t D a t a 44 45 Dose N Obs Mean Obs Std Dev 46 CO CO Gj CO 47 0 5 1.3 48 1 5 2 49 10 5 16.4 4.7 50 100 5 17.4 4 51 52 53 E s t i m a t e d V a l u e s o f I n t e r e s t 54 55 Dose Est Mean Est Std Scaled Residual 56 57 0 8.93 1.733 1.122 58 1 10.04 2.038 -1.582 59 10 15.42 3.683 0.5967 60 100 17.64 4.436 -0.1211 61 62 63 64 O t h e r m o d e l s; f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d 65 66 M o d e l A1: Yij = M u ( i ) + e( ij) 67 V a r { e ( i j ) } = S i g m a ^ 2 68 69 M o d e l A2: Yij = M u ( i ) + e( ij) 70 V a r { e ( i j ) } = S i g m a ( i ) ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-360 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A3: Yij = Mu(i) + e(ij) 3 V ar{e(ij)} = exp(l a l p h a + log(mean(i)) * rho) 4 5 M o d e l R: Yij = M u + e(i) 6 Var{e(ij)} = Sigma^2 7 8 9 Likelihoods of Interest 10 11 Model Log(likelihood) DF AIC 12 13 A1 -31.69739 5 73.39478 14 A2 -27.21541 8 70.43081 15 A3 -28.16434 6 68.32868 16 R -41.73188 2 87.46376 17 4 -30.27804 5 70.55608 18 19 20 Additive constant for all log-likelihoods = -18.38. This constant added to the 21 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 22 d e p e n d o n t h e m o d e l p a r a m e t e r s . 23 24 25 E x p l a n a t i o n o f T e s t s 26 27 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 28 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 29 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 30 31 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 32 33 34 T e s t s o f I n t e r e s t 35 36 T e s t -2*log(Likelihood Ratio) D. F. p-value 37 38 T e s t 1 29.03 6 < 0.0001 39 T e s t 2 8.964 3 0.02977 40 T e s t 3 1.898 2 0.3872 41 T e s t 6a 4.227 1 0.03978 42 43 44 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 45 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 46 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 47 48 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 49 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e . 50 51 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 52 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 53 54 T h e p - v a l u e f o r T e s t 6a is l e s s t h a n .1. M o d e l 4 m a y n o t a d e q u a t e l y 55 d e s c r i b e t h e d a t a ; y o u m a y w a n t t o c o n s i d e r a n o t h e r m o d e l . 56 57 58 B e n c h m a r k D o s e C o m p u t a t i o n s : 59 60 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 61 62 R i s k T y p e = E s t i m a t e d s t a n d a r d d e v i a t i o n s f r o m c o n t r o l 63 64 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 65 66 BMD = 1.62505 67 68 BMDL = 0.729987 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-361 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E .3 .1 9 .3 . Figurefor Selected Model: Exponential (M4) Exponential_beta Model 4 with 0.95 Confidence Level 2 17:34 02/16 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-362 DRAFT--DO NOT CITE OR QUOTE 1 E.3.20. Kociba et al., 1978: Uroporphyrin per Creatinine, Female 2 E.3.20.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2pValue AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M 2) 2 0.661 -93.561 4.357E+01 3.328E+01 Notes exponential (M 3) 2 0.661 -9 3 .5 6 1 4.3 5 7 E + 0 1 3.3 2 8 E + 0 1 p o w er hit b ound (d = 1) exponential (M 4) 1 0.576 -92.078 1.719E+01 5.516E +00 exponential (M 5) 0 N /A -90.190 1.080E+01 5.613E +00 H ill 0 N /A -90.190 1.099E+01 5.088E +00 linear b 2 0.720 -93.735 3.522E+01 2.500E+01 polynom ial, 3degree 2 0.720 -93.735 3.522E+01 2.500E+01 power 2 0 .7 2 0 -9 3 .7 3 5 3 .522E + 01 2 .5 0 0 E + 0 1 p o w er b oun d hit (p ow er = 1) power, unrestricted 1 0.515 -91.967 2.274E+01 3.334E +00 unrestricted (power = 0.731) a C onstant variance m odel selected (p = 0 .4 9 1 9 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3 .2 0 .2 . Outputfor Selected Model: Linear 6 K ociba et al., 1978: Uroporphyrin per Creatinine, Fem ale 7 8 9 10 P o l y n o m i a l M o d e l . ( V e r s i o n : 2 . 1 3 ; D a t e : 0 4 / 0 8 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 2 8 _ K o c i b a _ 1 9 7 8 _ U r o p o r _ L i n e a r C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 2 8 _ K o c i b a _ 1 9 7 8 _ U r o p o r _ L i n e a r C V _ 1 . p l t 13 T u e F e b _ 16 1 7 : 3 4 7 1 2 2 0 1 0 14 15 16 T a b l e 2 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 S i g n s o f t h e p o l y n o m i a l c o e f f i c i e n t s a r e n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t 29 30 T o t a l n u m b e r o f d o s e g r o u p s = 4 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 This document is a draftfor review purposes only and does not constitute Agency policy. E-363 DRAFT--DO NOT CITE OR QUOTE 1 Maximum number of 250 2 Relative Function has b e e n set to: 1e-008 3 Parameter has b een set to: 1e-008 4 5 6 7 Default Initial Parameter Values 8 alpha = 0.0030385 9 rho = 0 Specified 10 beta_0 = 0.154759 11 beta 1 = 0.0014231 12 13 14 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 15 16 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 17 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 18 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 19 20 alpha beta_0 beta_1 21 22 a l p h a 1 -2.2e-009 3.5e-009 23 24 beta_0 -2.2e-009 1 -0.55 25 26 beta 1 3.5e-009 -0.55 1 27 28 29 30 P a r a m e t e r E s t i m a t e s 31 32 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 33 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 34 alpha 0.00251184 0.000794315 0.000955015 0.00406867 35 beta_0 0.154759 0.0134422 0.128413 0.181105 36 beta 1 0.0014231 0.000267497 0.000898818 0.00194739 37 38 39 40 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 41 42 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 43 44 45 05 0.157 0.155 0.05 0.0501 0.1 46 15 0.143 0.156 0.037 0.0501 -0.588 47 10 5 0.181 0.169 0.053 0.0501 0.536 48 100 5 0.296 0.297 0.074 0.0501 -0.0477 49 50 51 52 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 53 54 55 M o d e l A1: Yij = Mu(i) + e(ij) 56 V a r { e i ;i j ) } = S i g m a ^ 2 57 58 M o d e l A2: Yij = Mu(i) + e(ij) 59 V a r { e i :i j ) } = S i g m a ( i ) ^2 60 61 M o d e l A3: Yij = Mu(i) + e(ij) 62 V a r { e i i j ) } = S i g m a ^ 2 63 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 64 w e r e s p e c i f i e d b y t h e u s e r 65 66 M o d e l R: Yi = M u + e(i) 67 V a r { e ( i ) } = S i g m a ^ 2 68 69 70 L i k e l i h o o d s o f I n t e r e s t This document is a draftfor review purposes only and does not constitute Agency policy. E-364 DRAFT--DO NOT CITE OR QUOTE 1 2 Model Log(likelihood) # Param's AIC 3 A1 50.195349 5 -90.390697 4 A2 51.400051 8 -86.800103 5 A3 50.195349 5 -90.390697 6 fitted 49.867385 3 -93.734769 7 R 41.049755 2 -78.099510 8 9 10 E x p l a n a t i o n o f T e s t s 11 12 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 13 (A2 vs. R) 14 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 15 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 16 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 17 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 18 19 T e s t s o f I n t e r e s t 20 21 Test -2*log(Likelihood Ratio) Test df p-value 22 23 T e s t 1 20.7006 6 0.002076 24 T e s t 2 2.40941 3 0.4919 25 T e s t 3 2.40941 3 0.4919 26 T e s t 4 0.655928 2 0.7204 27 28 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 29 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 30 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 31 32 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 33 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 34 35 36 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 37 t o b e a p p r o p r i a t e h e r e 38 39 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 40 t o a d e q u a t e l y d e s c r i b e t h e d a t a 41 42 43 B e n c h m a r k D o s e C o m p u t a t i o n 44 45 S p e c i f i e d e f f e c t = 1 46 47 R i s k T y p e = Estimated standard deviations from the control mean 48 49 C o n f i d e n c e l e v e l = 0.95 50 51 BMD = 35.2176 52 53 54 BMDL = 25.0024 55 56 This document is a draftfor review purposes only and does not constitute Agency policy. E-365 DRAFT--DO NOT CITE OR QUOTE 1 E.3.20.3. F igure f o r Selected M odel: L in ea r Linear Model with 0.95 Confidence Level Mean Response 2 17:34 02/16 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-366 DRAFT--DO NOT CITE OR QUOTE 1 E.3.21. Latchoumycandane and Mathur, 2002: Sperm Production 2 E.3.21.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of X2pFreedom Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M 2) 2 <0.0001 95.106 7.640E+01 3.992E+01 Notes exponential (M 3) 2 < 0 .0 0 0 1 9 5 .1 0 6 7 .6 4 0 E + 0 1 3.9 9 2 E + 0 1 p o w er hit bound (d = 1) exponential (M 4) 1 0.699 75.263 2.435E-01 1.016E-01 exponential (M 5) 0 N /A 77.263 3.697E-01 1.016E-01 Hill b 1 0 .8 5 9 7 5 .1 4 4 1.450E -01 1 .559E -02 n lo w er b oun d hit (n = 1) linear polynom ial, 3degree power 2 <0.0001 95.308 8.275E+01 4.852E+01 2 <0.0001 95.308 8.275E+01 4.852E+01 2 < 0 .0 0 0 1 9 5 .3 0 8 8.2 7 5 E + 0 1 4 .8 5 2 E + 0 1 p o w er b oun d hit (p ow er = 1) Hill, unrestricted c 0 N /A 77.113 6.943E -02 2.060E -06 unrestricted (n = 0.709) power, unrestricted 1 0.499 75.570 2.706E -07 2.706E -07 unrestricted (power = 0.067) a Constant variance m odel selected (p = 0 .8506) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .2 1 .2 . Outputfor Selected Model: Hill 6 Latchoum ycandane and Mathur, 2002: Sperm Production 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 3 0 _ L a t c h _ 2 0 0 2 _ S p e r m _ H i l l C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 3 0 _ L a t c h _ 2 0 0 2 _ S p e r m _ H i l l C V _ 1 . p l t 13 _ T u e F e b 16 1 8 : 1 3 : 2 0 2 0 1 0 14 15 16 ( x 1 0 A 6) T a b l e 1 w i t h o u t V i t a m i n E 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 P o w e r p a r a m e t e r r e s t r i c t e d t o b e g r e a t e r t h a n 1 This document is a draftfor review purposes only and does not constitute Agency policy. E-367 DRAFT--DO NOT CITE OR QUOTE 1 A constant variance model is fit 2 3 Total number of dose groups = 4 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 12 alpha 7.23328 13 rho 0 Specified 14 intercept 22.19 15 - 9 . 0 9 16 1 . 8 0 4 8 4 17 0 . 6 9 7 0 8 6 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( *** The model parameter(s) -rho -n 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 24 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 25 26 alpha intercept v k 27 28 a l p h a 1 6.3e-010 3e-008 8.3e-009 29 CO r o CO r o 30 i n t e r c e p t 6.3e-010 1 -0.23 31 32 v 3 e - 0 0 8 1 -0.17 33 34 k 8.3e-009 -0.23 -0.17 1 35 36 37 38 P a r a m e t e r E s t i m a t e s 39 40 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 41 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Lim 42 alpha 6.03567 1.74235 2.62073 9.45061 43 i n t e r c e p t 22.1885 1.00316 20.2223 24.1547 44 v -9.00869 1.26801 -11.4939 -6.52343 45 n 1 N A 46 k 0.386669 0.265663 -0.134021 0.907359 47 48 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 49 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 50 h a s n o s t a n d a r d e r r o r . 51 52 53 54 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 55 56 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 57 58 59 06 22.2 22.2 2.67 2.46 0.00151 60 16 15.7 15.7 2.65 2.46 -0.0218 61 10 6 1 3 . 7 13.5 2.19 2.46 0.134 62 100 6 13.1 13.2 3.16 2.46 -0.114 63 64 65 66 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 67 68 69 M o d e l A1: Yij = Mu(i) + e(ij) 70 V a r { e ( i j ) } = S i g m a ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-368 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A2: Yij = Mu(i) + e(ij) 3 Var{e(ij)} = Sigma(i)^2 4 5 Model A3: Yij = Mu(i) + e(ij) 6 Var{e(ij)} = Sigma^2 7 Model A3 uses any fixed variance parameters that 8 were specified by the user 9 10 M o d e l R: Yi = M u + e(i) 11 V a r { e ( i ) } = S i g m a ^ 2 12 13 14 L i k e l i h o o d s o f I n t e r e s t 15 16 Model Log(likelihood) # Param's AIC 17 A1 -33.556444 5 77 .112888 18 A2 -33.158811 8 82 . 317623 19 A3 -33.556444 5 77 .112888 20 fitted -33.572245 4 75 .144490 21 R -47.392394 2 98 . 7 8 4 7 8 8 22 23 24 E x p l a n a t i o n o f T e s t s 25 26 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 27 (A2 vs. R) 28 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 29 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 30 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 31 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 32 33 T e s t s o f I n t e r e s t 34 35 Test -2*log(Likelihood Ratio) Test df p-val 36 37 T e s t 1 28.4672 6 <.0001 38 T e s t 2 0.795266 3 0.8506 39 T e s t 3 0.795266 3 0.8506 40 T e s t 4 0.031602 1 0.8589 41 42 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 43 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 44 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 45 46 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 47 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 48 49 50 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 51 t o b e a p p r o p r i a t e h e r e 52 53 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 54 t o a d e q u a t e l y d e s c r i b e t h e d a t a 55 56 57 B e n c h m a r k D o s e C o m p u t a t i o n 58 59 S p e c i f i e d e f f e c t = 1 60 61 R i s k T y p e = Estimated standard deviations from the control mean 62 63 C o n f i d e n c e l e v e l = 0.95 64 65 BMD = 0.144988 66 67 BMDL = 0.0155926 68 This document is a draftfor review purposes only and does not constitute Agency policy. E-369 DRAFT--DO NOT CITE OR QUOTE 1 E .3 .2 1 .3 . Figurefor Selected Model: Hill Hill Model with 0.95 Confidence Level Mean Response 2 18:13 02/16 2010 3 dose 4 5 E .3.21.4. Outputfor Additional Model Presented: Hill, Unrestricted 6 Latchoum ycandane and Mathur, 2002: Sperm Production 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 3 0 _ L a t c h _ 2 0 0 2 _ S p e r m _ H i l l C V _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 3 0 _ L a t c h _ 2 0 0 2 _ S p e r m _ H i l l C V _ U _ 1 . p l t 13 _ T u e F e b 16 1 8 : 1 3 : 2 1 2 0 1 0 14 15 16 ( x 1 0 A 6) T a b l e 1 w i t h o u t V i t a m i n E 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 P o w e r p a r a m e t e r is n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t This document is a draftfor review purposes only and does not constitute Agency policy. E-370 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 4 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha 7.23328 12 rho 0 Specified 13 intercept 22.19 14 - 9 . 0 9 15 1 . 8 0 4 8 4 16 0 . 6 9 7 0 8 6 17 18 19 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 20 21 ( * * * T h e m o d e l p a r a m e t e r ( s ) -rho 22 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r 23 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 24 25 alpha intercept v n k 26 27 a l p h a 1 -7.6e-009 8e-008 5e-008 1.9e-008 28 29 intercept -7.6e-009 1 -0.5 -0.015 -0.13 30 31 v 8e-008 -0.5 1 0.75 0.55 32 33 n 5e-008 -0.015 0.75 1 0.86 34 35 k 1.9e-008 -0.13 0.55 0.86 1 36 37 38 39 P a r a m e t e r E s t i m a t e s 40 41 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 42 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 43 alpha 6.02773 1.74006 2.61728 9.43818 44 i n t e r c e p t 22.19 1.00231 20.2255 24.1545 45 v -9.23433 2.02073 -13.1949 -5.27378 46 n 0.709305 1.28329 -1.8059 3.22451 47 k 0.290697 0.548737 -0.784807 1.3662 48 49 50 51 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 52 53 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 54 55 56 06 22.2 22.2 2.67 2.46 2.62e-008 57 16 15.7 15.7 2.65 2.46 -1.5e-008 58 10 6 13.7 13.7 2.19 2.46 -4.56e-008 59 100 6 13.1 13.1 3.16 2.46 -3.52e-007 60 61 D e g r e e s o f f r e e d o m f o r T e s t A 3 v s f i t t e d < = 0 62 63 64 65 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 66 67 68 M o d e l A1: Yij = Mu(i) + e(ij) 69 V a r { e ( i j ) } = S i g m a ^ 2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-371 DRAFT--DO NOT CITE OR QUOTE 1 Model A2: Yij = Mu(i) + e(ij) 2 Var{e(ij)} = Sigma(i)^2 3 4 Model A3: Yij = Mu(i) + e(ij) 5 Var{e(ij)} = Sigma^2 6 Model A3 uses any fixed variance parameters that 7 were specified by the user 8 9 M o d e l R: Yi = M u + e(i) 10 V a r { e ( i ) } = S i g m a ^ 2 11 12 13 L i k e l i h o o d s o f I n t e r e s t 14 15 Model Log(likelihood) # Param's AIC 16 A1 -33.556444 5 77 .112888 17 A2 -33.158811 8 82 . 317623 18 A3 -33.556444 5 77 .112888 19 fitted -33.556444 5 77 .112888 20 R -47.392394 2 98 . 7 8 4 7 8 8 21 22 23 E x p l a n a t i o n o f T e s t s 24 25 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 26 (A2 vs. R) 27 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 28 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 29 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 30 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 31 32 T e s t s o f I n t e r e s t 33 34 Test -2*log(Likelihood Ratio) Test df p-vali 35 36 T e s t 1 28.4672 6 <.0001 37 T e s t 2 0.795266 3 0.8506 38 T e s t 3 0.795266 3 0.8506 39 T e s t 4 2.84217e-014 0 NA 40 41 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 42 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 43 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 44 45 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 46 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 47 48 49 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 50 t o b e a p p r o p r i a t e h e r e 51 52 N A - D e g r e e s o f f r e e d o m f o r T e s t 4 a r e l e s s t h a n o r e q u a l t o 0. T h e C h i - S q u a r e 53 t e s t f o r f i t is n o t v a l i d 54 55 56 B e n c h m a r k D o s e C o m p u t a t i o n 57 58 S p e c i f i e d e f f e c t = 1 59 60 R i s k T y p e = Estimated standard deviations from the control mean 61 62 C o n f i d e n c e l e v e l = 0.95 63 64 BMD = 0.0694325 65 66 B M D L = 2 . 0 6 0 0 7 e - 0 0 6 67 This document is a draftfor review purposes only and does not constitute Agency policy. E-372 DRAFT--DO NOT CITE OR QUOTE 1 E.3.21.5. F igure f o r A d d itio n a l M o d el Presented: H ill, U nrestricted Hill Model with 0.95 Confidence Level Mean Response 2 18:13 02/16 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-373 DRAFT--DO NOT CITE OR QUOTE 1 E.3.22. Li et al., 1997: FSH 2 E.3.22.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M 2) 8 <0.0001 1095.240 1.340E+04 1.060E+04 Notes exponential (M 3) 8 < 0 .0 0 0 1 1 0 9 5.240 1 .340E + 04 1 .060E + 04 p o w er hit boun d (d = 1) exponential (M 4) 7 <0.0001 1061.243 1.031E+03 4.015E+02 exponential (M 5) 7 < 0 .0 0 0 1 1061.243 1.031E +03 4 .0 1 5 E + 0 2 p o w er hit boun d (d = 1) H ill 7 <0.0001 1059.547 6.645E +02 error n lo w er b ound hit (n = 1) linear polynom ial, 8degree power b 8 <0.0001 1078.221 5.287E+03 3.602E+03 9 <0.0001 1155.670 error error 8 <0.0001 1078.221 5.287E+03 3.602E+03 power bound hit (power = 1) H ill, unrestricted 6 0.001 1039.902 2.809E+00 6.602E-01 unrestricted (n = 0.291) power, unrestricted c 7 0.002 1037.821 2.508E +00 2.525E-01 unrestricted (power = 0.279) a N on-con stant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .2 2 .2 . Outputfor Selected Model: Power 6 Li et al., 1997: FSH 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 7 2 _ L i _ 1 9 9 7 _ F S H _ P w r _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 7 2 _ L i _ 1 9 9 7 _ F S H _ P w r _ 1 . p l t 13 T u e F e b 16 2 0 : 0 7 : 3 1 2 0 1 0 14 15 16 F i g u r e 3: F S H i n f e m a l e S - D r a t s 2 4 h r a f t e r d o s i n g , 22 d a y o l d r a t s 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 T h e p o w e r is r e s t r i c t e d t o b e g r e a t e r t h a n o r e q u a l t o 1 This document is a draftfor review purposes only and does not constitute Agency policy. E-374 DRAFT--DO NOT CITE OR QUOTE 1 The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i)) rho) 2 3 Total number of dose groups = 10 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 12 lalpha = 9.8191 13 rho = 0 14 control = 22.1591 15 slope = 26.1213 16 power = 0.264963 17 18 19 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 20 21 ( * * * T h e m o d e l p a r a m e t e r ( s ) - p o w e r 22 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 23 a n d d o n o t a p p e a r in t h e c o r r e l a t i o n m a t r i x ) 24 25 lalpha rho control slope 26 27 l a l p h a 1 -0.99 -0.29 -0.023 28 29 rho -0.99 1 0.2 0.023 30 Lf) CO o Lf) CO o 31 c o n t r o l -0.29 0.2 1 32 33 slope -0.023 0.023 1 34 35 36 37 P a r a m e t e r E s t i m a t e s 38 39 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 40 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 41 lalpha 3.5473 1.23656 1.12369 5.9709 42 rho 1.26137 0.244246 0.782659 1.74009 43 control 88.9479 12.9114 63.6419 114.254 44 slope 0.0188972 0.00351723 0.0120035 0.0257908 45 p o w e r 1 NA 46 47 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 48 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 49 h a s n o s t a n d a r d e r r o r . 50 51 52 53 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 54 55 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 56 57 58 0 10 23.9 88.9 29.6 99.9 -2.06 59 3 10 22.2 89 48.5 99.9 -2.12 60 10 10 85.2 89.1 94.3 100 -0.124 61 30 10 73.3 89.5 48.5 100 -0.511 62 100 10 126 90.8 159 101 1.1 63 300 10 132 94.6 116 104 1.14 64 1000 10 117 108 51.2 113 0.25 65 3000 10 304 146 154 136 3.68 66 1 e + 0 0 4 10 347 278 151 205 1.06 67 3 e + 0 0 4 10 455 656 286 352 -1.8 68 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-375 DRAFT--DO NOT CITE OR QUOTE 1 Model Descriptions for likelihoods calculated 2 3 4 Model A1: Yij Mu(i) + e(ij) 5 Var{e(ij)} Sigma^2 6 7 Model A2: Yij Mu(i) + e(ij) 8 Var{e(ij)} Sigma(i)^2 9 10 M o d e l A3: Yij = Mu(i) + e(ij) 11 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 12 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 13 w e r e s p e c i f i e d b y t h e u s e r 14 15 M o d e l R: Yi = M u + e(i) 16 V a r { e ( i ) } = S i g m a ^ 2 17 18 19 L i k e l i h o o d s o f I n t e r e s t 20 21 Model Log(likelihood) # Param's AIC 22 A1 -535.687163 11 1 0 9 3 . 3 7 4 3 2 7 23 A2 -496.367061 20 1032.734122 24 A3 -502.709623 12 1 0 2 9 . 4 1 9 2 4 6 25 fitted -535.110448 4 1078.220896 26 R -574.835246 2 1153.670492 27 28 29 E x p l a n a t i o n o f T e s t s 30 31 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 32 (A2 vs. R) 33 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 34 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 35 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 36 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 37 38 T e s t s o f I n t e r e s t 39 40 Test -2*log(Likelihood Ratio) Test df p-val 41 42 T e s t 1 156.936 18 <.0001 43 T e s t 2 78.6402 9 <.0001 44 T e s t 3 12.6851 8 0.1232 45 T e s t 4 64.8016 8 <.0001 46 47 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 48 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 49 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 50 51 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 52 m o d e l a p p e a r s t o b e a p p r o p r i a t e 53 54 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 55 t o b e a p p r o p r i a t e h e r e 56 57 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 58 m o d e l 59 60 61 B e n c h m a r k D o s e C o m p u t a t i o n 62 63 S p e c i f i e d e f f e c t = 1 64 65 R i s k T y p e = Estimated standard deviations from the control mean 66 67 C o n f i d e n c e l e v e l = 0.95 68 69 B M D = 5 2 8 6 . 6 7 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-376 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 2 BMDL 3601.91 3 4 5 E .3 .2 2 .3 . Figurefor Selected Model: Power Power Model with 0.95 Confidence Level 6 7 8 9 E .3.22.4. Outputfor Additional Model Presented: Power, Unrestricted 10 L i et al., 1997: F S H 11 12 13 14 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 15 I n p u t D a t a F i l e : C : \ 1 \ 7 2 _ L i _ 1 9 9 7 _ F S H _ P w r _ U _ 1 . ( d ) 16 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 7 2 _ L i _ 1 9 9 7 _ F S H _ P w r _ U _ 1 . p l t 17 T u e F e b 16 2 0 : 0 7 : 3 3 2 0 1 0 18 19 20 F i g u r e 3: F S H i n f e m a l e S - D r a t s 2 4 h r a f t e r d o s i n g , 22 d a y o l d r a t s 21 22 23 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 24 25 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 26 27 28 D e p e n d e n t v a r i a b l e = M e a n This document is a draftfor review purposes only and does not constitute Agency policy. E-377 DRAFT--DO NOT CITE OR QUOTE 1 Independent variable = Dose 2 The power is not restricted 3 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) rho) 4 5 Total number of dose groups = 10 6 Total number of records with missing values = 0 7 Maximum number of iterations = 250 8 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 9 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 10 11 12 13 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 14 lalpha = 9.8191 15 rho = 0 16 control = 22.1591 17 slope = 26.1213 18 power = 0.264963 19 20 21 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 22 23 lalpha rho control slope power 24 25 l a l p h a 1 -0.99 -0.69 -0.15 0.28 26 27 rho -0.99 1 0.65 0.11 -0.26 28 29 c o n t r o l -0.69 0.65 1 -0.17 0.024 30 CO G o CO g o 31 slope -0.15 0.11 -0.17 1 32 33 p o w e r 0.28 -0.26 0.024 1 34 35 36 37 P a r a m e t e r E s t i m a t e s 38 39 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 40 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 41 lalpha 3.72156 1.13117 1.5045 5.93861 42 rho 1.17032 0.223249 0.732758 1.60788 43 control 15.7412 6.97367 2.07307 29.4094 44 slope 24.963 6.42976 12.3609 37.5651 45 power 0.278637 0.0312355 0.217417 0.339857 46 47 48 49 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 50 51 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 52 53 54 0 10 23.9 15.7 29.6 32.3 0.796 CO CO \-- 1 55 3 10 22.2 49.6 48.5 63.2 56 10 10 85.2 63.2 94.3 72.7 0.96 57 30 10 73.3 80.1 48.5 83.6 -0.259 58 100 10 126 106 159 98.4 0.654 59 300 10 132 138 116 115 -0.164 60 1000 10 117 187 51.2 137 -1.62 61 3000 10 304 248 154 162 1.1 62 1 e + 0 0 4 10 347 341 151 195 0.0999 63 3 e + 0 0 4 10 455 457 286 232 -0.0271 64 65 66 67 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 68 69 70 M o d e l A1: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-378 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma^2 2 3 Model A2: Yij Mu(i) + e(ij) 4 Var{e(ij)} Sigma(i)^2 5 6 Model A3: Yij = Mu(i) + e(ij) 7 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 8 Model A3 uses any fixed variance parameters that 9 were specified by the user 10 11 M o d e l R: Yi = M u + e(i) 12 V a r { e ( i ) } = S i g m a ^ 2 13 14 15 L i k e l i h o o d s o f I n t e r e s t 16 17 Model Log(likelihood) # Param's AIC 18 A1 -535.687163 11 1 0 9 3 . 3 7 4 3 2 7 19 A2 -496.367061 20 1032.734122 20 A3 -502.709623 12 1 0 2 9 . 4 1 9 2 4 6 21 fitted -513.910636 5 1037.821272 22 R -574.835246 2 1153.670492 23 24 25 E x p l a n a t i o n o f T e s t s 26 27 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 28 (A2 vs. R) 29 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 30 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 31 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 32 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 33 34 T e s t s o f I n t e r e s t 35 36 Test -2*log(Likelihood Ratio) Test df p-value 37 38 T e s t 1 156.936 18 <.0001 39 T e s t 2 78.6402 9 <.0001 40 T e s t 3 12.6851 8 0.1232 41 T e s t 4 22.402 7 0.002165 42 43 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 44 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 45 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 46 47 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 48 m o d e l a p p e a r s t o b e a p p r o p r i a t e 49 50 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 51 t o b e a p p r o p r i a t e h e r e 52 53 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 54 m o d e l 55 56 57 B e n c h m a r k D o s e C o m p u t a t i o n 58 59 S p e c i f i e d e f f e c t = 1 60 61 R i s k T y p e = Estimated standard deviations from the control mean 62 63 C o n f i d e n c e l e v e l = 0.95 64 65 B M D = 2 . 5 0 8 3 9 66 67 68 B M D L = 0 . 2 5 2 5 4 1 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-379 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E.3.22.5. F igure f o r A d d itio n a l M o d el Presented: Power, Unrestricted Power Model with 0.95 Confidence Level 2 20:07 02/16 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-380 DRAFT--DO NOT CITE OR QUOTE 1 E.3.23. Li et al., 2006: Estradiol, 3-Day 2 E.3.23.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M 2) 2 0.147 269.146 3.044E+02 1.108E+02 Notes exponential (M 3) 2 0 .1 4 7 2 6 9 .1 4 6 3 .0 4 4 E + 0 2 1.108E + 02 p ow er hit b ound (d = 1) exponential (M 4) 1 0.341 268.212 error error exponential (M 5) 0 N /A 270.212 error error H ill 0 N /A 270.212 error error linear b polynom ial, 3degree power 2 0.151 269.084 3.471E+02 1.082E+02 2 0.151 269.084 3.471E+02 1.082E+02 2 0.151 2 6 9 .0 8 4 3 .4 7 1 E + 0 2 1.082E + 02 p ow er b oun d hit (p ow er = 1) H ill, unrestricted 0 N /A 270.266 1.059E+17 1.059E+17 unrestricted (n = 0.025) power, unrestricted 1 0.327 268.266 3.727E +14 error a C onstant variance m odel selected (p = 0 .4 3 7 2 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3 .2 3 .2 . Outputfor Selected Model: Linear unrestricted (power = 0.012) 6 Li et al., 2006: Estradiol, 3-D ay 7 8 9 10 P o l y n o m i a l M o d e l . ( V e r s i o n : 2 . 1 3 ; D a t e : 0 4 / 0 8 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 3 1 _ L i _ 2 0 0 6 _ E s t r a _ L i n e a r C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 3 1 _ L i _ 2 0 0 6 _ E s t r a _ L i n e a r C V _ 1 . p l t 13 T u e F e b 16 1 8 : 1 3 : 5 6 2 0 1 0 14 15 16 F i g u r e 3, 3 - d a y e s t r a d i o l 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 S i g n s o f t h e p o l y n o m i a l c o e f f i c i e n t s a r e n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t This document is a draftfor review purposes only and does not constitute Agency policy. E-381 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 4 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha = 267.211 12 rho = 0S p e c i f i e d 13 beta_0 = 16.4428 14 b e t a 1 = 0 . 0 4 6 8 3 5 1 15 16 17 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 18 19 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 20 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 21 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 22 23 alpha beta 0 beta 1 24 25 a l p h a 1 -2.6e-013 -4.5e-015 26 CO CO o CO CO o 27 beta 0 -2.6e-013 1 28 29 beta 1 -4.5e-015 1 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 36 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 37 alpha 264.303 59.1 148.469 380.137 38 beta_0 16.4428 3.50431 9.57445 23.3111 39 beta 1 0.0468351 0.062677 -0.0760095 0.16968 40 41 42 43 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 44 45 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 46 47 48 0 10 10.2 16.4 12.2 16.3 -1.22 49 2 10 19.9 16.5 20 16.3 0.656 50 50 10 24.7 18.8 14.6 16.3 1.16 51 100 10 18.1 21.1 17.6 16.3 -0.591 52 53 54 55 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 56 57 58 M o d e l A1: Yij = Mu(i) + e(ij) 59 V a r { e i ;i j ) } = S i g m a ^ 2 60 61 M o d e l A2: Yij = Mu(i) + e(ij) 62 V a r { e i :i j ) } = S i g m a ( i ) ^2 63 64 M o d e l A3: Yij = Mu(i) + e(ij) 65 V a r { e i :i j ) } = S i g m a ^ 2 66 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 67 w e r e s p e c i f i e d b y t h e u s e r 68 69 M o d e l R: Yi = M u + e(i) 70 V a r { e ( i ) } = S i g m a ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-382 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Likelihoods of Interest 4 5 Model Log(likelihood) # Param's AIC 6 A1 -129.653527 5 269.307054 7 A2 -128.294657 8 272.589314 8 A3 -129.653527 5 269.307054 9 fitted -131.541911 3 269.083823 10 R -131.819169 2 267.638338 11 12 13 E x p l a n a t i o n o f T e s t s 14 15 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 16 (A2 vs. R) 17 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 18 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 19 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 20 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 21 22 T e s t s o f I n t e r e s t 23 24 Test -2*log(Likelihood Ratio ) Test df p-value 25 26 T e s t 1 7.04902 6 0.3163 27 T e s t 2 2.71774 3 0.4372 28 T e s t 3 2.71774 3 0.4372 29 T e s t 4 3.77677 2 0.1513 30 31 p - v a l u e f o r T e s t 1 is g r e a t e r t h a n .05. T h e r e m a y n o t b e 32 d i f f e n c e b e t w e e n r e s p o n s e s a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 33 M o d e l l i n g t h e d a t a w i t h a d o s e / r e s p o n s e c u r v e m a y n o t b e a p p r o p r i a t e 34 35 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 36 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 37 38 39 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 40 t o b e a p p r o p r i a t e h e r e 41 42 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 43 t o a d e q u a t e l y d e s c r i b e t h e d a t a 44 45 46 B e n c h m a r k D o s e C o m p u t a t i o n 47 48 S p e c i f i e d e f f e c t = 1 49 50 R i s k T y p e = Estimated standard deviations from the control mean 51 52 C o n f i d e n c e l e v e l = 0.95 53 54 BMD = 347.12 55 56 57 BMDL = 108.173 58 This document is a draftfor review purposes only and does not constitute Agency policy. E-383 DRAFT--DO NOT CITE OR QUOTE 1 E .3 .2 3 .3 . Figurefor Selected Model: Linear Linear Model with 0.95 Confidence Level Mean Response 2 18:13 02/16 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-384 DRAFT--DO NOT CITE OR QUOTE 1 E.3.24. Li et al., 2006: Progesterone, 3-Day 2 E.3.24.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M 2) 2 <0.001 330.234 5.252E +01 error Notes exponential (M 3) exponential (M4) b 2 <0.001 330.234 5.252E +01 error p o w er hit boun d (d = 1) 1 0.384 315.734 1.353E-01 8.351E-02 exponential (M 5) 0 N /A 317.734 5.225E-01 7.503E-02 H ill 1 0 .3 8 6 3 1 5 .7 2 9 1 .135E -02 1.161E -05 n lo w er b ound hit (n = 1) linear 2 <0.001 331.121 7.765E+01 5.264E+01 polynom ial, 3degree 2 <0.001 331.121 7.765E+01 5.264E+01 power 2 < 0.001 3 3 1.121 7 .7 6 5 E + 0 1 5 .264E + 01 p o w er b ound hit (p ow er = 1) power, unrestricted 1 0.405 315.670 1.066E-63 1.066E-63 unrestricted (power = 0.009) a N on-con stant variance m odel selected (p = 0 .0 0 1 3 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .2 4 .2 . Outputfor Selected Model: Exponential (M4) 6 Li et al., 2006: Progesterone, 3-D ay 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 3 2 _ L i _ 2 0 0 6 _ P r o g e s t _ E x p _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 T u e F e b 16 1 8 : 1 4 : 3 1 2 0 1 0 14 15 16 F i g u r e 4, 3 - d a y p r o g e s t e r o n e 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. This document is a draftfor review purposes only and does not constitute Agency policy. E-385 DRAFT--DO NOT CITE OR QUOTE 1 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 2 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 3 4 5 Dependent variable = Mean 6 Independent variable = Dose 7 Data are assumed to be distributed: normally 8 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 9 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) * rho) 10 11 T o t a l n u m b e r o f d o s e g r o u p s = 4 12 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 13 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 14 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 16 17 M L E s o l u t i o n p r o v i d e d : E x a c t 18 19 20 I n i t i a l P a r a m e t e r V a l u e s 21 22 Variable Model 4 23 24 lnalpha 11.3313 25 rho -1.44835 26 a 6 4 . 8 2 7 4 27 b 0 . 0 4 5 6 9 0 6 28 c 0 . 1 6 6 8 4 4 29 d 1 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 Variable Model 4 36 37 lnalpha 14.074 38 rho -2.27065 39 a 6 1 . 7 4 7 4 40 b 2 . 1 3 3 2 7 41 c 0 . 3 1 8 5 6 6 42 d 1 43 44 45 T a b l e o f S t a t s F r o m I n p u t D a t a 46 47 Dose N Obs Mean Obs Std Dev 48 49 0 10 61.74 11.1 50 2 10 30.56 40.48 51 50 10 16.93 33.3 52 100 10 11.36 43.75 53 54 55 E s t i m a t e d V a l u e s o f I n t e r e s t 56 57 Dose Est Mean Est Std Scaled Residual 58 59 0 61.75 10.55 -0.002085 60 2 20.26 37.38 0.8713 61 50 19.67 38.66 -0.224 62 100 19.67 38.66 -0.6801 63 64 65 66 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d : 67 68 Model A1: Yij = Mu(i) + e ( i j 69 V a r { e ( i j ) } = S i g m a ^ 2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-386 DRAFT--DO NOT CITE OR QUOTE 1 Model A2: Yij Mu(i) + e(ij) 2 Var{e(ij)} Sigma(i)^2 3 4 Model A3: Yij Mu(i) + e(ij) 5 Var{e(ij)} exp(lalpha + log(mean(i) rho) 6 7 M o d e l R: Yij M u + e(i) 8 Var{e(ij)} Sigma^2 9 10 11 L i k e l i h o o d s o f I n t e r e s t 12 13 Model Log(likelihood) DF AIC 14 15 A1 -159.6327 5 329.2653 16 A2 -151.8128 8 319.6255 17 A3 -152.4882 6 316.9763 18 R -165.6989 2 335.3978 19 4 -152.8668 5 315.7335 20 21 22 Additive constant for all log-likelihoods = -36.76. This constant added to the 23 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 24 d e p e n d o n t h e m o d e l p a r a m e t e r s . 25 26 27 E x p l a n a t i o n o f T e s t s 28 29 T e s t 1 D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 30 T e s t 2 A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 31 T e s t 3 A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 32 33 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 34 35 36 T e s t s o f I n t e r e s t 37 38 T e s t -2*log(Likelihood Ratio) D. F. p-value 39 40 T e s t 1 27.77 6 0.0001037 41 T e s t 2 15.64 3 0.001344 42 T e s t 3 1.351 2 0.5089 43 T e s t 6a 0.7572 1 0.3842 44 45 46 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 47 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 48 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 49 50 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 51 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e . 52 53 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 54 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 55 56 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 57 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 58 59 60 B e n c h m a r k D o s e C o m p u t a t i o n s : 61 62 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 63 64 R i s k T y p e = E s t i m a t e d s t a n d a r d d e v i a t i o n s f r o m c o n t r o l 65 66 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 67 68 BMD = 0.135296 69 70 BMDL = 0.0835054 This document is a draftfor review purposes only and does not constitute Agency policy. E-387 DRAFT--DO NOT CITE OR QUOTE 1 2 E .3 .2 4 .3 . Figurefor Selected Model: Exponential (M4) Exponential_beta Model 4 with 0.95 Confidence Level Exponential 60 40 Mean Response 20 0 -20 BMDLIBMD 0 20 40 60 80 3 18:14 02/16 2010 4 dose 5 6 E .3.24.4. Outputfor Additional Model Presented: Hill, Unrestricted 100 7 Li et al., 2006: Progesterone, 3-D ay 8 9 10 11 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 12 I n p u t D a t a F i l e : C : \ 1 \ 3 2 _ L i _ 2 0 0 6 _ P r o g e s t _ H i l l _ U _ 1 . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 3 2 _ L i _ 2 0 0 6 _ P r o g e s t _ H i l l _ U _ 1 . p l t 14 T u e F e b 16 1 8 : 1 4 : 4 1 2 0 1 0 15 16 17 F i g u r e 4, 3 - d a y p r o g e s t e r o n e 18 19 20 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 21 22 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 23 24 25 D e p e n d e n t v a r i a b l e = M e a n 26 I n d e p e n d e n t v a r i a b l e = D o s e 27 P o w e r p a r a m e t e r is n o t r e s t r i c t e d 28 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + r h o * l n ( m e a n ( i ) ) ) This document is a draftfor review purposes only and does not constitute Agency policy. E-388 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 4 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 lalpha 7.08699 12 rho 0 13 intercept 61.7404 14 - 5 0 . 3 8 3 5 15 1 . 4 3 9 9 7 16 1 . 6 1 5 9 17 18 19 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 20 21 ( * * * T h e m o d e l p a r a m e t e r ( s ) -k 22 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 23 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 24 25 lalpha rho intercept v n 26 27 l a l p h a 1 -0.99 -0.097 0.84 NA 28 29 rho -0.99 1 0.13 -0.81 NA 30 CO -tr O CO -tr O 31 i n t e r c e p t -0.097 0.13 1 NA 32 33 v 0.84 -0.81 1 NA 34 35 n N A N A N A N A N A 36 37 38 N A - T h i s p a r a m e t e r ' s v a r i a n c e h a s b e e n e s t i m a t e d as z e r o o r l e s s . 39 T H E M O D E L H A S P R O B A B L Y N O T C O N V E R G E D ! ! ! 40 41 42 43 P a r a m e t e r E s t i m a t e s 44 45 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 46 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 47 lalpha 13.9863 NA NA NA 48 rho -2.25026 NA NA NA 49 i n t e r c e p t 61.7404 NA NA NA 50 v -42.1239 NA NA NA 51 n 2.02774 NA NA NA 52 k 1e-013 NA 53 54 A t l e a s t s o m e v a r i a n c e e s t i m a t e s a r e n e g a t i v e . 55 T H I S U S U A L L Y M E A N S T H E M O D E L H A S N O T C O N V E R G E D ! 56 T r y a g a i n f r o m a n o t h e r s t a r t i n g p o i n t 57 58 59 60 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 61 62 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 63 64 65 0 10 61.7 61.7 11.1 10.5 9.74e-008 66 2 10 30.6 19.6 40.5 38.3 0.905 67 50 10 16.9 19.6 33.3 38.3 -0.222 68 100 10 11.4 19.6 43.7 38.3 -0.683 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-389 DRAFT--DO NOT CITE OR QUOTE 1 2 Model Descriptions for likelihoods calculated 3 4 5 Model A1: Yij Mu(i) + e(ij) 6 Var{e(ij)} Sigma^2 7 8 Model A2: Yij Mu(i) + e(ij) 9 Var{e(ij)} Sigma(i)^2 10 11 M o d e l A3: Yij = Mu(i) + e(ij) 12 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 13 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 14 w e r e s p e c i f i e d b y t h e u s e r 15 16 M o d e l R: Yi = M u + e(i) 17 V a r { e ( i ) } = S i g m a ^ 2 18 19 20 L i k e l i h o o d s o f I n t e r e s t 21 22 Model Log(likelihood) # Param's AIC 23 A1 -159.632675 5 329.265349 24 A2 -151.812765 8 319.625529 25 A3 -152.488175 6 316.976349 26 fitted -152.873643 5 315.747285 27 R -165.698875 2 335.397750 28 29 30 E x p l a n a t i o n o f T e s t s 31 32 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 33 (A2 vs. R) 34 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 35 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 36 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 37 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 38 39 T e s t s o f I n t e r e s t 40 41 Test -2*log(Likelihood Ratio) Test df p-value 42 43 T e s t 1 27.7722 6 0.0001037 44 T e s t 2 15.6398 3 0.001344 45 T e s t 3 1.35082 2 0.5089 46 T e s t 4 0.770936 1 0.3799 47 48 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 49 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 50 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 51 52 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 53 m o d e l a p p e a r s t o b e a p p r o p r i a t e 54 55 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 56 t o b e a p p r o p r i a t e h e r e 57 58 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 59 t o a d e q u a t e l y d e s c r i b e t h e d a t a 60 61 62 B e n c h m a r k D o s e C o m p u t a t i o n 63 64 S p e c i f i e d e f f e c t = 1 65 66 R i s k T y p e = Estimated standard deviations from the control mean 67 68 C o n f i d e n c e l e v e l = 0.95 69 70 B M D = 5 . 8 1 7 0 3 e - 0 1 4 This document is a draftfor review purposes only and does not constitute Agency policy. E-390 DRAFT--DO NOT CITE OR QUOTE 1 2 BMDL = 5.81703e-014 3 4 5 E .3.24.5. Figurefor Additional Model Presented: Hill, Unrestricted Hill Model with 0.95 Confidence Level Mean Response 6 18:14 02/16 2010 7 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-391 DRAFT--DO NOT CITE OR QUOTE 1 E.3.25. Markowski et al., 2001: FR10 Run Opportunities 2 E.3.25.1. S u m m a ry Table o f B M D S M odeling R esults Model a exponential (M2) b Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) 2 0.248 117.557 1.653E+02 5.025E+01 Notes exponential (M 3) 2 0.248 1 1 7.557 1 .653E + 02 5.0 2 5 E + 0 1 p o w er hit bound (d = 1) exponential (M 4) 1 0.412 117.445 4.742E+01 1.729E-01 exponential (M 5) 0 N /A 118.918 3.178E+01 3.967E-05 Hill 0 N /A 118.918 2.348E+01 6.728E-06 linear 2 0.190 118.089 2.081E+02 1.051E+02 polynom ial, 3degree 2 0.190 118.089 2.081E+02 1.051E+02 power 2 0 .1 9 0 1 1 8.089 2 .0 8 1 E + 0 2 1 .051E + 02 p o w er b oun d hit (p ow er = 1) power, unrestricted 1 0.238 118.164 9.153E+01 5.911E -07 unrestricted (pow er = 0.237) a C onstant variance m odel selected (p = 0 .1 7 1 9 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3 .2 5 .2 . Outputfor Selected Model: Exponential (M2) 6 M arkowski et al., 2001: FR 10 Run Opportunities 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 3 3 _ M a r k _ 2 0 0 1 _ F R 1 0 o p p _ E x p C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 T u e F e b 16 1 8 : 1 5 : 2 6 2 0 1 0 14 15 16 T a b l e 3 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 30 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 31 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. This document is a draftfor review purposes only and does not constitute Agency policy. E-392 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Dependent variable = Mean 4 Independent variable = Dose 5 Data are assumed to be distributed: normally 6 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 7 r h o is s e t t o 0. 8 A constant v a r i a n c e model is fit. 9 10 T o t a l n u m b e r o f d o s e g r o u p s = 4 11 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 12 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 13 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 16 M L E s o l u t i o n p r o v i d e d : E x a c t 17 18 19 I n i t i a l P a r a m e t e r V a l u e s 20 21 Variable Model 2 22 23 lnalpha 3.5321 24 rho(S) 0 25 a 6 . 9 8 1 6 9 26 b 0 . 0 0 3 0 9 8 9 1 27 c 0 28 d 1 29 30 (S) = S p e c i f i e d 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 Variable Model 2 37 38 lnalpha 3.64823 39 rho 0 40 a 1 1 . 9 4 4 3 41 b 0 . 0 0 4 4 2 6 2 42 c 0 43 d 1 44 45 46 T a b l e o f S t a t s F r o m I n p u t D a t a 47 48 Dose N Obs Mean Obs Std Dev 49 50 07 13.29 8.65 51 20 4 11.25 5.56 52 60 6 5.75 3.53 53 180 7 7 6.01 54 55 56 E s t i m a t e d V a l u e s o f I n t e r e s t 57 58 Dose Est Mean Est Std Scaled Residual 59 60 0 11.94 6.197 0.5745 61 20 10.93 6.197 0.1025 62 60 9.158 6.197 -1.347 63 180 5.385 6.197 0.6897 64 65 66 67 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d 68 69 M o d e l A1: Yij = M u ( i ) + e(i j ) 70 V a r { e ( i j ) } = S i g m a ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-393 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A2: Yij = Mu(i) + e(ij) 3 V a r { e i ;ij)} = S i g m a ( i ) ^ 2 4 5 Model A3: Yij = Mu(i) + e(ij) 6 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 7 8 M o d e l R: Yij = M u + e(i) 9 V a r { e i :ij)} = S i g m a ^ 2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) DF AIC 15 16 A1 -54.38526 5 118.7705 17 A2 -51.88568 8 119.7714 18 A3 -54.38526 5 118.7705 19 R -57.45429 2 118.9086 20 2 -55.77871 3 117.5574 21 22 23 Additive constant for all log-likelihoods = -22.05. This constant added to the 24 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 25 d e p e n d o n t h e m o d e l p a r a m e t e r s . 26 27 28 E x p l a n a t i o n o f T e s t s 29 30 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 31 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 32 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 33 T e s t 4: D o e s M o d e l 2 f i t t h e d a t a ? (A3 vs. 2) 34 35 36 T e s t s o f I n t e r e s t 37 38 T e s t -2*log(Likelihood Ratio) D. F. p-value 39 40 T e s t 1 11.14 6 0.08423 41 T e s t 2 4.999 3 0.1719 42 T e s t 3 4.999 3 0.1719 43 T e s t 4 2.787 2 0.2482 44 45 46 T h e p - v a l u e f o r T e s t 1 is g r e a t e r t h a n .05. T h e r e m a y n o t b e a 47 d i f f e n c e b e t w e e n r e s p o n s e s a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 48 M o d e l l i n g t h e d a t a w i t h a d o s e / r e s p o n s e c u r v e m a y n o t b e a p p r o p r i a t e . 49 50 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 51 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e . 52 53 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 54 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 55 56 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. M o d e l 2 s e e m s 57 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 58 59 60 B e n c h m a r k D o s e C o m p u t a t i o n s : 61 62 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 63 64 R i s k T y p e = E s t i m a t e d s t a n d a r d d e v i a t i o n s f r o m c o n t r o l 65 66 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 67 68 BMD = 165.284 69 70 BMDL = 50.2488 This document is a draftfor review purposes only and does not constitute Agency policy. E-394 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E .3 .2 5 .3 . Figurefor Selected Model: Exponential (M2) Exponential_beta Model 2 with 0.95 Confidence Level 2 18:15 02/16 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-395 DRAFT--DO NOT CITE OR QUOTE 1 E.3.26. Markowski et al., 2001: FR2 Revolutions 2 E.3.26.1. S u m m a ry Table o f B M D S M odeling R esults Model a Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M 2) 2 0.192 217.636 1.627E+02 5.807E+01 Notes exponential (M 3) 2 0 .1 9 2 2 1 7 .6 3 6 1.627E + 02 5 .807E + 01 p ow er hit b ound (d = 1) exponential (M 4) 1 0.298 217.415 4.668E+01 1.965E-01 exponential (M 5) Hill b 0 N /A 218.532 3.308E+01 1.193E+01 0 N/A 218.532 2.364E+01 7.336E+00 n upper bound hit (n = 18) linear 2 0.150 218.129 1.989E+02 1.025E+02 polynom ial, 3degree 2 0.150 218.129 1.989E+02 1.025E+02 power 2 0 .1 5 0 2 1 8 .1 2 9 1.989E + 02 1.025E + 02 p ow er b oun d hit (p ow er = 1) power, unrestricted c 1 0.160 218.302 9.101E+01 1.800E-13 unrestricted (power = 0.272) a C onstant variance m odel selected (p = 0 .1 0 9 2 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .2 6 .2 . Outputfor Selected Model: Hill 6 M arkowski et al., 2001: FR2 R evolutions 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 3 4 _ M a r k _ 2 0 0 1 _ F R 2 r e v _ H i l l C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 3 4 _ M a r k _ 2 0 0 1 _ F R 2 r e v _ H i l l C V _ 1 . p l t 13 ~ T u e F e b 16 1 8 : 1 6 : 0 3 2 0 1 0 14 15 16 T a b l e 3 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 P o w e r p a r a m e t e r r e s t r i c t e d t o b e g r e a t e r t h a n 1 28 A c o n s t a n t v a r i a n c e m o d e l is f i t 29 This document is a draftfor review purposes only and does not constitute Agency policy. E-396 DRAFT--DO NOT CITE OR QUOTE 1 Total number of dose groups = 4 2 Total number of records with missing values = 0 3 Maximum number of iterations = 250 4 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 5 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 6 7 8 9 Default Initial Parameter Values 10 alpha 2598.74 11 rho 0 Specified 12 intercept 119.29 13 - 6 2 . 7 9 14 1 . 8 0 6 0 2 15 3 5 . 8 5 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 21 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 22 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 23 24 alpha intercept v n k 25 26 a l p h a 1 -8.1e-009 4.5e-008 -3e-005 3e-005 27 28 intercept -8.1e-009 1 -0.81 -0.00013 -0.0022 29 30 v 4.5e-008 -0.81 1 0.0002 0.0014 31 32 n -3e-005 -0.00013 0.0002 1 -1 33 34 k 3e-005 -0.0022 0.0014 -1 1 35 36 37 38 P a r a m e t e r E s t i m a t e s 39 40 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 41 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 42 alpha 2183.85 630.425 948.245 3419.46 43 i n t e r c e p t 119.29 17.6629 84.6713 153.909 44 v -56.5223 21.9082 -99.4615 -13.5831 45 n 18 8854.08 -17335.7 17371.7 46 k 21.6708 855.263 -1654.61 1697.95 47 48 49 50 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 51 52 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res. 53 54 55 0 7 1 1 9 119 69.9 46.7 2.74e-008 56 20 4 109 108 61 46.7 8.42e-010 57 60 6 5 6 . 5 62.8 31.2 46.7 -0.329 58 180 7 68.1 62.8 33.2 46.7 0.304 59 60 D e g r e e s o f f r e e d o m f o r T e s t A 3 v s f i t t e d < = 0 61 62 63 64 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 65 66 67 M o d e l A1: Yij = Mu(i) + e(ij) 68 V a r { e ( i j ) } = S i g m a ^ 2 69 70 M o d e l A2: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-397 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma(i)^2 2 3 Model A3: Yij = Mu(i) + e(ij) 4 Var{e(ij)} = Sigma^2 5 Model A3 uses any fixed variance parameters that 6 were specified by the user 7 8 M o d e l R: Yi = M u + e(i) 9 Var{e(i)} = Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) # Param's AIC 15 A1 -104.165520 5 218.331040 16 A2 -101.140174 8 218.280349 17 A3 -104.165520 5 218.331040 18 fitted -104.266162 5 218.532324 19 R -107.599268 2 219.198536 20 21 22 E x p l a n a t i o n o f T e s t s 23 24 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 25 (A2 vs. R) 26 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 27 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 28 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 29 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 30 31 T e s t s o f I n t e r e s t 32 33 Test -2*log(Likelihood Ratio) Test df p-vali 34 35 T e s t 1 12.9182 6 0.04435 36 T e s t 2 6.05069 3 0.1092 37 T e s t 3 6.05069 3 0.1092 38 T e s t 4 0.201283 0 NA 39 40 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 41 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 42 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 43 44 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 45 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 46 47 48 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 49 t o b e a p p r o p r i a t e h e r e 50 51 N A - D e g r e e s o f f r e e d o m f o r T e s t 4 a r e l e s s t h a n o r e q u a l t o 0. T h e C h i - S q u a r e 52 t e s t f o r f i t is n o t v a l i d 53 54 55 B e n c h m a r k D o s e C o m p u t a t i o n 56 57 S p e c i f i e d e f f e c t = 1 58 59 R i s k T y p e = Estimated standard deviations from the control mean 60 61 C o n f i d e n c e l e v e l = 0.95 62 63 BMD = 23.6366 64 65 BMDL = 7.33648 66 This document is a draftfor review purposes only and does not constitute Agency policy. E-398 DRAFT--DO NOT CITE OR QUOTE 1 E .3 .2 6 .3 . Figurefor Selected Model: Hill Hill Model with 0.95 Confidence Level Mean Response 2 18:16 02/16 2010 3 dose 4 5 E .3.26.4. Outputfor Additional Model Presented: Power, Unrestricted 6 M arkowski et al., 2001: FR2 R evolutions 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 3 4 _ M a r k _ 2 0 0 1 _ F R 2 r e v _ P o w e r C V _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 3 4 _ M a r k _ 2 0 0 1 _ F R 2 r e v _ P o w e r C V _ U _ 1 . p l t 13 ~ T u e F e b 16 1 8 7 1 6 : 0 4 2 0 1 0 14 15 16 T a b l e 3 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 T h e p o w e r is n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t This document is a draftfor review purposes only and does not constitute Agency policy. E-399 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 4 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha = 2598.74 12 rho = 0 Specified 13 control = 119.29 14 slope = -1.79436 15 power = 0.708231 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 21 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 22 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 23 24 alpha control slope power 25 26 a l p h a 1 9.7e-009 -1.9e-008 -1.6e-008 27 28 control 9.7e-009 1 -0.49 -0.28 29 30 slope -1.9e-008 -0.49 1 0.96 31 32 power -1.6e-008 -0.28 0.96 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 39 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 40 alpha 2351 678.674 1020.82 3681.17 41 control 120.074 18.0837 84.6305 155.517 42 slope -14.1965 22.2073 -57.722 29.329 43 power 0.27229 0.301344 -0 .318334 0.862913 44 45 46 47 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 48 49 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res. 50 51 52 0 7 119 120 69.9 48.5 -0.0428 53 20 4 109 88 61 48.5 0.846 54 60 6 5 6 . 5 76.8 31.2 48.5 -1.02 55 180 7 68.1 61.7 33.2 48.5 0.352 56 57 58 59 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 60 61 62 M o d e l A1: Yij = Mu(i) + e(ij) 63 V a r j e i ;i j ) } = S i g m a ^ 2 64 65 M o d e l A2: Yij = Mu(i) + e(ij) 66 V a r j e i :i j ) } = S i g m a ( i ) ^2 67 68 M o d e l A3: Yij = Mu(i) + e(ij) 69 V a r j e i :i j ) } = S i g m a ^ 2 70 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t This document is a draftfor review purposes only and does not constitute Agency policy. E-400 DRAFT--DO NOT CITE OR QUOTE 1 were specified by the user 2 3 M o d e l R: Yi = M u + e(i) 4 Var{e(i)} = Sigma^2 5 6 7 Likelihoods of Interest 8 9 Model Log(likelihood) # Param's AIC 10 A1 -104.165520 5 218.331040 11 A2 -101.140174 8 218.280349 12 A3 -104.165520 5 218.331040 13 fitted -105.151136 4 218.302271 14 R -107.599268 2 219.198536 15 16 17 E x p l a n a t i o n o f T e s t s 18 19 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 20 (A2 vs. R) 21 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 22 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 23 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 24 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 25 26 T e s t s o f I n t e r e s t 27 28 Test -2*log(Likelihood Ratio) Test df p-val 29 30 T e s t 1 12 .9182 6 0. 0 4 4 3 5 31 T e s t 2 6. 0 5 0 6 9 3 0 . 1 0 9 2 32 T e s t 3 6. 0 5 0 6 9 3 0 . 1 0 9 2 33 T e s t 4 1. 9 7 1 2 3 1 0 . 1 6 0 3 34 35 p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t 36 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 37 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 38 39 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 40 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 41 42 43 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 44 t o b e a p p r o p r i a t e h e r e 45 46 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 47 t o a d e q u a t e l y d e s c r i b e t h e d a t a 48 49 50 B e n c h m a r k D o s e C o m p u t a t i o n 51 52 S p e c i f i e d e f f e c t = 1 53 54 R i s k T y p e = Estimated standard deviations from the control mean 55 56 C o n f i d e n c e l e v e l = 0.95 57 58 B M D = 9 1 . 0 1 4 5 59 60 61 B M D L = 1 . 8 e - 0 1 3 62 This document is a draftfor review purposes only and does not constitute Agency policy. E-401 DRAFT--DO NOT CITE OR QUOTE 1 E.3.26.5. F igure f o r A d d itio n a l M o d el Presented: Power, U nrestricted Power Model with 0.95 Confidence Level Mean Response 2 18:16 02/16 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-402 DRAFT--DO NOT CITE OR QUOTE 1 E.3.27. M arkow ski et al., 2001: FR 5 R un O pportunities 2 E .3 .2 7 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M 2) 2 0.149 133.830 9.491E+01 4.324E+01 Notes exponential (M 3) 2 0.149 133.830 9 .491E + 01 4.3 2 4 E + 0 1 p ow er hit b ound (d = 1) exponential (M 4) 1 0.303 133.087 2.961E+01 9.356E+00 exponential (M 5) Hill b 0 N /A 134.032 2.871E+01 1.226E+01 1 0.939 132.032 2.214E+01 1.117E+01 n upper bound hit (n = 18) linear 2 0.091 134.825 1.349E+02 8.118E+01 polynom ial, 3degree 2 0.091 134.825 1.349E+02 8.118E+01 power 2 0.091 134.825 1.349E + 02 8 .118E + 01 p ow er b oun d hit (p ow er = 1) power, unrestricted c 1 0.133 134.281 3.721E+01 1.439E-07 unrestricted (power = 0.336) a C onstant variance m odel selected (p = 0 .2 2 6 2 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .2 7 .2 . Outputfor Selected Model: Hill 6 M arkowski et al., 2001: FR5 Run Opportunities 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 3 5 _ M a r k _ 2 0 0 1 _ F R 5 o p p _ H i l l C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 3 5 _ M a r k _ 2 0 0 1 _ F R 5 o p p _ H i l l C V _ 1 . p l t 13 ~ T u e F e b 16 1 8 : 1 6 : 3 9 2 0 1 0 14 15 16 T a b l e 3 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 P o w e r p a r a m e t e r r e s t r i c t e d t o b e g r e a t e r t h a n 1 28 A c o n s t a n t v a r i a n c e m o d e l is f i t 29 This document is a draftfor review purposes only and does not constitute Agency policy. E-403 DRAFT--DO NOT CITE OR QUOTE 1 Total number of dose groups = 4 2 Total number of records with missing values = 0 3 Maximum number of iterations = 250 4 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 5 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 6 7 8 9 Default Initial Parameter Values 10 alpha 77.4849 11 rho 0 Specified 12 intercept 26.14 13 - 1 3 . 3 4 14 2 . 3 6 0 0 2 15 3 5 . 0 6 5 4 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 ( *** The model parameter(s) -rho -n 21 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 22 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 23 24 alpha intercept v k 25 26 a l p h a 1 -3.6e-009 9.8e-009 3.6e-008 27 28 i n t e r c e p t - 3 . 6 e - 0 0 9 1 -0.81 -0.51 29 30 9.8e-009 -0.81 1 0.36 31 32 k 3.6e-008 -0.51 0.36 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 39 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Lim 40 alpha 64.5863 18.6445 28.0438 101.129 41 i n t e r c e p t 26.14 3.03753 20.1865 32.0935 42 v -13.1569 3.7676 -20.5413 -5.77257 43 n 18 N A 44 k 21.5963 2.68136 16.3409 26.8517 45 46 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 47 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 48 h a s n o s t a n d a r d e r r o r . 49 50 51 52 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 53 54 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled R 55 56 57 07 26.1 26.1 12.3 8.04 1.02e-008 58 20 4 23.5 23.5 7.04 8.04 -1.39e-007 59 60 6 1 2 . 8 13 6.17 8.04 -0.0558 60 180 7 13.1 13 7.14 8.04 0.0517 61 62 63 64 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 65 66 67 M o d e l A1: Yij = Mu(i) + e(ij) 68 V a r { e ( i j ) } = S i g m a ^ 2 69 70 M o d e l A2: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-404 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma(i)^2 2 3 Model A3: Yij = Mu(i) + e(ij) 4 Var{e(ij)} = Sigma^2 5 Model A3 uses any fixed variance parameters that 6 were specified by the user 7 8 M o d e l R: Yi = M u + e(i) 9 Var{e(i)} = Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) # Param's AIC 15 A1 -62.013133 5 134.026266 16 A2 -59.839035 8 135.678070 17 A3 -62.013133 5 134.026266 18 fitted -62.016024 4 132.032049 19 R -67.530040 2 139.060081 20 21 22 E x p l a n a t i o n o f T e s t s 23 24 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 25 (A2 vs. R) 26 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 27 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 28 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 29 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 30 31 T e s t s o f I n t e r e s t 32 33 Test -2*log(Likelihood Ratio) Test df p-val 34 35 T e s t 1 15.382 6 0.01748 36 T e s t 2 4.3482 3 0.2262 37 T e s t 3 4.3482 3 0.2262 38 T e s t 4 0.0057833 1 0.9394 39 40 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 41 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 42 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 43 44 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 45 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 46 47 48 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 49 t o b e a p p r o p r i a t e h e r e 50 51 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 52 t o a d e q u a t e l y d e s c r i b e t h e d a t a 53 54 55 B e n c h m a r k D o s e C o m p u t a t i o n 56 57 S p e c i f i e d e f f e c t = 1 58 59 R i s k T y p e = Estimated standard deviations from the control mean 60 61 C o n f i d e n c e l e v e l = 0.95 62 63 BMD = 22.144 64 65 BMDL = 11.165 66 This document is a draftfor review purposes only and does not constitute Agency policy. E-405 DRAFT--DO NOT CITE OR QUOTE 1 E .3 .2 7 .3 . Figurefor Selected Model: Hill Hill Model with 0.95 Confidence Level Mean Response 2 18:16 02/16 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-406 DRAFT--DO NOT CITE OR QUOTE 1 E.3.27.4. O utputf o r A d d itio n a l M o d el Presented: Power, U nrestricted 2 Markowski et al., 2001: FR5 Run Opportunities 3 4 5 6 Power Model. (Version: 2.15; Date: 04/07/2008) 7 Input Data File: C:\1\35_Mark_2001_FR5opp_PwrCV_U_1.(d) 8 Gnuplot Plotting File: C:\1\35_Mark_2001_FR5opp_PwrCV_U_1.plt 9 ~ Tue Feb 16 18716:40 2010 10 11 12 T a b l e 3 13 14 15 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 16 17 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 18 19 20 D e p e n d e n t v a r i a b l e = M e a n 21 I n d e p e n d e n t v a r i a b l e = D o s e 22 r h o is s e t t o 0 23 T h e p o w e r is n o t r e s t r i c t e d 24 A c o n s t a n t v a r i a n c e m o d e l is f i t 25 26 T o t a l n u m b e r o f d o s e g r o u p s = 4 27 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 28 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 29 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 30 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 31 32 33 34 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 35 alpha = 77.4849 36 rho = 0 Specified 37 control = 26.14 38 slope = -0.39517 39 power = 0.725538 40 41 42 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 43 44 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 45 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 46 a n d d o n o t a p p e a r in t h e c o r r e l a t i o n m a t r i x ) 47 48 alpha control slope power 49 50 a l p h a 1 7.4e-009 4.3e-008 4.8e-008 51 CO o CO o 52 control 7.4e-009 1 -0.51 53 54 slope 4.3e-008 -0.51 1 0.97 55 56 power 4.8e-008 0.97 1 57 58 59 60 P a r a m e t e r E s t i m a t e s 61 62 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 63 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 64 alpha 70.9323 20.4764 30.7993 111.065 65 control 26.3567 3.13032 20.2213 32.492 66 slope -2.49841 3.16984 -8.71118 3.71437 67 power 0.336003 0.242031 -0.138368 0.810375 68 This document is a draftfor review purposes only and does not constitute Agency policy. E-407 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Table of Data and Estimated Values of Interest 4 5 Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled R 6 7 8 07 26.1 26.4 12.3 8.42 -0.0681 9 20 4 23.5 19.5 7.04 8.42 0.945 10 60 6 1 2 . 8 16.5 6.17 8.42 -1.07 11 180 7 13.1 12.1 7.14 8.42 0.341 12 13 14 15 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 16 17 18 M o d e l A1: Yij = Mu(i) + e(ij) 19 V a r { e i ;i j ) } = S i g m a ^ 2 20 21 M o d e l A2: Yij = Mu(i) + e(ij) 22 V a r { e i :i j ) } = S i g m a ( i ) ^2 23 24 M o d e l A3: Yij = Mu(i) + e(ij) 25 V a r { e i :i j ) } = S i g m a ^ 2 26 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 27 w e r e s p e c i f i e d b y t h e u s e r 28 29 M o d e l R: Yi = M u + e(i) 30 V a r { e ( i ) } = S i g m a ^ 2 31 32 33 L i k e l i h o o d s o f I n t e r e s t 34 35 Model Log(likelihood) # Param's AIC 36 A1 -62.013133 5 134.026266 37 A2 -59.839035 8 135.678070 38 A3 -62.013133 5 134.026266 39 fitted -63.140714 4 134.281428 40 R -67.530040 2 139.060081 41 42 43 E x p l a n a t i o n o f T e s t s 44 45 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 46 (A2 vs. R) 47 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 48 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 49 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 50 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 51 52 T e s t s o f I n t e r e s t 53 54 Test -2*log(Likelihood Ratio) Test df p-value 55 56 T e s t 1 15.382 6 0.01748 57 T e s t 2 4.3482 3 0.2262 58 T e s t 3 4.3482 3 0.2262 59 T e s t 4 2.25516 1 0.1332 60 61 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s to 62 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 63 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 64 65 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 66 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 67 68 69 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 70 t o b e a p p r o p r i a t e h e r e This document is a draftfor review purposes only and does not constitute Agency policy. E-408 DRAFT--DO NOT CITE OR QUOTE 1 2 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 3 to adequately describe the data 4 5 6 Benchmark Dose Computation 7 8 Specified effect = 1 9 10 R i s k T y p e = Estimated standard deviations from the control mean 11 12 C o n f i d e n c e l e v e l = 0.95 13 14 B M D = 3 7 . 2 1 3 1 15 16 17 B M D L = 1 . 4 3 9 2 6 e - 0 0 7 18 19 20 E .3.27.5. Figurefor Additional Model Presented: Power, Unrestricted Power Model with 0.95 Confidence Level Mean Response 21 18:16 02/16 2010 22 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-409 DRAFT--DO NOT CITE OR QUOTE 1 E .3.28. M iettinen et al., 2006: C ariogenic L esions, Pups 2 E .3 .2 8 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2pValue AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes gamma 3 0.345 1 62.699 7 .5 0 5 E + 0 1 4 .0 8 6 E + 0 1 p o w er b oun d hit (p ow er = 1) logistic 3 0.315 162.909 8.991E+01 5.250E+01 log-logistic a 3 0.506 161.767 3.130E+01 1.054E+01 slope bound hit (slope = 1) log-probit multistage, 4degree probit 3 0.257 163.393 1 .390E + 02 6 .7 2 9 E + 0 1 slop e b ound hit (slop e = 1) 3 0.345 162.699 7.505E+01 4.086E+01 final B = 0 3 0.299 163.031 9.941E+01 6.208E+01 W eibull 3 0.345 1 62.699 7 .5 0 5 E + 0 1 4 .0 8 6 E + 0 1 p o w er b oun d hit (p ow er = 1) gamma, unrestricted log-logistic, unrestricted b log-probit, unrestricted W eibull, unrestricted 1 .3 3 5 E - 2 0.797 161.805 1.591E-02 unrestricted (pow er = 0.184) 240 2 0.723 161.998 3.713E-01 error unrestricted (slope = 0.403) 2 0.726 161.987 5.098E-01 error unrestricted (slope = 0.25) 2 0.761 161.897 1.174E-01 error unrestricted (pow er = 0.281) aBest-fitting m odel, BM D S output presented in this appendix b Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .2 8 .2 . Outputfor Selected Model: Log-Logistic 6 M iettinen et al., 2006: C ariogenic L esions, Pups 7 8 9 10 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 3 6 _ M i e t _ 2 0 0 6 _ C a r i o g e n i c _ L o g L o g i s t i c _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 3 6 _ M i e t _ 2 0 0 6 _ C a r i o g e n i c _ L o g L o g i s t i c _ 1 . p l t 13 ~ T u e F e b 16 1 8 : 1 7 : 1 6 2 0 1 0 14 15 16 T a b l e 2 c o n v e r t i n g t h e p e r c e n t a g e i n t o t h e n u m b e r o f a n i m a l s , a n d c o n t r o l is C o n t r o l II f r o m t h e 17 s t u d y . D o s e is i n n g p e r k g a n d is f r o m T a b l e 1 18 19 20 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 21 22 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 23 24 25 D e p e n d e n t v a r i a b l e = D i c h E f f This document is a draftfor review purposes only and does not constitute Agency policy. E -410 D RAFT-- DO N O T CITE OR QUOTE 1 Independent variable = Dose 2 Slope pa r a m e t e r is r estricted as slope >= 1 3 4 Total number of observations = 5 5 Total number of records with missing values = 0 6 Maximum number of iterations = 250 7 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 8 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 9 10 11 12 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 13 14 15 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 16 background = 0.595238 17 intercept = -5.52519 18 slope = 1 19 20 21 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 22 23 ( * * * T h e m o d e l p a r a m e t e r ( s ) - s l o p e 24 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 25 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 26 27 background intercept 28 29 b a c k g r o u n d 1 -0.64 30 31 i n t e r c e p t -0.64 1 32 33 34 35 P a r a m e t e r E s t i m a t e s 36 37 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 38 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 39 b a c k g r o u n d 0.658158 40 intercept -5.64068 41 s l o p e 1 42 43 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 44 45 46 47 A n a l y s i s o f D e v i a n c e T a b l e 48 49 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 50 Full model -77.6769 5 51 F i t t e d m o d e l -78.8837 2 2.41374 3 0.4911 52 R e d u c e d m o d e l -83.2067 1 11.0597 4 0.0259 53 54 AIC: 161.767 55 56 57 G o o d n e s s o f F i t 58 S c a l e d 59 Dose Est. Prob. Expected Observed Size Residual 60 61 0.0000 0.6582 27.643 25.000 42 -0. 860 62 30.0000 0.6911 20.041 23.000 29 1. 1 8 9 63 100.0000 0.7477 18.693 19.000 25 0. 1 4 1 64 300.0000 0.8345 20.027 20.000 24 -0. 015 65 1 0 0 0 . 0 0 0 0 0.9249 29.596 29.000 32 -0. 400 66 67 C h i ^ 2 = 2 . 3 3 d.f. = 3 P-value = 0.5062 68 69 70 B e n c h m a r k D o s e C o m p u t a t i o n This document is a draftfor review purposes only and does not constitute Agency policy. E-411 DRAFT--DO NOT CITE OR QUOTE 1 2 Specified effect 3 4 Risk Type 5 6 Confidence level 7 8 BMD 9 10 B M D L 11 12 0.1 Extra risk 0.95 31.2951 10.5354 13 E .3 .2 8 .3 . Figurefor Selected Model: Log-Logistic Log-Logistic Model with 0.95 Confidence Level 14 15 16 17 E .3.28.4. Outputfor Additional Model Presented: Log-Logistic, Unrestricted 18 M iettinen et al., 2006: C ariogen ic L esio n s, Pups 19 20 21 22 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 23 I n p u t D a t a F i l e : C : \ 1 \ 3 6 _ M i e t _ 2 0 0 6 _ C a r i o g e n i c _ L o g L o g i s t i c _ U _ 1 . ( d ) 24 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 3 6 _ M i e t _ 2 0 0 6 _ C a r i o g e n i c _ L o g L o g i s t i c _ U _ 1 . p l t 25 ~ T u e F e b 16 1 8 : 1 7 : 1 8 2 0 1 0 26 27 This document is a draftfor review purposes only and does not constitute Agency policy. E-412 DRAFT--DO NOT CITE OR QUOTE 1 Table 2 c onverting the percen t a g e into the number of animals, and control is Control II from the 2 study. Dose is in ng per kg and is from Table 1 3 4 5 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 6 7 P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))] 8 9 10 D e p e n d e n t v a r i a b l e = D i c h E f f 11 I n d e p e n d e n t v a r i a b l e = D o s e 12 S l o p e p a r a m e t e r is n o t r e s t r i c t e d 13 14 T o t a l n u m b e r o f o b s e r v a t i o n s = 5 15 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 16 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 17 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 18 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 19 20 21 22 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 23 24 25 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 26 background = 0.595238 27 intercept = -1.68849 28 slope = 0.382632 29 30 31 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 32 33 background intercept slope 34 35 b a c k g r o u n d 1 -0.41 0.24 36 37 i n t e r c e p t -0.41 1 -0.96 38 39 s l o p e 0.24 -0.96 1 40 41 42 43 P a r a m e t e r E s t i m a t e s 44 45 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 46 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 47 b a c k g r o u n d 0.597778 48 intercept -1.79836 49 slope 0.402606 50 51 I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 52 53 54 55 A n a l y s i s o f D e v i a n c e T a b l e 56 57 Model Log(likelihood) # Param's Deviance Test d.f. P-value 58 Full model -77.6769 5 59 F i t t e d m o d e l -77.9988 3 0.643944 2 0.7247 60 R e d u c e d m o d e l -83.2067 1 11.0597 4 0.0259 61 62 AIC: 161.998 63 64 65 G o o d n e s s o f F i t 66 S c a l e d 67 Dose Est. Prob. Expected Observed Size Residual 68 69 0.0000 0.5978 25.107 25.000 42 -0.034 70 30.0000 0.7564 21.936 23.000 29 0.460 This document is a draftfor review purposes only and does not constitute Agency policy. E-413 DRAFT--DO NOT CITE OR QUOTE 1 100.0000 0.8045 20.112 19.000 25 -0.561 2 300.0000 0.8480 20.351 20.000 24 - 0 . 2 0 0 3 1000.0000 0.8905 28.495 29.000 32 0.286 4 5 Chi^2 = 0.65 d.f. = 2 P-value = 0.7227 6 7 8 Benchmark Dose Computation 9 10 S p e c i f i e d e f f e c t : 0.1 11 12 R i s k T y p e : Extra risk 13 14 C o n f i d e n c e l e v e l : 0.95 15 16 BMD ^ 0.371315 17 18 B e n c h m a r k d o s e c o m p u t a t i o n f a i l e d . L o w e r l i m i t i n c l u d e s zer o . 19 20 21 E .3.28.5. Figurefor Additional Model Presented: Log-Logistic, Unrestricted Log-Logistic Model ..................................... -- 1-- 1-- 1 i 1 1 1-- 1-- 1-- 1 1 1 1 i 1-- 1-- 1-- .....................................-- 1-- 1-- .......................-- 1-- 1-- .........................-- 1-- 1-- 1-- 1-- r 1 Log-Logistic --------------- 0.9 0.8 T3 CD O o1-I-- (0 LL 0.6 0.5 22 23 0.4 .......................................................... ........ i i ........ ............................................... i i ........ ......................................................... i i ........ .......................................................... ........ i i ........ ............................................... i i ........ ......................................................... i i i ____ i____ L 0 200 400 600 800 1000 18:17 02/16 2010 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-414 DRAFT--DO NOT CITE OR QUOTE 1 E .3.29. M urray et al., 1979: F ertility in F2 G eneration 2 E .3 .2 9 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes gamma 0 N /A 61.729 7.016E+00 1.698E+00 logistic negative intercept (intercept = 1 0.072 60.497 4.007E+00 2.836E+00 2.53) log-logistic 0 N /A 61.729 7.902E+00 1.584E+00 m ultistage, 1degree multistage, 2degree a probit 1 0.053 61.644 2.380E+00 1.320E+00 1 0.094 59.935 4.548E+00 1.635E+00 negative intercept (intercept = 1 0.070 60.613 3.707E+00 2.615E+00 1.446) W eibull 0 N /A 61.729 8.115E+00 1.698E+00 log-probit, unrestricted 0 N /A 61.729 6.373E +00 1.503E+00 unrestricted (slope = 2.306) a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3.29.2. Outputfor Selected Model: Multistage, 2-Degree 6 Murray et al., 1979: Fertility in F2 Generation 7 8 9 10 M u l t i s t a g e M o d e l . ( V e r s i o n : 3.0; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ M u r r a y _ 1 9 7 9 _ f e r t _ i n d e x _ f 2 _ M u l t i 2 _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ M u r r a y _ 1 9 7 9 _ f e r t _ i n d e x _ f 2 _ M u l t i 2 _ 1 . p l t 13 ~ ~ T u e F e b 16 2 0 : 0 8 : 0 6 2 0 1 0 14 15 16 T a b l e 1 b u t e x p r e s s e d as n u m b e r o f d a m s w h o d o n o t p r o d u c e o f f s p r i n g 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) * [ 1 - E X P ( 22 - b e t a i * d o s e ^ 1 - b e t a 2 * d o s e ^ 2 ) ] 23 24 T h e p a r a m e t e r b e t a s a r e r e s t r i c t e d t o b e p o s i t i v e 25 26 27 D e p e n d e n t v a r i a b l e = D i c h E f f 28 I n d e p e n d e n t v a r i a b l e = D o s e 29 30 T o t a l n u m b e r o f o b s e r v a t i o n s = 3 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 T o t a l n u m b e r o f p a r a m e t e r s i n m o d e l = 3 33 T o t a l n u m b e r o f s p e c i f i e d p a r a m e t e r s = 0 34 D e g r e e o f p o l y n o m i a l = 2 35 This document is a draftfor review purposes only and does not constitute Agency policy. E-415 DRAFT--DO NOT CITE OR QUOTE 1 2 Maximum number of iterations = 250 3 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 4 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 5 6 7 8 Default Initial Parameter Values 9 Background = 0.0624181 10 Beta(1) = 0 11 Beta(2) = 0.00532688 12 13 14 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 15 16 ( * * * T h e m o d e l p a r a m e t e r ( s ) - B e t a ( 1 ) 17 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 18 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 19 20 Background Beta(2) 21 22 B a c k g r o u n d 1 -0.44 23 24 B e t a ( 2 ) -0.44 1 25 26 27 28 P a r a m e t e r E s t i m a t e s 29 30 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 31 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 32 B a c k g r o u n d 0 . 0 7 7 2 2 0 1 33 B e t a ( 1 ) 0 34 Beta(2) 0.00509404 35 36 I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 37 38 39 40 A n a l y s i s o f D e v i a n c e T a b l e 41 42 Model L o g ( l i k e l i h o o d ) # P a r a m 's D e v i a n c e T e s t d . f . P-value 43 Full model -25.8194 3 44 F i t t e d m o d e l -27.9673 2 4.29584 0.03821 45 R e d u c e d m o d e l -34.0009 1 16.363 0.0002798 46 47 AIC: 59.9347 48 49 50 G o o d n e s s o f F i t 51 S c a l e d 52 Dose Est. Prob. Expected Observed Size Residual 53 54 0.0000 0.0772 2.471 4.000 32 1.013 55 1.0000 0.0819 1.638 0.000 20 -1.336 56 10.0000 0.4455 8.911 9.000 20 0.040 57 58 C h i ^ 2 = 2 . 8 1 d.f. = 1 P-value = 0.0936 59 60 61 B e n c h m a r k D o s e C o m p u t a t i o n 62 63 S p e c i f i e d e f f e c t = 0.1 64 65 R i s k T y p e = Extra risk 66 67 C o n f i d e n c e l e v e l = 0.95 68 69 BMD = 4.54787 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-416 DRAFT--DO NOT CITE OR QUOTE 1 BMDL 1.63487 2 3 BMDU = 6.79105 4 5 T a k e n t ogether, (1.63487, 6.79105) is a 90 6 interval for the BMD 7 8 % two-sided confidence 9 E .3.29.3. Figurefor Selected Model: Multistage, 2-Degree Multistage Model with 0.95 Confidence Level Fraction Affected 10 11 This document is a draftfor review purposes only and does not constitute Agency policy. E-417 DRAFT--DO NOT CITE OR QUOTE 1 E.3.30. N ational T oxicology Program , 1982: T oxic H epatitis, M ale M ice 2 E .3 .3 0 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2pValue AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes gamma 1 0.026 113.097 1.552E+01 5.155E+00 logistic negative intercept (intercept = 2 0.093 110.712 1.769E+01 1.383E+01 -3.087) log-logistic 1 0.027 113.093 1.499E+01 6.628E+00 log-probit multistage, 3degree a probit W eibull 1 0.027 113.111 1.360E+01 7.237E+00 1 0.028 112.555 1.488E+01 4.676E+00 negative intercept (intercept = 2 0.088 110.696 1.564E+01 1.261E+01 -1.731) 1 0.026 113.056 1.619E+01 4.903E+00 aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3.30.2. Outputfor Selected Model: Multistage, 3-Degree 6 National T oxicology Program, 1982: T oxic Hepatitis, M ale M ice 7 8 9 10 M u l t i s t a g e M o d e l . . ( V e r s i o n : 3.0; D a t e : 0 5 / 1 6 / 2 0 0 8 :) 11 I n p u t D a t a Fil e : C : \ 1 \ 3 7 N T P 1 9 8 2 T o x H e p M u l t i 3 1. (d) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 3 7 N T P 1 9 8 2 T o x H e p M u l t i 3 1 . p l t 13 T u e F e b 16 1 8 : 1 7 : 14 15 16 0 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) * [ 1 - E X P ( 22 - b e t a i * d o s e ^ 1 - b e t a 2 * d o s e ^ 2 - b e t a 3 * d o s e ^ 3 ) ] 23 24 T h e p a r a m e t e r b e t a s a r e r e s t r i c t e d t o b e p o s i t i v e 25 26 27 D e p e n d e n t v a r i a b l e = D i c h E f f 28 I n d e p e n d e n t v a r i a b l e = D o s e 29 30 T o t a l n u m b e r o f o b s e r v a t i o n s = 4 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 T o t a l n u m b e r o f p a r a m e t e r s i n m o d e l = 4 33 T o t a l n u m b e r o f s p e c i f i e d p a r a m e t e r s = 0 34 D e g r e e o f p o l y n o m i a l = 3 35 36 37 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 38 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 This document is a draftfor review purposes only and does not constitute Agency policy. E-418 DRAFT--DO NOT CITE OR QUOTE 1 Parameter Convergence has been set to: 1e-008 2 3 4 5 Default Initial Parameter Values 6 Background = 0.0525767 7 Beta(1) = 0.00243254 8 Beta(2) = 0 9 Beta(3) = 5.29052e-006 10 11 12 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 13 14 ( * * * T h e m o d e l p a r a m e t e r ( s ) - B e t a ( 2 ) 15 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 16 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 17 18 Background Beta(1) Beta(3) 19 20 B a c k g r o u n d 1 -0.69 0.66 21 22 B e t a ( 1 ) -0.69 1 -0.98 23 24 B e t a ( 3 ) 0.66 -0.98 1 25 26 27 28 P a r a m e t e r E s t i m a t e s 29 30 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 31 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 32 B a c k g r o u n d 0 . 0 3 8 3 4 7 4 * 33 Beta(1) 0.00605732 * 34 B e t a ( 2 ) 0* 35 Beta(3) 4.60855e-006 * 36 37 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 38 39 40 41 A n a l y s i s o f D e v i a n c e T a b l e 42 43 Model L o g ( l i k e l i h o o d ) # Parami' s D e v i a n c e T e s t d . f . P-value 44 Full model -51.0633 4 45 F i t t e d m o d e l -53.2776 3 4.42854 1 0.03534 46 R e d u c e d m o d e l -121.743 1 141.358 3 <.0001 47 48 AIC: 112.555 49 50 51 G o o d n e s s o f F i t 52 53 Dose Est. Prob. Expected Observed Size Residual 54 55 0.0000 0.0383 2.799 1.000 73 -1.097 56 1.4000 0.0465 2.278 5.000 49 1.847 57 7.1000 0.0803 3.937 3.000 49 -0.492 58 71.0000 0.8798 43.990 44.000 50 0.004 59 60 C h i ^ 2 = 4 . 8 6 d.f. = 1 P-value = 0.0275 61 62 63 B e n c h m a r k D o s e C o m p u t a t i o n 64 65 S p e c i f i e d e f f e c t = 0.1 66 67 R i s k T y p e = Extra risk 68 69 C o n f i d e n c e l e v e l = 0.95 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-419 DRAFT--DO NOT CITE OR QUOTE 1 BMD 14.8848 2 3 BMDL = 4.67636 4 5 BMDU = 28.8293 6 7 T a k e n t ogether, (4.67636, 28.8293) is a 90 8 interval for the BMD 9 10 % two-sided confidence 11 E .3 .3 0 .3 . Figurefor Selected Model: Multistage, 3-Degree Multistage Model with 0.95 Confidence Level Fraction Affected 12 18:17 02/16 2010 13 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-420 DRAFT--DO NOT CITE OR QUOTE 1 E.3.31. N ational T oxicology Program , 2006: A lveolar M etaplasia 2 E .3 .3 1 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes gamma 4 < 0 .0 0 1 3 4 0 .1 2 7 2 .2 4 0 E + 0 0 1 .791E + 00 p o w er b oun d hit (p o w er = 1) logistic negative intercept (intercept = 4 <0.001 358.346 4.997E+00 4.149E+00 -0.687) log-logistic a 4 0.409 312.970 6.644E-01 5.041E-01 slope bound hit (slope = 1) log-probit 4 < 0 .0 0 1 3 4 0 .2 9 6 3 .2 9 1 E + 0 0 2 .5 1 7 E + 0 0 slop e b ound hit (slop e = 1) multistage, 5degree probit 4 <0.001 340.127 2.240E+00 1.791E+00 final B = 0 negative intercept (intercept = 4 <0.001 362.181 5.656E+00 4.810E+00 -0.381) W eibull 4 < 0 .0 0 1 3 4 0 .1 2 7 2 .2 4 0 E + 0 0 1 .791E + 00 p o w er b oun d hit (p o w er = 1) gamma, unrestricted log-logistic, unrestricted b log-probit, unrestricted W eibull, unrestricted 3 0.407 314.135 2.211E -02 8.081E-04 unrestricted (power = 0.297) 3 0.739 312.487 3.062E-01 7.972E -02 unrestricted (slope = 0.785) 3 0.727 312.543 3.316E-01 8.968E-02 unrestricted (slope = 0.471) 3 0.586 313.176 9.000E-02 1.341E-02 unrestricted (power = 0.465) aBest-fitting m odel, BM D S output presented in this appendix b Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .3 1 .2 . Outputfor Selected Model: Log-Logistic 6 National T oxicology Program, 2006: A lveolar M etaplasia 7 8 9 10 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 4 0 _ N T P _ 2 0 0 6 _ A l v M e t a _ L o g L o g i s t i c _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 4 0 _ N T P _ 2 0 0 6 _ A l v M e t a _ L o g L o g i s t i c _ 1 . p l t 13 ~ T u e F e b 16 1 8 : 1 9 : 3 0 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e This document is a draftfor review purposes only and does not constitute Agency policy. E-421 D RAFT-- DO N O T CITE OR QUOTE 1 Slope pa r a m e t e r is r estricted as slope >= 1 2 3 Total number of observations = 6 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 12 13 14 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 15 background = 0.0377358 16 intercept = -2.03745 17 slope = 1 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( * * * T h e m o d e l p a r a m e t e r ( s ) - s l o p e 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 24 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 25 26 background intercept 27 28 b a c k g r o u n d 1 -0.4 29 30 i n t e r c e p t -0.4 1 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 37 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 38 b a c k g r o u n d 0 . 0 4 4 8 7 5 3 39 intercept -1.78837 40 s l o p e 1 41 42 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 43 44 45 46 A n a l y s i s o f D e v i a n c e T a b l e 47 48 Model Log(likelihood) # Param's Deviance Test d.f. P-value 49 Full model -152. 615 6 50 F i t t e d m o d e l -154. 485 2 3.7393 4 0.4424 51 R e d u c e d m o d e l -216. 802 1 128.374 5 <.0001 52 53 AIC: 312.97 54 55 56 G o o d n e s s o f F i t 57 S c a l e d 58 Dose Est. Prob. Expected Observed Size Residual 59 60 0.0000 0.0449 2.378 2.000 53 -0. 251 61 2.1400 0.2966 16.017 19.000 54 0. 8 8 9 62 7.1400 0.5647 29.928 33.000 53 0. 8 51 63 15.7000 0.7366 38.301 35.000 52 -1. 039 64 32.9000 0.8531 45.214 45.000 53 -0. 083 65 71.4000 0.9262 48.162 46.000 52 -1. 147 66 67 C h i ^ 2 = 3 . 9 8 d.f. = 4 P-value = 0.4088 68 69 70 B e n c h m a r k D o s e C o m p u t a t i o n This document is a draftfor review purposes only and does not constitute Agency policy. E-422 DRAFT--DO NOT CITE OR QUOTE 1 2 Specified effect 3 4 Risk Type 5 6 Confidence level 7 8 BMD 9 10 B M D L 11 12 0.1 Extra risk 0.95 0.664411 0.504109 13 E .3 .3 1 .3 . Figurefor Selected Model: Log-Logistic Log-Logistic Model with 0.95 Confidence Level 14 15 16 17 E .3.31.4. Outputfor Additional Model Presented: Log-Logistic, Unrestricted 18 N ation al T o x ic o lo g y Program , 2006: A lv eo la r M etaplasia 19 20 21 22 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 23 I n p u t D a t a F i l e : C : \ 1 \ 4 0 _ N T P _ 2 0 0 6 _ A l v M e t a _ L o g L o g i s t i c _ U _ 1 . ( d ) 24 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 4 0 _ N T P _ 2 0 0 6 _ A l v M e t a _ L o g L o g i s t i c _ U _ 1 . p l t 25 ~ T u e F e b 16 1 8 : 1 9 : 3 1 ~ 2 0 1 0 26 27 28 0 This document is a draftfor review purposes only and does not constitute Agency policy. E-423 DRAFT--DO NOT CITE OR QUOTE 1 2 3 The form of the probability function is: 4 5 P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))] 6 7 8 Dependent variable = DichEff 9 Independent variable = Dose 10 S l o p e p a r a m e t e r is n o t r e s t r i c t e d 11 12 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 13 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 14 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 15 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 16 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 17 18 19 20 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 21 22 23 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 24 background = 0.0377358 25 intercept = -1.26694 26 slope = 0.784484 27 28 29 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 30 31 background intercept slope 32 33 b a c k g r o u n d 1 -0.24 0.11 34 35 i n t e r c e p t -0.24 1 -0.9 36 37 s l o p e 0.11 -0.9 1 38 39 40 41 P a r a m e t e r E s t i m a t e s 42 43 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 44 Variable Estimate Std. Err. Lower Conf. Limit Uppe 45 b a c k g r o u n d 0 . 0 3 7 5 2 8 6 * * 46 intercept -1.26811 * * 47 slope 0.785033 48 49 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 50 51 52 53 A n a l y s i s o f D e v i a n c e T a b l e 54 55 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 56 Full model -152.615 6 57 F i t t e d m o d e l -153.244 3 1.2566 3 0.7395 58 R e d u c e d m o d e l -216.802 1 128.374 5 <.0001 59 60 AIC: 312.487 61 62 63 G o o d n e s s o f F i t 64 65 Dose Est. Prob. Expected Observed Size Residual 66 67 0.0000 0.0375 1.989 2.000 53 0.008 68 2.1400 0.3631 19.609 19.000 54 - 0 . 1 7 2 69 7.1400 0.5845 30.980 33.000 53 0.563 70 15.7000 0.7205 37.468 35.000 52 - 0 .763 This document is a draftfor review purposes only and does not constitute Agency policy. E-424 DRAFT--DO NOT CITE OR QUOTE 1 32.9000 0.8207 43.498 45.000 2 71.4000 0.8934 46.455 46.000 3 4 Chi^2 = 1.26 d.f. = 3 P-value = 0.7388 5 6 7 Benchmark Dose Computation 8 9 Specified effect : 0.1 10 11 R i s k T y p e : Extra risk 12 13 C o n f i d e n c e l e v e l : 0.95 14 15 BMD ^ 0.306194 16 17 BMDL ^ 0.0797223 18 19 53 52 0.538 -0.204 20 E .3.31.5. Figurefor Additional Model Presented: Log-Logistic, Unrestricted Log-Logistic Model with 0.95 Confidence Level 21 18:19 02/16 2010 22 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-425 DRAFT--DO NOT CITE OR QUOTE 1 E.3.32. N ational T oxicology Program , 2006: E osinophilic Focus, L iver 2 E .3 .3 2 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes gamma 4 0 .3 6 7 3 3 0 .4 5 7 5 .6 7 6 E + 0 0 4 .5 3 2 E + 0 0 p o w er b oun d hit (p ow er = 1) logistic negative intercept (intercept = 4 0.167 333.343 1.258E+01 1.071E+01 -1.747) log-logistic 3 0.117 334.148 4.727E+00 2.867E+00 log-probit 4 0.084 334.683 1.078E+01 8.514E+00 multistage, 5degree probit a 3 0.313 331.771 6.568E+00 4.666E+00 4 0.187 332.962 1.196E+01 1.031E+01 negative intercept (intercept = -1.061) W eibull 4 0 .3 6 7 3 3 0 .4 5 7 5 .6 7 5 E + 0 0 4 .5 3 2 E + 0 0 p o w er b oun d hit (p ow er = 1) log-probit, unrestricted 3 0.087 334.849 4.750E+00 1.757E+00 unrestricted (slope = 0.643) aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3 .3 2 .2 . Outputfor Selected Model: Probit 6 National T oxicology Program, 2006: E osinophilic Focus, Liver 7 8 9 10 P r o b i t M o d e l . ( V e r s i o n : 3.1; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 4 5 _ N T P _ 2 0 0 6 _ L i v E o s F o c _ P r o b i t _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 4 5 _ N T P _ 2 0 0 6 _ L i v E o s F o c _ P r o b i t _ 1 . p l t 13 T u e F e b 16 1 8 : 2 5 : 5 6 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = C u m N o r m ( I n t e r c e p t + S l o p e * D o s e ) , 22 23 w h e r e C u m N o r m ( . ) is t h e c u m u l a t i v e n o r m a l d i s t r i b u t i o n f u n c t i o n 24 25 26 D e p e n d e n t v a r i a b l e = D i c h E f f 27 I n d e p e n d e n t v a r i a b l e = D o s e 28 S l o p e p a r a m e t e r is n o t r e s t r i c t e d 29 30 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 33 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 34 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 35 This document is a draftfor review purposes only and does not constitute Agency policy. E-426 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Default Initial (and Specified) Parameter Values 4 background = 0 Specified 5 intercept = -1.11935 6 slope = 0.0279665 7 8 9 Asymptotic Correlation Matrix of Parameter Estimates 10 11 ( * * * T h e m o d e l p a r a m e t e r ( s ) - b a c k g r o u n d 12 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 13 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 14 15 intercept slope 16 17 i n t e r c e p t 1 -0.69 18 19 slope -0.69 1 20 21 22 23 P a r a m e t e r E s t i m a t e s 24 25 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 26 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 27 intercept -1.06148 0.109177 -1.27546 -0.847497 28 slope 0.0269279 0.00327788 0.0205034 0.0333525 29 30 31 32 A n a l y s i s o f D e v i a n c e T a b l e 33 34 Model Log(likelihood) # Param's Deviance Test d.f. P-value 35 F u l l m o d e l -161.07 6 36 F i t t e d m o d e l -164.481 2 6.8221 4 0.1456 37 R e d u c e d m o d e l -202.816 1 83.4925 5 <.0001 38 39 AIC: 332.962 40 41 42 G o o d n e s s o f F i t 43 S c a l e d 44 Dose Est . Prob. Expected Observed Size Residual 45 46 0.0000 0. 1 4 4 2 7.645 3. 0 0 0 53 -1. 816 47 2.1400 0. 1 5 7 7 8.517 8. 0 0 0 54 -0. 193 48 7.1400 0. 1 9 2 4 10.195 14. 000 53 1. 3 2 6 49 15.7000 0. 2 6 1 5 13.860 17. 000 53 0. 982 50 32.9000 0. 4 3 0 3 22.807 22. 000 53 -0. 224 51 71.4000 0. 8 0 5 4 42.688 42. 000 53 -0. 239 52 53 C h i ^ 2 = 6 . 1 6 d.f. = 4 P-value = 0.1873 54 55 56 B e n c h m a r k D o s e C o m p u t a t i o n 57 58 S p e c i f i e d e f f e c t 0.1 59 60 R i s k T y p e Extra risk 61 62 C o n f i d e n c e l e v e l 0.95 63 64 BMD 11.9584 65 66 BMDL 10.3075 67 This document is a draftfor review purposes only and does not constitute Agency policy. E-427 DRAFT--DO NOT CITE OR QUOTE 1 E .3 .3 2 .3 . Figurefor Selected Model: Probit Probit Model with 0.95 Confidence Level Fraction Affected 2 18:25 02/16 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-428 DRAFT--DO NOT CITE OR QUOTE 1 E.3.33. N ational T oxicology Program , 2006: Fatty C hange D iffuse, L iver 2 E .3 .3 3 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes gamma 4 0.668 252.294 4.224E+00 3.166E+00 logistic negative intercept (intercept = 4 0.005 269.825 1.092E+01 9.292E+00 -2.298) log-logistic 4 0.292 255.082 4.697E+00 3.153E+00 log-probit 4 0 .1 1 8 2 5 7 .5 4 8 6 .2 3 6 E + 0 0 5 .2 0 4 E + 0 0 slop e b ound hit (slop e = 1) multistage, 5degree probit 4 0.808 251.545 4.021E+00 3.250E+00 negative intercept (intercept = 4 0.005 269.430 1.052E+01 9.068E+00 -1.36) Weibull a 4 0.679 252.218 4.252E+00 3.174E+00 log-probit, unrestricted 4 0.282 255.258 4.581E+00 3.193E+00 unrestricted (slope = 0.824) aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3 .3 3 .2 . Outputfor Selected Model: Weibull 6 National T oxicology Program, 2006: Fatty Change D iffuse, Liver 7 8 9 10 W e i b u l l M o d e l u s i n g W e i b u l l M o d e l ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 4 7 _ N T P _ 2 0 0 6 _ L i v F a t D i f f _ W e i b u l l _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 4 7 _ N T P _ 2 0 0 6 _ L i v F a t D i f f _ W e i b u l l _ 1 . p l t 13 T u e F e b 16 1 8 : 2 6 : 5 7 2 0 1 0 14 15 16 N T P l i v e r f a t t y c h a n g e d i f f u s e 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) * [ 1 - E X P ( - s l o p e * d o s e ^ p o w e r ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 P o w e r p a r a m e t e r is r e s t r i c t e d as p o w e r > = 1 27 28 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 29 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 30 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 31 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 32 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 33 34 35 This document is a draftfor review purposes only and does not constitute Agency policy. E-429 DRAFT--DO NOT CITE OR QUOTE 1 Default Initial (and Specified) Parameter Values 2 Background = 0.00925926 3 Slope = 0.00962604 4 Power = 1.28042 5 6 7 Asymptotic Correlation Matrix of Parameter Estimates 8 9 ( *** The model parameter(s) -Background 10 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 11 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 12 13 Slope Power 14 15 S l o p e 1 -0.97 16 17 Power -0.97 1 18 19 20 21 P a r a m e t e r E s t i m a t e s 22 23 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 24 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 25 B a c k g r o u n d 0 NA 26 Slope 0.0223474 0.00951041 0.0037073 0.0409874 27 Power 1.07133 0.122134 0.831952 1.31071 28 29 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 30 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 31 h a s n o s t a n d a r d e r r o r . 32 33 34 35 A n a l y s i s o f D e v i a n c e T a b l e 36 37 Model Log(likelihood) # Param's Deviance Test d.f. P-value 38 Full model -122.992 6 39 F i t t e d m o d e l -124.109 2 2.23388 4 0.6928 40 R e d u c e d m o d e l -204.846 1 163.708 5 <.0001 41 42 AIC: 252.218 43 44 45 G o o d n e s s o f F i t 46 S c a l e d 47 Dose Est. Prob. Expected Observed Size Residu 48 49 0 .0000 0. 0 0 0 0 0. 0 0 0 0. 0 0 0 53 0. 0 0 0 50 2 .1400 0. 04 92 2. 6 5 9 2. 0 0 0 54 -0 414 51 7 .1400 0. 1 6 7 7 8. 8 8 9 12. 000 53 1. 1 44 52 15 .7000 0. 3 4 7 5 18 420 17. 000 53 -0 409 53 32 .9000 0. 6 1 0 7 32 365 30. 000 53 -0 666 54 71 .4000 0. 8 8 5 1 46. 909 48. 000 53 0. 4 7 0 55 56 i^ 2 = 2.31 d.f. = 4 P-value = 0.6785 57 58 59 B e n c h m a r k D o s e C o m p u t a t i o n 60 61 S p e c i f i e d e f f e c t 0.1 62 63 R i s k T y p e = Extra risk 64 65 C o n f i d e n c e l e v e l 0.95 66 67 BMD = 4.25219 68 69 BMDL = 3.17375 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-430 DRAFT--DO NOT CITE OR QUOTE 1 2 E .3 .3 3 .3 . Figurefor Selected Model: Weibull Weibull Model with 0.95 Confidence Level Fraction Affected 3 18:26 02/16 2010 4 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-431 DRAFT--DO NOT CITE OR QUOTE 1 E.3.34. N ational T oxicology Program , 2006: G ingival H yperplasia, Squam ous, 2 Y ears 2 E .3 .3 4 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes gamma 4 0 .0 1 2 3 1 8 .8 6 7 2 .2 9 5 E + 0 1 1.417E +01 p o w er b oun d hit (p ow er = 1) logistic negative intercept (intercept = 4 0.008 320.908 3.594E+01 2.564E+01 -1.711) log-logistic a 4 0.015 317.969 1.838E+01 1.044E+01 slope bound hit (slope = 1) log-probit 4 0.003 32 3 .6 3 3 4 .3 1 3 E + 0 1 2 .7 9 4 E + 0 1 slop e b ound hit (slop e = 1) multistage, 5degree probit 4 0.012 318.867 2.295E+01 1.417E+01 final B = 0 negative intercept (intercept = 4 0.008 320.687 3.436E+01 2.425E+01 -1.034) W eibull 4 0 .0 1 2 3 1 8 .8 6 7 2 .2 9 5 E + 0 1 1.417E +01 p o w er b oun d hit (p ow er = 1) gamma, unrestricted log-logistic, unrestricted b log-probit, unrestricted W eibull, unrestricted 3 0.651 307.529 2.480E-01 5.096E -09 unrestricted (pow er = 0.199) 3 0.675 307.416 3.710E-01 1.505E-07 unrestricted (slope = 0.265) 3 0.688 307.354 4.688E-01 8.851E-07 unrestricted (slope = 0.156) 3 0.663 307.471 3.076E-01 3.210E-08 unrestricted (power = 0.23) aBest-fitting m odel, BM D S output presented in this appendix b Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .3 4 .2 . Outputfor Selected Model: Log-Logistic 6 National T oxicology Program, 2006: Gingival Hyperplasia, Squamous, 2 Years 7 8 9 10 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 4 2 _ N T P _ 2 0 0 6 _ G i n g H y p S q _ L o g L o g i s t i c _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 4 2 _ N T P _ 2 0 0 6 _ G i n g H y p S q _ L o g L o g i s t i c _ 1 . p l t 13 ~ T u e F e b _ 16 1 8 : 2 0 : 2 9 ~ 2 0 1 0 14 15 16 [ i n s e r t s t u d y n o t e s ] 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e This document is a draftfor review purposes only and does not constitute Agency policy. E -432 D RAFT-- DO N O T CITE OR QUOTE 1 Slope pa r a m e t e r is r estricted as slope >= 1 2 3 Total number of observations = 6 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 12 13 14 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 15 background = 0.0188679 16 intercept = -4.5509 17 slope = 1 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( * * * T h e m o d e l p a r a m e t e r ( s ) - s l o p e 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 24 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 25 26 background intercept 27 28 b a c k g r o u n d 1 -0.71 29 30 i n t e r c e p t -0.71 1 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 37 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 38 b a c k g r o u n d 0.117717 39 intercept -5.10866 40 s l o p e 1 41 42 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 43 44 45 46 A n a l y s i s o f D e v i a n c e T a b l e 47 48 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 49 F u l l m o d e l -149.95 6 50 F i t t e d m o d e l -156.985 2 14.0696 4 0.007076 51 R e d u c e d m o d e l -162.631 1 25.3627 5 0.0001186 52 53 AIC: 317.969 54 55 56 G o o d n e s s o f F i t 57 S c a l e d 58 Dose Est. Prob. Expected Observed Size Residual 59 60 0.0000 0.1177 6.239 1.000 53 -2.233 61 2.1400 0.1290 6.965 7.000 54 0.014 62 7.1400 0.1542 8.174 14.000 53 2.216 63 15.7000 0.1942 10.292 13.000 53 0.940 64 32.9000 0.2641 13.995 15.000 53 0.313 65 71.4000 0.3837 20.335 16.000 53 -1.225 66 67 C h i ^ 2 = 12 38 d.f. = 4 P -value = 0.0147 68 69 70 B e n c h m a r k D o s e C o m p u t a t i o n This document is a draftfor review purposes only and does not constitute Agency policy. E-433 DRAFT--DO NOT CITE OR QUOTE 1 2 Specified effect 3 4 Risk Type 5 6 Confidence level 7 8 BMD 9 10 B M D L 11 12 0.1 Extra risk 0.95 18.3832 10.4359 13 E .3 .3 4 .3 . Figurefor Selected Model: Log-Logistic Log-Logistic Model with 0.95 Confidence Level 14 15 16 17 E .3.34.4. Outputfor Additional Model Presented: Log-Logistic, Unrestricted 18 N ation al T o x ic o lo g y Program , 2006: G ingival H yperplasia, Squam ous, 2 Y ears 19 20 21 22 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 23 I n p u t D a t a F i l e : C : \ 1 \ 4 2 _ N T P _ 2 0 0 6 _ G i n g H y p S q _ L o g L o g i s t i c _ U _ 1 . ( d ) 24 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 4 2 _ N T P _ 2 0 0 6 _ G i n g H y p S q _ L o g L o g i s t i c _ U _ 1 . p l t 25 ~ T u e F e b _ 16 1 8 : 2 0 : 2 9 ~ 2 0 1 0 26 27 28 [ i n s e r t s t u d y n o t e s ] This document is a draftfor review purposes only and does not constitute Agency policy. E-434 DRAFT--DO NOT CITE OR QUOTE 1 2 3 The form of the probability function is: 4 5 P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))] 6 7 8 Dependent variable = DichEff 9 Independent variable = Dose 10 S l o p e p a r a m e t e r is n o t r e s t r i c t e d 11 12 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 13 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 14 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 15 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 16 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 17 18 19 20 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 21 22 23 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 24 background = 0.0188679 25 intercept = -2.04571 26 slope = 0.299277 27 28 29 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 30 31 background intercept slope 32 33 b a c k g r o u n d 1 -0.3 0.12 34 <--1 o 35 i n t e r c e p t -0.3 1 36 37 s l o p e 0.12 -0.91 1 38 39 40 41 P a r a m e t e r E s t i m a t e s 42 43 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 44 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf 45 b a c k g r o u n d 0 . 0 1 8 5 1 2 6 * * * 46 intercept -1.93464 * * * 47 slope 0.264795 48 49 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 50 51 52 53 A n a l y s i s o f D e v i a n c e T a b l e 54 55 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 56 F u l l m o d e l -149.95 6 57 F i t t e d m o d e l -150.708 3 1.5163 3 0.6785 58 R e d u c e d m o d e l -162.631 1 25.3627 5 0.0001186 59 60 AIC: 307.416 61 62 63 G o o d n e s s o f F i t 64 65 Dose Est. Prob. Expected Observed Size Residual 66 67 0.0000 0.0185 0.981 1.000 53 0.019 68 2.1400 0.1659 8.959 7.000 54 -0. 7 1 7 69 7.1400 0.2105 11.155 14.000 53 0.959 70 15.7000 0.2447 12.972 13.000 53 0.009 This document is a draftfor review purposes only and does not constitute Agency policy. E-435 DRAFT--DO NOT CITE OR QUOTE 1 32.9000 0.2806 14.873 15.000 2 71.4000 0.3219 17.059 16.000 3 4 Chi^2 = 1.53 d.f. = 3 P-value = 0.6750 5 6 7 Benchmark Dose Computation 8 9 Specified effect : 0.1 10 11 R i s k T y p e : Extra risk 12 13 C o n f i d e n c e l e v e l : 0.95 14 15 BMD ^ 0.370958 16 17 BMDL ^ 1.50494e-007 18 19 53 53 0.039 -0.311 20 E .3.34.5. Figurefor Additional Model Presented: Log-Logistic, Unrestricted Log-Logistic Model with 0.95 Confidence Level 21 18:20 02/16 2010 22 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-436 DRAFT--DO NOT CITE OR QUOTE 1 E.3.35. N ational T oxicology Program , 2006: H epatocyte H ypertrophy, 2 Y ears 2 E .3 .3 5 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2pValue AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes gamma 4 < 0.001 2 9 0 .3 6 5 1 .647E + 00 1 .340E + 00 p o w er b oun d hit (p ow er = 1) logistic negative intercept (intercept = 4 <0.001 310.492 4.315E+00 3.650E+00 -1.237) log-logistic 5 0 .0 1 0 2 7 8 .0 8 2 6.9 7 8 E -0 1 5.4 5 4 E -0 1 slop e b ound hit (slop e = 1) log-probit 4 < 0.001 2 9 7 .1 6 8 2 .9 3 0 E + 0 0 2 .2 6 7 E + 0 0 slop e b ound hit (slop e = 1) multistage, 5degree a probit 4 <0.001 290.365 1.647E+00 1.340E+00 final B = 0 negative intercept (intercept = 4 <0.001 313.841 4.564E+00 3.923E+00 -0.714) W eibull 4 < 0.001 2 9 0 .3 6 5 1 .647E + 00 1 .340E + 00 p o w er b oun d hit (p ow er = 1) gamma, unrestricted log-logistic, unrestricted log-probit, unrestricted W eibull, unrestricted 4 0.029 275.042 error error unrestricted (pow er = 0.478) 4 0.005 280.068 6.672E-01 2.939E-01 unrestricted (slope = 0.984) 4 0.006 279.204 7.167E-01 3.322E-01 unrestricted (slope = 0.594) 4 0.019 275.967 3.709E-01 1.315E-01 unrestricted (power = 0.64) aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3.35.2. Outputfor Selected Model: Multistage, 5-Degree 6 National T oxicology Program, 2006: Hepatocyte Hypertrophy, 2 Years 7 8 9 10 M u l t i s t a g e M o d e l . ( V e r s i o n : 3.0; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 4 3 _ N T P _ 2 0 0 6 _ H e p H y p e r _ M u l t i 5 _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 4 3 _ N T P _ 2 0 0 6 _ H e p H y p e r _ M u l t i 5 _ 1 . p l t 13 ~ T u e F e b 16 1 8 7 2 1 : 0 0 2 0 1 0 14 15 16 [ i n s e r t s t u d y n o t e s ] 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) * [ 1 - E X P ( 22 - b e t a 1 * d o s e ^ 1 - b e t a 2 * d o s e ^ 2 - b e t a 3 * d o s e ^ 3 - b e t a 4 * d o s e ^ 4 - b e t a 5 * d o s e ^ 5 ) ] 23 24 T h e p a r a m e t e r b e t a s a r e r e s t r i c t e d t o b e p o s i t i v e 25 26 27 D e p e n d e n t v a r i a b l e = D i c h E f f This document is a draftfor review purposes only and does not constitute Agency policy. E -437 D RAFT-- DO N O T CITE OR QUOTE 1 Independent variable = Dose 2 3 Total number of observations = 6 4 Total number of records with missing values = 0 5 Total number of parameters in model = 6 6 Total number of specified parameters = 0 7 Degree of polynomial = 5 8 9 10 M a x i m u m n u m b e r o f = 250 11 R e l a t i v e F u n c t i o n has b e e n set to: 1e-008 12 P a r a m e t e r has b e e n set to: 1e-008 13 14 15 16 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 17 Background 0.232262 18 Beta(1) 0.045074 19 Beta(2) 0 20 Beta(3) 0 21 Beta(4) 0 22 Beta(5) 2.59945e-010 23 24 25 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 26 27 ( *** The model parameter(s) -Beta(2) -Beta(3) -Beta(4) -Beta(5) 28 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 29 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 30 31 Background Beta(1) 32 33 B a c k g r o u n d 1 -0.64 34 35 B e t a ( 1 ) -0.64 1 36 37 38 39 P a r a m e t e r E s t i m a t e s 40 41 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 42 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 43 Background 0.0541647 * * * 44 Beta(1) 0.0639585 * * * 45 B e t a ( 2 ) 0* * * 46 B e t a ( 3 ) 0* * * 47 B e t a ( 4 ) 0* * * 48 B e t a ( 5 ) 0* * * 49 50 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 51 52 53 54 A n a l y s i s o f D e v i a n c e T a b l e 55 56 Model Log(likelihood) # Param's Deviance Test d.f. P-value 57 Full model -129.986 6 58 F i t t e d m o d e l -143.183 2 26.3932 4 2.6361629e-005 59 R e d u c e d m o d e l -219.97 1 179.968 5 <.0001 60 61 AIC: 290.365 62 63 64 G o o d n e s s o f F i t 65 S c a l e d 66 Dose Est. Prob. Expected Observed Size Residual 67 68 0.0000 0.0542 2.871 0.000 53 -1.742 69 2.1400 0.1752 9.458 19.000 54 3 . 4 1 6 70 7.1400 0.4009 21.248 19.000 53 -0.630 This document is a draftfor review purposes only and does not constitute Agency policy. E-438 DRAFT--DO NOT CITE OR QUOTE 1 15.7000 0.6535 34.635 42.000 53 2.126 2 32.9000 0.8847 46.887 41.000 53 -2.532 3 71.4000 0.9902 52.479 52.000 53 -0.667 4 5 Chi^2 = 26.48 d.f. = 4 P-value = 0.0000 6 7 8 Benchmark Dose Computation 9 10 S p e c i f i e d e f f e c t 0.1 11 12 R i s k T y p e Extra risk 13 14 C o n f i d e n c e l e v e l 0.95 15 16 BMD 1.64733 17 18 BMDL 1.34007 19 20 BMDU 2.0581 21 22 T a k e n t o g e t h e r , ( 1 . 3 4 0 0 7 , 2 . 0 5 8 1 ) is a 90 two-sided confidence 23 i n t e r v a l f o r t h e B M D 24 25 26 E .3 .3 5 .3 . Figurefor Selected Model: Multistage, 5-Degree Multistage Model with 0.95 Confidence Level 27 18:21 02/16 2010 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-439 DRAFT--DO NOT CITE OR QUOTE 1 E.3.36. National Toxicology Program, 2006: Necrosis, Liver 2 E.3.36.1. S u m m a ry Table o f B M D S M odeling R esults Model logistic Degrees of Freedom x2p Value 4 0.397 AIC 238.314 BMD (ng/kg-d) 3 .4 8 4 E + 0 1 BMDL (ng/kg-d) 2 .8 4 2 E + 0 1 Notes negative intercept (intercept = -2.601) log-logistic 4 0 .8 1 0 2 3 5 .2 6 5 1.791E +01 1.194E +01 slop e b ound hit (slop e = 1) log-probit 4 0 .2 9 0 2 3 9 .1 0 7 3.2 0 5 E + 0 1 2 .3 8 2 E + 0 1 slop e b ound hit (slop e = 1) multistage, 5degree probit 4 0.763 235.581 2.019E+01 1.419E+01 final h = 0 negative intercept (intercept = 4 0.445 237.888 3.266E+01 2.637E+01 -1.508) W eibull 4 0.763 2 3 5 .5 8 1 2 .0 1 9 E + 0 1 1.419E +01 p o w er b oun d hit (p ow er = 1) gamma, unrestricted log-logistic, unrestricted log-probit, unrestricted a W eibull, unrestricted 3 0.869 236.344 1.114E+01 3.487E+00 unrestricted (power = 0.599) 3 0.833 236.483 1.112E+01 3.581E+00 unrestricted (slope = 0.695) 3 0.768 236.742 1.061E+01 3.498E+00 unrestricted (slope = 0.367) 3 0.856 236.393 1.117E+01 3.554E+00 unrestricted (power = 0.64) aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3.36.2. Outputfor Selected Model: Log-Probit, Unrestricted 6 National T oxicology Program, 2006: N ecrosis, Liver 7 8 9 10 P r o b i t M o d e l . ( V e r s i o n : 3.1; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 5 0 _ N T P _ 2 0 0 6 _ L i v N e c _ L o g P r o b i t _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 5 0 _ N T P _ 2 0 0 6 _ L i v N e c _ L o g P r o b i t _ U _ 1 . p l t 13 ~ T u e F e b 16 1 8 : 3 4 7 3 1 2 0 1 0 14 15 16 N T P l i v e r n e c r o s i s 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = B a c k g r o u n d 22 + ( 1 - B a c k g r o u n d ) * C u m N o r m ( I n t e r c e p t + S l o p e * L o g ( D o s e ) ) , 23 24 w h e r e C u m N o r m ( . ) is t h e c u m u l a t i v e n o r m a l d i s t r i b u t i o n f u n c t i o n 25 26 27 D e p e n d e n t v a r i a b l e = D i c h E f f 28 I n d e p e n d e n t v a r i a b l e = D o s e 29 S l o p e p a r a m e t e r is n o t r e s t r i c t e d 30 This document is a draftfor review purposes only and does not constitute Agency policy. E-440 DRAFT--DO NOT CITE OR QUOTE 1 Total number of observations = 6 2 Total number of records with missing values = 0 3 Maximum number of iterations = 250 4 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 5 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 6 7 8 9 User has chosen the log transformed model 10 11 12 D e f a u l t I n i t i a l (and S p e c i f i e d ) P a r a m e t e r V a l u e s 13 background = 0.0188679 14 intercept = -1.98094 15 slope = 0.316942 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 background intercept slope 21 22 b a c k g r o u n d 1 -0.69 0.59 23 24 i n t e r c e p t -0.69 1 -0.97 25 26 s l o p e 0.59 -0.97 1 27 28 29 30 P a r a m e t e r E s t i m a t e s 31 32 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 33 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 34 b a c k g r o u n d 0 . 0 2 2 8 3 3 9 0 . 0 2 3 0 8 1 8 -0.0224057 0.0680734 35 intercept -2.14844 0.527256 -3.18184 -1.11503 36 slope 0.367034 0.139055 0.0944904 0.639577 37 38 39 40 A n a l y s i s o f D e v i a n c e T a b l e 41 42 Model Log(likelihood) # Param's Deviance Test d.f. P-value 43 Full model -114.813 6 44 F i t t e d m o d e l -115.371 3 1.1157 3 0.7733 45 R e d u c e d m o d e l -127.98 1 26.3331 5 <.0001 46 47 AIC: 236.742 48 49 50 G o o d n e s s o f F i t 51 S c a l e d 52 Est. Prob. Expected Observed Residual 53 54 0.0000 0.0228 1.210 1. 0 0 0 53 -0.193 55 2.1400 0.0529 2.858 4. 0 0 0 54 0.694 56 7.1400 0.0979 5.187 4. 0 0 0 53 -0.549 57 15.7000 0.1475 7.819 8. 0 0 0 53 0.070 58 32.9000 0.2116 11.215 10. 000 53 -0.409 59 71.4000 0.2968 15.729 17. 000 53 0.382 60 61 C h i ^ 2 = 1 . 1 4 d.f. = 3 P-value = 0.7678 62 63 64 B e n c h m a r k D o s e C o m p u t a t i o n 65 66 S p e c i f i e d e f f e c t = 0.1 67 68 R i s k T y p e = Extra risk 69 70 C o n f i d e n c e l e v e l = 0.95 This document is a draftfor review purposes only and does not constitute Agency policy. E-441 DRAFT--DO NOT CITE OR QUOTE 1 2 BMD 10.6107 3 4 BMDL = 3.49791 5 6 7 E .3.36.3. Figurefor Selected Model: Log-Probit, Unrestricted LogProbit Model with 0.95 Confidence Level Fraction Affected 8 18:34 02/16 2010 9 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-442 DRAFT--DO NOT CITE OR QUOTE 1 E.3.37. National Toxicology Program, 2006: Oval Cell Hyperplasia 2 E.3.37.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes gamma 3 0.072 199.446 8.970E+00 5.499E+00 logistic negative intercept (intercept = 4 0.069 199.875 9.792E+00 8.245E+00 -3.116) log-logistic 3 0.039 202.012 9.708E+00 7.247E+00 log-probit multistage, 5degree probit a W eibull b 3 0.068 200.421 9.968E+00 7.758E+00 2 0.066 198.641 5.424E+00 3.514E+00 4 0.112 198.166 9.103E+00 7.701E+00 negative intercept (intercept = -1.821) 3 0.075 198.690 7.712E+00 4.692E+00 aBest-fitting m odel, BM D S output presented in this appendix b Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .3 7 .2 . Outputfor Selected Model: Probit 6 National T oxicology Program, 2006: Oval Cell Hyperplasia 7 8 9 10 P r o b i t M o d e l . ( V e r s i o n : 3.1; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 5 3 _ N T P _ 2 0 0 6 _ O v a l H y p e r _ P r o b i t _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 5 3 _ N T P _ 2 0 0 6 _ O v a l H y p e r _ P r o b i t _ 1 . p l t 13 ~ T u e F e b _ 16 1 9 : 5 1 : 5 2 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = C u m N o r m ( I n t e r c e p t + S l o p e * D o s e ) , 22 23 w h e r e C u m N o r m ( . ) is t h e c u m u l a t i v e n o r m a l d i s t r i b u t i o n f u n c t i o n 24 25 26 D e p e n d e n t v a r i a b l e = D i c h E f f 27 I n d e p e n d e n t v a r i a b l e = D o s e 28 S l o p e p a r a m e t e r is n o t r e s t r i c t e d 29 30 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 33 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 34 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 35 36 37 This document is a draftfor review purposes only and does not constitute Agency policy. E-443 DRAFT--DO NOT CITE OR QUOTE 1 Default Initial (and Specified) Parameter Values 2 background = 0 Specified 3 intercept = -1.92612 4 slope = 0.0670004 5 6 7 Asymptotic Correlation Matrix of Parameter Estimates 8 9 ( *** The model parameter(s) -background 10 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 11 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 12 13 intercept slope 14 15 i n t e r c e p t 1 -0.8 16 17 s l o p e -0.8 1 18 19 20 21 P a r a m e t e r E s t i m a t e s 22 23 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 24 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 25 intercept -1.82129 0.16954 -2.15359 -1.489 26 slope 0.0767832 0.00835175 0.060414 0.0931523 27 28 29 30 A n a l y s i s o f D e v i a n c e T a b l e 31 32 Model Log(likelihood) # Param's Deviance Test d.f. P-value 33 Full model -92.4898 6 34 F i t t e d m o d e l -97.0832 2 9.18683 4 0.0566 35 R e d u c e d m o d e l -210.191 1 235.402 5 <.0001 36 37 AIC: 198.166 38 39 40 G o o d n e s s o f F i t 41 S c a l e d 42 Dose Est. Prob. Expected Observed Size Residual 43 44 0.0000 0.0343 1.817 0. 0 0 0 53 -1. 372 45 2.1400 0.0488 2.633 4. 0 0 0 54 0. 864 46 7.1400 0.1015 5.379 3. 0 0 0 53 -1. 082 47 15.7000 0.2690 14.258 20. 000 53 1. 7 7 9 48 32.9000 0.7596 40.256 38. 000 53 -0. 725 49 71.4000 0.9999 52.993 53. 000 53 0. 0 82 50 51 C h i ^ 2 = 7 . 5 0 d.f. = 4 P-value = 0.1119 52 53 54 B e n c h m a r k D o s e C o m p u t a t i o n 55 56 S p e c i f i e d e f f e c t = 0.1 57 58 R i s k T y p e = Extra risk 59 60 C o n f i d e n c e l e v e l = 0.95 61 62 BMD = 9.1026 63 64 BMDL = 7.7011 65 This document is a draftfor review purposes only and does not constitute Agency policy. E-444 DRAFT--DO NOT CITE OR QUOTE 1 E.3.37.3. F igure f o r Selected M odel: Probit Probit Model with 0.95 Confidence Level Fraction Affected 2 19:51 02/16 2010 3 dose 4 5 E .3.37.4. Outputfor Additional Model Presented: Weibull 6 National T oxicology Program, 2006: Oval Cell Hyperplasia 7 8 9 10 W e i b u l l M o d e l u s i n g W e i b u l l M o d e l ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 5 3 _ N T P _ 2 0 0 6 _ O v a l H y p e r _ W e i b u l l _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 5 3 _ N T P _ 2 0 0 6 _ O v a l H y p e r _ W e i b u l l _ 1 . p l t 13 ~ T u e F e b _ 16 1 9 : 5 1 : 5 3 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) * [ 1 - E X P ( - s l o p e * d o s e ^ p o w e r ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 P o w e r p a r a m e t e r is r e s t r i c t e d as p o w e r > = 1 27 28 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 This document is a draftfor review purposes only and does not constitute Agency policy. E-445 DRAFT--DO NOT CITE OR QUOTE 1 Total number of records with missing values = 0 2 Maximum number of iterations = 250 3 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 4 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 5 6 7 8 Default Initial (and Specified) Parameter Values 9 Background = 0.00925926 10 Slope = 0.0044452 11 Power = 1.63009 12 13 14 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 15 16 Background Slope Power 17 CO o CO o 18 B a c k g r o u n d 1 0.61 19 20 S l o p e 1 -0.99 21 22 P o w e r 0.61 -0.99 1 23 24 25 26 P a r a m e t e r E s t i m a t e s 27 28 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 29 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 30 B a c k g r o u n d 0.021258 0.0198428 -0.0176332 0.0601492 31 Slope 0.0028715 0.00303327 -0.0030736 0.0088166 32 Power 1.76359 0.309457 1.15706 2.37011 33 34 35 36 A n a l y s i s o f D e v i a n c e T a b l e 37 38 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 39 Full model -92.4898 6 40 F i t t e d m o d e l -96.3448 3 7.70998 3 0.0524 41 R e d u c e d m o d e l -210.191 1 235.402 5 <.0001 42 43 AIC: 198.69 44 45 46 G o o d n e s s o f F i t 47 S c a l e d 48 Dose Est. Prob. Expected Observed Size Residual 49 50 0.0000 0.0213 1.127 0.000 53 -1.073 51 2.1400 0.0320 1.725 4.000 54 1 . 7 6 0 52 7.1400 0.1073 5.685 3.000 53 -1.192 53 15.7000 0.3234 17.138 20.000 53 0.840 54 32.9000 0.7490 39.698 38.000 53 -0.538 55 71.4000 0.9953 52.750 53.000 53 0.501 56 57 C h i ^ 2 = 6 . 9 2 d.f. = 3 P-value = 0.0746 58 59 60 B e n c h m a r k D o s e C o m p u t a t i o n 61 62 S p e c i f i e d e f f e c t 0.1 63 64 R i s k T y p e Extra risk 65 66 C o n f i d e n c e l e v e l 0.95 67 68 BMD 7.71171 69 70 BMDL 4.69152 This document is a draftfor review purposes only and does not constitute Agency policy. E-446 DRAFT--DO NOT CITE OR QUOTE 1 2 E.3.37.5. F igure f o r A d d itio n a l M o d el Presented: W eibull Weibull Model with 0.95 Confidence Level Fraction Affected 3 19:51 02/16 2010 4 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-447 DRAFT--DO NOT CITE OR QUOTE 1 E.3.38. National Toxicology Program, 2006: Pigmentation, Liver 2 E.3.38.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes gamma 3 0.385 197.655 1.547E+00 8.055E-01 logistic negative intercept (intercept = 4 <0.001 203.517 2.259E+00 1.872E+00 -1.925) log-logistic 3 0.978 195.600 2.212E+00 1.452E+00 log-probit a multistage, 5degree probit W eibull 3 0.980 195.450 2.072E+00 1.399E+00 3 0.210 199.850 9.396E-01 7.079E-01 final b = 0 negative intercept (intercept = 4 <0.001 210.309 2.259E+00 1.916E+00 -1.057) 3 0.290 198.489 1.280E+00 7.518E-01 aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3 .3 8 .2 . Outputfor Selected Model: Log-Probit 6 National T oxicology Program, 2006: Pigmentation, Liver 7 8 9 10 P r o b i t M o d e l . ( V e r s i o n : 3.1; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 5 4 _ N T P _ 2 0 0 6 _ P i g m e n t _ L o g P r o b i t _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 5 4 _ N T P _ 2 0 0 6 _ P i g m e n t _ L o g P r o b i t _ 1 . p l t 13 ~ T u e F e b 16 1 9 : 5 2 : 1 9 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = B a c k g r o u n d 22 + ( 1 - B a c k g r o u n d ) * C u m N o r m ( I n t e r c e p t + S l o p e * L o g ( D o s e ) ) , 23 24 w h e r e C u m N o r m ( . ) is t h e c u m u l a t i v e n o r m a l d i s t r i b u t i o n f u n c t i o n 25 26 27 D e p e n d e n t v a r i a b l e = D i c h E f f 28 I n d e p e n d e n t v a r i a b l e = D o s e 29 S l o p e p a r a m e t e r is r e s t r i c t e d as s l o p e > = 1 30 31 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 32 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 33 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 34 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 35 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 36 37 38 This document is a draftfor review purposes only and does not constitute Agency policy. E-448 DRAFT--DO NOT CITE OR QUOTE 1 User has chosen the log transformed model 2 3 4 Default Initial (and Specified) Parameter Values 5 background = 0.0754717 6 intercept = -1.91144 7 slope = 1.07385 8 9 10 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 11 12 background intercept slope 13 Lf) o Lf) o 14 b a c k g r o u n d 1 0.35 15 16 i n t e r c e p t 1 -0.94 17 18 s l o p e 0.35 -0.94 1 19 20 21 22 P a r a m e t e r E s t i m a t e s 23 24 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 25 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 26 b a c k g r o u n d 0 . 0 7 3 5 9 5 6 0 . 0 3 4 3 2 8 4 0.00631316 0.140878 27 intercept -2.19294 0.400053 -2.97703 -1.40885 28 slope 1.25068 0.169731 0.918012 1.58335 29 30 31 32 A n a l y s i s o f D e v i a n c e T a b l e 33 34 Model Log(likelihood) # Param's Deviance Test d.f. P-value 35 Full model -94.6177 6 36 F i t t e d m o d e l -94.7248 3 0.214232 3 0.975 37 R e d u c e d m o d e l -210.717 1 232.198 5 <.0001 38 39 AIC: 195.45 40 41 42 G o o d n e s s o f F i t 43 S c a l e d 44 Dose Est. Prob. Expected Observed Size Residual 45 46 .0000 0.0736 3.901 4.000 53 0. 0 52 47 .1400 0.1729 9.338 9.000 54 -0. 122 48 .1400 0.6338 33.591 34.000 53 0. 1 17 49 .7000 0.9023 47.822 48.000 53 0. 0 82 50 .9000 0.9863 52.275 52.000 53 -0. 325 51 .4000 0.9992 52.959 53.000 53 0. 2 0 2 52 53 C h i ^ 2 = 0 . 1 8 d.f. = 3 P-value = 0.9801 54 55 56 B e n c h m a r k D o s e C o m p u t a t i o n 57 58 S p e c i f i e d e f f e c t 0.1 59 60 R i s k T y p e Extra risk 61 62 C o n f i d e n c e l e v e l 0.95 63 64 BMD 2.07241 65 66 BMDL 1.39932 67 68 This document is a draftfor review purposes only and does not constitute Agency policy. E-449 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 E.3.38.3. F igure f o r Selected M odel: Log-P robit LogProbit Model with 0.95 Confidence Level 2 19:52 02/16 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-450 DRAFT--DO NOT CITE OR QUOTE 1 E.3.39. National Toxicology Program, 2006: Toxic Hepatopathy 2 E.3.39.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2pValue AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes gamma 4 0.772 185.634 4.668E+00 3.317E+00 logistic negative intercept (intercept = 4 0.012 198.445 7.070E+00 5.925E+00 -2.925) log-logistic 3 0.362 190.061 5.676E+00 4.040E+00 log-probit multistage, 5degree a probit W eibull 3 0.378 189.858 6.061E+00 4.079E+00 4 0.577 186.521 4.163E+00 2.701E+00 final B = 0 negative intercept (intercept = 4 0.019 197.159 6.784E+00 5.712E+00 -1.724) 4 0.745 185.657 4.454E+00 3.159E+00 aBest-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3.39.2. Outputfor Selected Model: Multistage, 5-Degree 6 National T oxicology Program, 2006: Toxic Hepatopathy 7 8 9 10 M u l t i s t a g e M o d e l . ( V e r s i o n : 3.0; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 5 5 _ N T P _ 2 0 0 6 _ T o x H e p a _ M u l t i 5 _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 5 5 _ N T P _ 2 0 0 6 _ T o x H e p a _ M u l t i 5 _ 1 . p l t 13 T u e F e b 16 1 9 : 5 2 : 4 9 2 0 1 0 14 15 16 0 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) * [ 1 - E X P ( 22 - b e t a i * d o s e ^ 1 - b e t a 2 * d o s e ^ 2 - b e t a 3 * d o s e ^ 3 - b e t a 4 * d o s e ^ 4 - b e t a 5 * d o s e ^ 5 ) ] 23 24 T h e p a r a m e t e r b e t a s a r e r e s t r i c t e d t o b e p o s i t i v e 25 26 27 D e p e n d e n t v a r i a b l e = D i c h E f f 28 I n d e p e n d e n t v a r i a b l e = D o s e 29 30 T o t a l n u m b e r o f o b s e r v a t i o n s = 6 31 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 T o t a l n u m b e r o f p a r a m e t e r s i n m o d e l = 6 33 T o t a l n u m b e r o f s p e c i f i e d p a r a m e t e r s = 0 34 D e g r e e o f p o l y n o m i a l = 5 35 36 37 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 38 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 This document is a draftfor review purposes only and does not constitute Agency policy. E-451 DRAFT--DO NOT CITE OR QUOTE 1 Parameter Convergence has been set to: 1e-008 2 3 4 5 Default Initial Parameter Values 6 Background = 0 7 Beta(1) = 0 8 Beta(2) = 0 9 Beta(3) = 0 10 Beta(4) = 0 11 B e t a ( 5 ) = 5 . 4 0 9 8 3 e + 0 1 0 12 13 14 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 15 16 ( *** The model parameter(s) -Background -Beta(3) -Beta(4) -Beta(5) 17 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 18 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 19 20 Beta(1) Beta(2) 21 22 B e t a ( 1 ) 1 -0.91 23 24 B e t a ( 2 ) -0.91 1 25 26 27 28 P a r a m e t e r E s t i m a t e s 29 30 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 31 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 32 B a c k g r o u n d 0* * * 33 Beta(1) 0.019656 * * * 34 Beta(2) 0.00135796 * * * 35 B e t a ( 3 ) 0* * * 36 B e t a ( 4 ) 0* * * 37 B e t a ( 5 ) 0* * * 38 39 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 40 41 42 43 A n a l y s i s o f D e v i a n c e T a b l e 44 45 Model Log(likelihood) # Param's Deviance Test d.f. P-value 46 Full model -89.8076 6 47 F i t t e d m o d e l -91.2606 2 2.90597 4 0.5737 48 R e d u c e d m o d e l -218.207 1 256.799 5 <.0001 49 50 AIC: 186.521 51 52 53 G o o d n e s s o f F i t 54 S c a l e d 55 Est. Prob. Expected Observed Residual 56 57 0.0000 0. 0 0 0 0 0.000 0. 0 0 0 53 0. 0 0 0 58 2.1400 0. 0 4 7 1 2.545 2. 0 0 0 54 -0. 350 59 7.1400 0. 1 8 9 1 10.021 8. 0 0 0 53 -0. 709 60 15.7000 0. 4 7 4 5 25.146 30. 000 53 1. 3 3 5 61 32.9000 0. 87 96 46.616 45. 000 53 -0. 682 62 71.4000 0. 9 9 9 8 52.987 53. 000 53 0. 1 1 3 63 64 C h i ^ 2 = 2 . 8 9 d.f. = 4 P-value = 0.5771 65 66 67 B e n c h m a r k D o s e C o m p u t a t i o n 68 69 S p e c i f i e d e f f e c t = 0.1 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-452 DRAFT--DO NOT CITE OR QUOTE 1 Risk Type Extra risk 2 3 Confidence level = 0.95 4 5 BMD = 4.16294 6 7 BMDL = 2.70063 8 9 BMDU = 6.00186 10 11 T a k e n t o g e t h e r , ( 2 . 7 0 0 6 3 , 6 . 0 0 1 8 6 ) is a 90 12 i n t e r v a l f o r t h e B M D 13 14 two-sided confidence 15 E .3 .3 9 .3 . Figurefor Selected Model: Multistage, 5-Degree Multistage Model with 0.95 Confidence Level Fraction Affected 16 19:52 02/16 2010 17 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-453 DRAFT--DO NOT CITE OR QUOTE 1 E .3.40. O hsako et al., 2001: A no-G enital L ength, PN D 120 2 E .3 .4 0 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M 2) 3 0.019 171.804 5.650E+02 3.785E+02 Notes exponential (M 3) 3 0.019 171.804 5 .6 5 0 E + 0 2 3 .7 8 5 E + 0 2 p o w er hit b ound (d = 1) exponential (M 4) 2 0.117 168.204 2.854E+01 1.054E+01 exponential (M 5) Hill b 1 0.049 169.789 2.948E+01 1.135E+01 2 0.148 167.727 3.722E+01 9.752E+00 n lower bound hit (n = 1) linear polynom ial, 4degree power 3 0.018 171.954 5.852E+02 4.047E+02 3 0.018 171.954 5.852E+02 4.047E+02 3 0 .0 1 8 171.954 5 .8 5 2 E + 0 2 4 .0 4 7 E + 0 2 p o w er b oun d hit (p ow er = 1) H ill, unrestricted c 1 0.055 169.600 5.101E+01 3.066E+00 unrestricted (n = 0.502) power, unrestricted 2 0.151 167.689 6.200E+01 2.291E +00 unrestricted (power = 0.252) a C onstant variance m odel selected (p = 0.1 6 5 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .4 0 .2 . Outputfor Selected Model: Hill 6 O hsako et al., 2001: A no-G enital Length, P N D 120 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 5 6 _ O h s a k o _ 2 0 0 1 _ A n o g e n _ H i l l C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 5 6 _ O h s a k o _ 2 0 0 1 _ A n o g e n _ H i l l C V _ 1 . p l t 13 T u e F e b _ 16 1 9 : 5 3 : 2 5 2 0 1 0 14 15 16 F i g u r e 7 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 This document is a draftfor review purposes only and does not constitute Agency policy. E-454 DRAFT--DO NOT CITE OR QUOTE 1 Power parameter restricted to be greater than 1 2 A constant v ariance model is fit 3 4 Total number of dose groups = 5 5 Total number of records with missing values = 0 6 Maximum number of iterations = 250 7 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 8 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 9 10 11 12 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 13 alpha 7.27386 14 rho 0 Specified 15 intercept 28.905 16 - 5 . 1 0 6 5 17 1 . 4 0 2 2 6 18 3 3 . 9 6 6 9 19 20 21 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 22 23 ( *** The model parameter(s) -rho -n 24 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 25 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 26 27 alpha intercept v k 28 29 a l p h a 1 -2.2e-009 -2.4e-008 -7.2e-009 30 31 i n t e r c e p t 2.2e-009 1 -0.66 -0.5 32 33 v 2.4e-008 -0.66 1 -0.11 34 35 k 7.2e-009 -0.5 -0.11 1 36 37 38 39 P a r a m e t e r E s t i m a t e s 40 41 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 42 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 43 alpha 7.08444 1.3634 4.41223 9.75666 44 i n t e r c e p t 28.9809 0.745637 27.5195 30.4423 45 v -4.79692 0.983318 -6.72418 -2.86965 46 n 1 N A 47 k 29.8628 24.4463 -18.0511 77.7767 48 49 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 50 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 51 h a s n o s t a n d a r d e r r o r . 52 53 54 55 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 56 57 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled R 58 59 60 0 12 28.9 29 3.13 2.66 -0.0988 61 12.5 10 27.9 27.6 2.5 2.66 0.442 62 50 10 25.2 26 3.21 2.66 -0.963 63 200 10 26 24.8 2.85 2.66 1.42 64 800 12 23.8 24.4 1.56 2.66 -0.726 65 66 67 68 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-455 DRAFT--DO NOT CITE OR QUOTE 1 Model A1: Yij = Mu(i) + e(ij) 2 Var{e(ij)} = Sigma^2 3 4 Model A2: Yij = Mu(i) + e(ij) 5 Var{e(ij)} = Sigma(i)^2 6 7 Model A3: Yij = Mu(i) + e(ij) 8 Var{e(ij)} = Sigma^2 9 Model A3 uses any fixed variance parameters that 10 w e r e s p e c i f i e d b y t h e u s e r 11 12 M o d e l R: Yi = M u + e(i) 13 V a r { e ( i ) } = S i g m a ^ 2 14 15 16 L i k e l i h o o d s o f I n t e r e s t 17 18 Model Log(likelihood) # Param's AIC 19 A1 -77.952340 6 167.904680 20 A2 -74.703868 10 1 6 9.407736 21 A3 -77.952340 6 167.904680 22 fitted -79.863340 4 167.726680 23 R -89.824703 2 183.649405 24 25 26 E x p l a n a t i o n o f T e s t s 27 28 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 29 (A2 vs. R) 30 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 31 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 32 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 33 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 34 35 T e s t s o f I n t e r e s t 36 37 Test -2*log(Likelihood Ratio) Test df p-value 38 39 T e s t 1 30.2417 8 0.0001916 40 T e s t 2 6.49694 4 0.165 41 T e s t 3 6.49694 4 0.165 42 T e s t 4 3.822 2 0.1479 43 44 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 45 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 46 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 47 48 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 49 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 50 51 52 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 53 t o b e a p p r o p r i a t e h e r e 54 55 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 56 t o a d e q u a t e l y d e s c r i b e t h e d a t a 57 58 59 B e n c h m a r k D o s e C o m p u t a t i o n 60 61 S p e c i f i e d e f f e c t = 1 62 63 R i s k T y p e = Estimated standard deviations from the control mean 64 65 C o n f i d e n c e l e v e l = 0.95 66 67 BMD = 37.2249 68 69 BMDL = 9.75249 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-456 DRAFT--DO NOT CITE OR QUOTE 1 E .3 .4 0 .3 . Figurefor Selected Model: Hill Hill Model with 0.95 Confidence Level Mean Response 2 19:53 02/16 2010 3 dose 4 5 E .3.40.4. Outputfor Additional Model Presented: Hill, Unrestricted 6 O hsako et al., 2001: A no-G enital Length, P N D 120 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 5 6 _ O h s a k o _ 2 0 0 1 _ A n o g e n _ H i l l C V _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 5 6 _ O h s a k o _ 2 0 0 1 _ A n o g e n _ H i l l C V _ U _ 1 . p l t 13 T u e F e b _ 16 1 9 : 5 3 2 6 2 0 1 0 14 15 16 F i g u r e 7 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 P o w e r p a r a m e t e r is n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t This document is a draftfor review purposes only and does not constitute Agency policy. E-457 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 5 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha 7.27386 12 rho 0 Specified 13 intercept 28.905 14 - 5 . 1 0 6 5 15 1 . 4 0 2 2 6 16 3 3 . 9 6 6 9 17 18 19 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 20 21 ( j T h e m o d e l p a r a m e t e r ( s ) - r h o 22 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r 23 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 24 25 alpha intercept v n k 26 27 a l p h a 1 2.1e-009 -1.8e-008 -1.7e-008 1.6e -008 28 29 i n t e r c e p t 2.1e-009 1 0.012 0.0075 - 0.13 30 31 v -1.8e-008 0.012 1 0.98 - 0.99 32 33 n -1.7e-008 0.0075 0.98 1 - 0.97 34 35 k 1.6e-008 -0.13 -0.99 -0.97 1 36 37 38 39 P a r a m e t e r E s t i m a t e s 40 41 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 42 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 43 alpha 7. 0 6 7 8 5 1.36021 4.40189 9.73381 44 i n t e r c e p t 28 .9608 0.755363 27.4803 30.4413 45 v -6. 94236 12.2514 -30.9547 17.07 46 n 0.501942 0.915162 -1.29174 2.29563 47 k 131.957 1071.9 -1968.92 2232.84 48 49 50 51 T a b l e o f D a t a a n d ]E s t i m a t e d V a l u e s o f I n t e r e s t 52 53 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res. 54 55 56 0 12 28.9 29 3.13 2.66 -0 .0727 57 12.5 10 27.9 27.3 2.5 2.66 0.72 58 50 10 25.2 26.3 3.21 2.66 -1.37 59 200 10 26 25.1 2.85 2.66 1.04 60 800 12 23.8 24 1.56 2 . 6 6 - 0.287 61 62 63 64 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 65 66 67 M o d e l A1: Yij = Mu(i) + e(ij) 68 V a r { e ( i j ) } = S i g m a ^ 2 69 70 M o d e l A2: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-458 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma(i)^2 2 3 Model A3: Yij = Mu(i) + e(ij) 4 Var{e(ij)} = Sigma^2 5 Model A3 uses any fixed variance parameters that 6 were specified by the user 7 8 M o d e l R: Yi = M u + e(i) 9 Var{e(i)} = Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) # Param's AIC 15 A1 -77.952340 6 167.904680 16 A2 -74.703868 10 1 6 9.407736 17 A3 -77.952340 6 167.904680 18 fitted -79.800035 5 169.600070 19 R -89.824703 2 183.649405 20 21 22 E x p l a n a t i o n o f T e s t s 23 24 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 25 (A2 vs. R) 26 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 27 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 28 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 29 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 30 31 T e s t s o f I n t e r e s t 32 33 Test -2*log(Likelihood Ratio) Test df p-value 34 35 T e s t 1 30.2417 8 0.0001916 36 T e s t 2 6.49694 4 0.165 37 T e s t 3 6.49694 4 0.165 38 T e s t 4 3.69539 1 0.05456 39 40 p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e 41 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 42 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 43 44 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 45 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 46 47 48 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 49 t o b e a p p r o p r i a t e h e r e 50 51 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 52 m o d e l 53 54 55 B e n c h m a r k D o s e C o m p u t a t i o n 56 57 S p e c i f i e d e f f e c t = 1 58 59 R i s k T y p e = Estimated standard deviations from the control mean 60 61 C o n f i d e n c e l e v e l = 0.95 62 63 BMD = 51.0107 64 65 BMDL = 3.06631 66 This document is a draftfor review purposes only and does not constitute Agency policy. E-459 DRAFT--DO NOT CITE OR QUOTE 1 E.3.40.5. F igure f o r A d d itio n a l M o d el Presented: H ill, U nrestricted Hill Model with 0.95 Confidence Level Mean Response 2 19:53 02/16 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-460 DRAFT--DO NOT CITE OR QUOTE 1 E.3.41. Sew all et al., 1995: T4 In Serum 2 E .3 .4 1 .1 . Summary Table o f BMDS Modeling Results Model a Degrees of x2p Freedom Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M 2) 3 0.424 205.966 5.762E+01 3.783E+01 Notes exponential (M 3) 3 0.424 2 0 5 .9 6 6 5.7 6 2 E + 0 1 3.7 8 3 E + 0 1 p o w er hit bound (d = 1) exponential (M 5) Hill b 2 0.611 2 0 6 .1 5 2 2 .5 2 3 E + 0 1 8 .4 4 2 E + 0 0 p o w er hit bound (d = 1) 2 0.702 205.875 2.071E+01 5.164E+00 n lower bound hit (n = 1) linear polynom ial, 4degree power 3 0.332 206.584 6.788E+01 4.858E+01 3 0.332 206.584 6.788E+01 4.858E+01 3 0 .3 3 2 2 0 6 .5 8 4 6 .7 8 8 E + 0 1 4 .8 5 8 E + 0 1 p o w er b oun d hit (p ow er = 1) Hill, unrestricted c 1 0.844 207.205 1.657E+01 1.903E+00 unrestricted (n = 0.427) power, unrestricted 2 0.983 205.200 1.658E+01 1.820E+00 unrestricted (power = 0.403) a Constant variance m odel selected (p = 0.4 0 7 8 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 E .3 .4 1 .2 . Outputfor Selected Model: Hill 5 Sew all et al., 1995: T4 In Serum 6 7 8 9 Hill Model. (Version: 2.14; Date: 06/26/2008) 10 I n p u t D a t a F i l e : C : \ 1 \ 5 8 _ S e w a l l _ 1 9 9 5 _ T 4 _ H i l l C V _ 1 . ( d ) 11 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 5 8 _ S e w a l l _ 1 9 9 5 _ T 4 _ H i l l C V _ 1 . p l t 12 T u e _ F e b 1 6 ~ 1 9 : 5 4 : 3 0 2 0 1 0 13 14 15 F i g u r e 1, S a l i n e n o n i n i t i a t e d 16 17 18 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 19 20 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 21 22 23 D e p e n d e n t v a r i a b l e = M e a n 24 I n d e p e n d e n t v a r i a b l e = D o s e 25 r h o is s e t t o 0 26 P o w e r p a r a m e t e r r e s t r i c t e d t o b e g r e a t e r t h a n 1 27 A c o n s t a n t v a r i a n c e m o d e l is f i t 28 29 T o t a l n u m b e r o f d o s e g r o u p s = 5 30 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 This document is a draftfor review purposes only and does not constitute Agency policy. E-461 DRAFT--DO NOT CITE OR QUOTE 1 Maximum number of 250 2 Relative Function has b e e n set to: 1e-008 3 Parameter has b een set to: 1e-008 4 5 6 7 Default Initial Parameter Values 8 alpha 33.0913 9 rho 0 Specified 10 intercept 30.6979 11 - 1 2 . 2 9 3 7 12 0 . 6 9 5 3 8 4 13 2 4 . 6 6 7 4 14 15 16 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 17 18 ( *** The model parameter(s) -rho -n 19 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 20 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 21 22 alpha intercept v k 23 24 a l p h a 1 1.2e-008 4.1e-008 -2.4e-008 25 26 i n t e r c e p t 1.2e-008 1 0.14 -0.66 27 28 v 4.1e-008 0.14 1 -0.76 29 30 k -2.4e-008 -0.66 -0.76 1 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 37 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 38 alpha 29.8807 6.29941 17.5341 42.2274 39 ntercept 29.9609 1.64749 26.7319 33.1899 40 v -14.2338 4.35645 -22.7723 -5.69537 41 n 1 N A 42 k 33.2198 37.0852 -39.4658 105.905 43 44 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 45 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 46 h a s n o s t a n d a r d e r r o r . 47 48 49 50 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 51 52 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled R 53 54 55 09 30 .7 30 4.66 5.47 0. 404 56 Lf) CO 9 27 .9 28.6 7.17 5.47 -0. 399 57 1 0 . 7 9 25 .9 26.5 6.81 5.47 -0. 328 58 35 9 23 .6 22.7 5.38 5.47 0. 4 9 3 59 125 9 18 .4 18.7 4.12 5.47 -0. 171 60 61 62 63 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 64 65 66 M o d e l A1: Yij = Mu(i) + e(ij) 67 V a r { e i ;i j ) } = S i g m a ^ 2 68 69 M o d e l A2: Yij = Mu(i) + e(ij) 70 V a r { e i :i j ) } = S i g m a ( i ) ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. E-462 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A3: Yij = Mu(i) + e(ij) 3 Var{e(ij)} = Sigma^2 4 Model A3 uses any fixed variance parameters that 5 were specified by the user 6 7 M o d e l R: Yi = M u + e(i) 8 Var{e(i)} = Sigma^2 9 10 11 L i k e l i h o o d s o f I n t e r e s t 12 13 Model Log(likelihood) # Param's AIC 14 A1 -98.583448 6 209.166896 15 A2 -96.590204 10 213.180407 16 A3 -98.583448 6 209.166896 17 fitted -98.937315 4 205.874631 18 R -109.013252 2 222.026503 19 20 21 E x p l a n a t i o n o f T e s t s 22 23 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 24 (A2 vs. R) 25 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 26 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 27 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 28 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 29 30 T e s t s o f I n t e r e s t 31 32 Test -2*log(Likelihood Ratio) Test df p-val 33 34 T e s t 1 24.8461 8 0.001651 35 T e s t 2 3.98649 4 0.4078 36 T e s t 3 3.98649 4 0.4078 37 T e s t 4 0.707735 2 0.702 38 39 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 40 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 41 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 42 43 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 44 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 45 46 47 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 48 t o b e a p p r o p r i a t e h e r e 49 50 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 51 t o a d e q u a t e l y d e s c r i b e t h e d a t a 52 53 54 B e n c h m a r k D o s e C o m p u t a t i o n 55 56 S p e c i f i e d e f f e c t = 1 57 58 R i s k T y p e = Estimated standard deviations from the control mean 59 60 C o n f i d e n c e l e v e l = 0.95 61 62 BMD = 20.7117 63 64 BMDL = 5.16405 65 This document is a draftfor review purposes only and does not constitute Agency policy. E-463 DRAFT--DO NOT CITE OR QUOTE 1 E .3 .4 1 .3 . Figurefor Selected Model: Hill Hill Model with 0.95 Confidence Level Mean Response 2 19:54 02/16 2010 3 dose 4 5 E .3.41.4. Outputfor Additional Model Presented: Hill, Unrestricted 6 Sew all et al., 1995: T4 In Serum 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 5 8 _ S e w a l l _ 1 9 9 5 _ T 4 _ H i l l C V _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 5 8 _ S e w a l l _ 1 9 9 5 _ T 4 _ H i l l C V _ U _ 1 . p l t 13 T u e _ F e b 1 6 ~ 1 9 : 5 4 : 3 1 2 0 1 0 14 15 16 F i g u r e 1, S a l i n e n o n i n i t i a t e d 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 P o w e r p a r a m e t e r is n o t r e s t r i c t e d 28 A c o n s t a n t v a r i a n c e m o d e l is f i t This document is a draftfor review purposes only and does not constitute Agency policy. E-464 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 5 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha 33.0913 12 rho 0 Specified 13 intercept 30.6979 14 - 1 2 . 2 9 3 7 15 0 . 6 9 5 3 8 4 16 2 4 . 6 6 7 4 17 18 19 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 20 21 ( *** T h e m o d e l p a r a m e t e r ( s ) -rho 22 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r 23 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 24 25 alpha intercept v n k 26 27 a l p h a 1 -0.0004 0.0059 0.0048 -0.0059 28 29 i n t e r c e p t -0.0004 1 -0.026 -0.44 0.07 30 31 v 0.0059 -0.026 1 0.77 -1 32 33 n 0.0048 -0.44 0.77 1 -0.82 34 35 k -0.0059 0.07 -1 -0.82 1 36 37 38 39 P a r a m e t e r E s t i m a t e s 40 41 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 42 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 43 alpha 29.4396 6.20653 17.2751 41.6042 44 i n t e r c e p t 30.6757 1.77521 27.1963 34.155 45 v -141.324 1202.4 -2497.98 2215.33 46 n 0.426599 0.262207 -0.0873175 0.940515 47 k 31487 770429 -1.47853e+006 1 .5415e+006 48 49 50 51 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 52 53 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 54 55 56 09 30 .7 30.7 4.66 5.43 0.0123 57 Lf) CO 9 27 .9 27.8 7.17 5.43 0.0279 58 1 0 . 7 9 25 .9 26.1 6.81 5.43 -0.137 59 35 9 23 .6 23.3 5.38 5.43 0.132 60 125 9 18 .4 18.5 4.12 5.43 -0.0354 61 62 63 64 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 65 66 67 M o d e l A1: Yij = Mu(i) + e(ij) 68 V a r { e ( i j ) } = S i g m a ^ 2 69 70 M o d e l A2: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-465 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma(i)^2 2 3 Model A3: Yij = Mu(i) + e(ij) 4 Var{e(ij)} = Sigma^2 5 Model A3 uses any fixed variance parameters that 6 were specified by the user 7 8 M o d e l R: Yi = M u + e(i) 9 Var{e(i)} = Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) # Param's AIC 15 A1 -98.583448 6 209.166896 16 A2 -96.590204 10 213.180407 17 A3 -98.583448 6 209.166896 18 fitted -98.602701 5 207.205403 19 R -109.013252 2 222.026503 20 21 22 E x p l a n a t i o n o f T e s t s 23 24 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 25 (A2 vs. R) 26 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 27 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 28 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 29 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 30 31 T e s t s o f I n t e r e s t 32 33 Test -2*log(Likelihood Ratio) Test df p-value 34 35 T e s t 1 24.8461 8 0.001651 36 T e s t 2 3.98649 4 0.4078 37 T e s t 3 3.98649 4 0.4078 38 T e s t 4 0.0385071 1 0.8444 39 40 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 41 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 42 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 43 44 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 45 m o d e l a p p e a r s t o b e a p p r o p r i a t e h e r e 46 47 48 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 49 t o b e a p p r o p r i a t e h e r e 50 51 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 52 t o a d e q u a t e l y d e s c r i b e t h e d a t a 53 54 55 B e n c h m a r k D o s e C o m p u t a t i o n 56 57 S p e c i f i e d e f f e c t = 1 58 59 R i s k T y p e = Estimated standard deviations from the control mean 60 61 C o n f i d e n c e l e v e l = 0.95 62 63 BMD = 16.5689 64 65 BMDL = 1.90347 66 This document is a draftfor review purposes only and does not constitute Agency policy. E-466 DRAFT--DO NOT CITE OR QUOTE 1 E.3.41.5. F igure f o r A d d itio n a l M o d el Presented: H ill, U nrestricted Hill Model with 0.95 Confidence Level Mean Response 2 19:54 02/16 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-467 DRAFT--DO NOT CITE OR QUOTE 1 E.3.42. Shi et al., 2007: Estradiol 17B, PE9 2 E.3.42.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of x2p Freedom Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M 2) 3 0.001 395.701 1.729E+01 8.956E +00 Notes exponential (M 3) 3 0.001 39 5 .7 0 1 1.729E +01 8 .9 5 6 E + 0 0 p o w er hit b ound (d = 1) exponential (M4) b 2 0.494 383.635 5.559E-01 2.236E-01 exponential (M 5) 2 0.494 3 8 3 .6 3 5 5 .559E -01 2.2 3 6 E -0 1 p o w er hit b ound (d = 1) H ill 2 0.773 382.743 4.434E -01 error n lo w er bound hit (n = 1) linear polynomial, 4degree power 3 0.001 397.484 2.243E+01 1.523E+01 3 0.001 397.484 2.243E+01 1.523E+01 3 0.001 3 9 7 .4 8 4 2.2 4 3 E + 0 1 1.523E +01 p o w er b oun d hit (p ow er = 1) Hill, unrestricted 1 0.874 384.251 3.998E -01 error unrestricted (n = 0.616) power, unrestricted 2 0.506 383.589 3.409E-01 5.002E-03 unrestricted (pow er = 0.155) a N on-constant variance m odel selected (p = 0 .0521) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 E .3 .4 2 .2 . Outputfor Selected Model: Exponential (M4) 6 Shi et al., 2007: Estradiol 17B, PE9 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 5 9 _ S h i _ 2 0 0 7 _ E s t r a d i o l _ E x p _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 T u e F e b 16 1 9 : 5 5 : 0 6 2 0 1 0 14 15 16 F i g u r e 4 P E 9 o n l y 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; 27 s i g n = -1 f o r d e c r e a s i n g t r e n d . 28 29 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. This document is a draftfor review purposes only and does not constitute Agency policy. E-468 DRAFT--DO NOT CITE OR QUOTE 1 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 2 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 3 4 5 Dependent variable = Mean 6 Independent variable = Dose 7 Data are assumed to be distributed: normally 8 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 9 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) * rho) 10 11 T o t a l n u m b e r o f d o s e g r o u p s = 5 12 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 13 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 14 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 16 17 M L E s o l u t i o n p r o v i d e d : E x a c t 18 19 20 I n i t i a l P a r a m e t e r V a l u e s 21 22 Variable Model 4 23 24 lnalpha 2.65881 25 rho 0.913414 26 a 108 27 b 0 . 1 3 6 2 8 7 28 c 0 . 3 4 0 1 3 6 29 d 1 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 Variable Model 4 36 37 lnalpha 1.81331 38 rho 1.12126 39 a 1 0 0 . 5 2 6 40 b 1 . 5 3 8 2 3 41 c 0 . 4 3 1 7 9 6 42 d 1 43 44 45 T a b l e o f S t a t s F r o m I n p u t D a t a 46 47 Dose N Obs Mean Obs Std 48 49 0 10 102.9 41.41 50 0.143 10 86.19 19.58 51 0.714 10 63.33 29.36 52 7.14 10 48.1 18.82 53 28.6 10 38.57 22.59 54 55 56 E s t i m a t e d V a l u e s o f I n t e r e s t 57 58 Dose Est Mean Est Std Scaled Residual 59 60 0 100.5 32.83 0.2245 61 0 . 1 4 3 89.25 30.71 -0.3147 62 0 . 7 1 4 62.45 25.14 0.1108 63 7.14 43.41 20.5 0.723 64 28.6 43.41 20.5 -0.7458 65 66 67 68 O t h e r m o d e l s f o r w h i c h l i k e l i h o o d s a r e c a l c u l a t e d : 69 70 Model A1: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-469 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma^2 2 3 Model A2: Yij Mu(i) + e(ij) 4 Var{e(ij)} Sigma(i)^2 5 6 Model A3: Yij Mu(i) + e(ij) 7 Var{e(ij)} e x p ( l a l p h a + log(mean(i)) * rho) 8 9 M o d e l R: Yij M u + e(i) 10 Var{e(ij)} Sigma^2 11 12 13 L i k e l i h o o d s o f I n t e r e s t 14 15 Model Log(likelihood) DF AIC 16 17 A1 -188.3615 6 388.7231 18 A2 -183.667 10 387.3339 19 A3 -186.1132 7 386.2263 20 R -203.3606 2 410.7211 21 4 -186.8176 5 383.6352 22 23 24 Additive constant for all log-likelihoods = -45.95. This constant added to the 25 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 26 d e p e n d o n t h e m o d e l p a r a m e t e r s . 27 28 29 E x p l a n a t i o n o f T e s t s 30 31 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 32 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 33 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 34 35 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 36 37 38 T e s t s o f I n t e r e s t 39 40 T e s t -2*log(Likelihood Ratio) D. F. p-value 41 42 T e s t 1 39.39 8 < 0.0001 43 T e s t 2 9.389 4 0.05208 44 T e s t 3 4.892 3 0.1798 45 T e s t 6a 1.409 2 0.4944 46 47 48 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 49 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 50 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 51 52 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 53 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e . 54 55 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 56 v a r i a n c e a p p e a r s t o b e a p p r o p r i a t e h e r e . 57 58 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 59 t o a d e q u a t e l y d e s c r i b e t h e d a t a . 60 61 62 B e n c h m a r k D o s e C o m p u t a t i o n s : 63 64 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 65 66 R i s k T y p e = E s t i m a t e d s t a n d a r d d e v i a t i o n s f r o m c o n t r o l 67 68 C o n f i d e n c e L e v e l = 0 . 9 5 0 0 0 0 69 70 BMD = 0.555948 This document is a draftfor review purposes only and does not constitute Agency policy. E-470 DRAFT--DO NOT CITE OR QUOTE 1 2 BMDL 0.223612 3 4 5 E .3 .4 2 .3 . Figurefor Selected Model: Exponential (M4) Exponential_beta Model 4 with 0.95 Confidence Level Mean Response 6 19:55 02/16 2010 7 This document is a draftfor review purposes only and does not constitute Agency policy. E-471 DRAFT--DO NOT CITE OR QUOTE 1 E.3.43. Smialowicz et al., 2008: PFC per 10A6 Cells 2 E.3.43.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M2) 3 0.048 903.586 8.234E+01 4.833E+01 Notes exponential (M3) 3 0.048 9 0 3 .5 8 6 8.2 3 4 E + 0 1 4 .8 3 3 E + 0 1 p o w er hit boun d (d = 1) exponential (M4) 2 0.019 905.578 8.032E+01 6.220E+00 exponential (M5) 2 0.019 9 0 5 .5 7 8 8.0 3 2 E + 0 1 6 .2 2 0 E + 0 0 p o w er hit boun d (d = 1) H ill 2 0.026 90 4 .9 7 5 1.617E +01 2 .2 1 4 E + 0 0 n lo w er b ound hit (n = 1) linear polynom ial, 4degree power c 3 0.016 905.992 1.450E+02 1.102E+02 2 <0.0001 1198.471 1.375E+03 3.331E+01 3 0.016 9 0 5 .9 9 2 1 .450E + 02 1 .102E + 02 p o w er bound hit (p ow er = 1) H ill, unrestricted 1 0.183 901.442 8.297E+00 4.172E-01 unrestricted (n = 0.266) power, unrestricted b 2 0.446 899.282 7.676E+00 4.087E-01 unrestricted (power = 0.249) a C onstant variance m od el selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3.43.2. Outputfor Selected Model: Power, Unrestricted 6 S m ia lo w icz et al., 2008: PFC per 10A6 C ells 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 6 0 _ S m i a l _ 2 0 0 8 _ P F C c e l l s _ P w r C V _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 6 0 _ S m i a l _ 2 0 0 8 _ P F C c e l l s _ P w r C V _ U _ 1 . p l t 13 _ T u e F e b 16 1 9 : 5 5 7 5 3 2 0 1 0 14 15 16 A n t i R e s p o n s e t o S R B C s , P F C p e r 1 0 t o 6 c e l l s , T a b l e 4 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 This document is a draftfor review purposes only and does not constitute Agency policy. E-472 DRAFT--DO NOT CITE OR QUOTE 1 The power is not restricted 2 A constant v ariance model is fit 3 4 Total number of dose groups = 5 5 Total number of records with missing values = 0 6 Maximum number of iterations = 250 7 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 8 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 9 10 11 12 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 13 alpha = 232385 14 rho = 0 Specified 15 control = 1491 16 slope = -384.362 17 power = 0.215085 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( * * * T h e m o d e l p a r a m e t e r ( s ) - r h o 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 24 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 25 26 alpha control slope power 27 28 a l p h a 1 -1.5e-009 -8.2e-009 -1.1e-008 29 Lf) o Lf) o 30 control 1.5e-009 1 -0.79 31 32 slope 8.2e-009 -0.79 1 0.96 33 34 power 1.1e-008 0.96 1 35 36 37 38 P a r a m e t e r E s t i m a t e s 39 40 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 41 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 42 alpha 220294 38061.1 145696 294893 43 control 1470.38 124.07 1227.21 1713.55 44 slope -282.777 145.113 -567.193 1.64025 45 power 0.248621 0.0856348 0.0807799 0.416462 46 47 48 49 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 50 51 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 52 53 54 0 15 1.49e+003 1.47e+003 716 469 0.17 55 1.07 14 1 . 1 3 e + 0 0 3 1.18e+003 171 469 -0.429 56 10.7 15 945 961 516 469 -0.129 57 107 15 677 567 465 469 0.91 58 321 8 161 283 117 469 -0.735 59 60 61 62 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 63 64 65 M o d e l A1: Yij Mu(i) + e(ij) 66 Var{e(ij)} Sigma^2 67 68 M o d e l A2: Yij Mu(i) + e(ij) 69 Var{e(ij)} Sigma(i)^2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-473 DRAFT--DO NOT CITE OR QUOTE 1 Model A3: Yij = Mu(i) + e(ij) 2 Var{e(ij)} = Sigma^2 3 Model A3 uses any fixed variance parameters that 4 were specified by the user 5 6 M o d e l R: Yi = M u + e(i) 7 Var{e(i)} = Sigma^2 8 9 10 L i k e l i h o o d s o f I n t e r e s t 11 12 Model Log(likelihood) # Param's AIC 13 A1 -444.832859 6 901.665718 14 A2 -425.402825 10 870.805651 15 A3 -444.832859 6 901.665718 16 fitted -445.641102 4 899.282205 17 R -463.753685 2 931.507371 18 19 20 E x p l a n a t i o n o f T e s t s 21 22 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 23 (A2 vs. R) 24 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 25 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 26 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 27 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 28 29 T e s t s o f I n t e r e s t 30 31 Test -2*log(Likelihood Ratio) Test df p-value 32 33 T e s t 1 76.7017 8 <.0001 34 T e s t 2 38.8601 4 <.0001 35 T e s t 3 38.8601 4 <.0001 36 T e s t 4 1.61649 2 0.4456 37 38 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 39 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 40 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 41 42 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. C o n s i d e r r u n n i n g a 43 n o n - h o m o g e n e o u s v a r i a n c e m o d e l 44 45 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t t o c o n s i d e r a 46 d i f f e r e n t v a r i a n c e m o d e l 47 48 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 49 t o a d e q u a t e l y d e s c r i b e t h e d a t a 50 51 52 B e n c h m a r k D o s e C o m p u t a t i o n 53 54 S p e c i f i e d e f f e c t = 1 55 56 R i s k T y p e = Estimated standard deviations from the control mean 57 58 C o n f i d e n c e l e v e l = 0.95 59 60 B M D = 7 . 6 7 5 6 4 61 62 63 B M D L = 0 . 4 0 8 6 6 1 64 This document is a draftfor review purposes only and does not constitute Agency policy. E-474 DRAFT--DO NOT CITE OR QUOTE 1 E .3.43.3. Figurefor Selected Model: Power, Unrestricted Power Model with 0.95 Confidence Level 2 19:55 02/16 2010 3 dose 4 5 E .3 .4 3 .4 . Outputfor Additional Model Presented: Power 6 S m ia lo w icz et al., 2008: PFC per 10A6 C ells 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 6 0 _ S m i a l _ 2 0 0 8 _ P F C c e l l s _ P w r C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 6 0 _ S m i a l _ 2 0 0 8 _ P F C c e l l s _ P w r C V _ 1 . p l t 13 _ T u e F e b 16 1 9 : 5 5 : 5 3 2 0 1 0 14 15 16 A n t i R e s p o n s e t o S R B C s , P F C p e r 1 0 t o 6 c e l l s , T a b l e 4 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 r h o is s e t t o 0 27 T h e p o w e r is r e s t r i c t e d t o b e g r e a t e r t h a n o r e q u a l t o 1 28 A c o n s t a n t v a r i a n c e m o d e l is f i t This document is a draftfor review purposes only and does not constitute Agency policy. E-475 DRAFT--DO NOT CITE OR QUOTE 1 2 Total number of dose groups = 5 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha = 232385 12 rho = 0 Specified 13 control = 1491 14 slope = -2925.99 15 power = -0.136613 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 ( *** The model parameter(s) -rho -power 21 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 22 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 23 24 alpha control slope 25 26 a l p h a 1 3.6e-009 -1.2e-008 27 28 control 3.6e-009 1 -0.53 29 30 slope -1.2e-008 -0.53 1 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 37 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 38 alpha 250878 43345.1 165923 335833 39 control 1176.24 72.2586 1034.61 1317.86 40 slope -3.45384 0.592114 -4.61436 -2.29332 41 p o w e r 1 NA 42 43 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 44 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 45 h a s n o s t a n d a r d e r r o r . 46 47 48 49 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 50 51 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev S c a l e d Re; 52 53 54 0 15 1.49e+003 1.18e+003 716 501 2.43 55 1.07 14 1 . 1 3 e + 0 0 3 1.17e+003 171 501 -0.325 56 10.7 15 945 1.14e+003 516 501 -1.5 57 107 15 677 807 465 501 -1 58 321 8 161 67.6 117 501 0.528 59 60 61 62 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 63 64 65 M o d e l A1: Yij Mu(i) + e(ij) 66 Var{e(ij)} Sigma^2 67 68 M o d e l A2: Yij Mu(i) + e(ij) 69 Var{e(ij)} Sigma(i)^2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-476 DRAFT--DO NOT CITE OR QUOTE 1 Model A3: Yij = Mu(i) + e(ij) 2 Var{e(ij)} = Sigma^2 3 Model A3 uses any fixed variance parameters that 4 were specified by the user 5 6 M o d e l R: Yi = M u + e(i) 7 Var{e(i)} = Sigma^2 8 9 10 L i k e l i h o o d s o f I n t e r e s t 11 12 Model Log(likelihood) # Param's AIC 13 A1 -444.832859 6 901.665718 14 A2 -425.402825 10 870.805651 15 A3 -444.832859 6 901.665718 16 fitted -449.996183 3 905.992366 17 R -463.753685 2 931.507371 18 19 20 E x p l a n a t i o n o f T e s t s 21 22 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 23 (A2 vs. R) 24 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 25 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 26 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 27 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 28 29 T e s t s o f I n t e r e s t 30 31 Test -2*log(Likelihood Ratio) Test df p-value 32 33 T e s t 1 76.7017 8 <.0001 34 T e s t 2 38.8601 4 <.0001 35 T e s t 3 38.8601 4 <.0001 36 T e s t 4 10.3266 3 0.01598 37 38 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 39 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 40 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 41 42 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. C o n s i d e r r u n n i n g a 43 n o n - h o m o g e n e o u s v a r i a n c e m o d e l 44 45 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t t o c o n s i d e r a 46 d i f f e r e n t v a r i a n c e m o d e l 47 48 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 49 m o d e l 50 51 52 B e n c h m a r k D o s e C o m p u t a t i o n 53 54 S p e c i f i e d e f f e c t = 1 55 56 R i s k T y p e = Estimated standard deviations from the control mean 57 58 C o n f i d e n c e l e v e l = 0.95 59 60 B M D = 1 4 5 . 0 2 61 62 63 B M D L = 1 1 0 . 1 6 1 64 This document is a draftfor review purposes only and does not constitute Agency policy. E-477 DRAFT--DO NOT CITE OR QUOTE 1 E .3 .4 3 .5 . Figurefor Additional Model Presented: Power Power Model with 0.95 Confidence Level Mean Response 2 19:55 02/16 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-478 DRAFT--DO NOT CITE OR QUOTE 1 E.3.44. Sm ialow icz et al., 2008: PFC per Spleen 2 E .3 .4 4 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M2) 3 0.133 377.395 1.320E+02 8.431E+01 Notes exponential (M3) 3 0.133 37 7 .3 9 5 1 .320E + 02 8 .4 3 1 E + 0 1 p o w er hit boun d (d = 1) exponential (M4) 3 0.133 377.395 1.320E+02 8.184E+01 exponential (M5) 2 0.061 37 9 .3 9 5 1 .320E + 02 8 .1 8 4 E + 0 1 p o w er hit boun d (d = 1) H ill 2 0.069 379.150 1.401E+02 error n lo w er b ound hit (n = 1) linear polynom ial, 4degree power c 3 0.044 379.895 2.151E+02 1.704E+02 3 0.044 379.895 2.151E+02 1.704E+02 3 0.044 37 9 .8 9 5 2 .1 5 1 E + 0 2 1 .704E + 02 p o w er bound hit (p ow er = 1) H ill, unrestricted 2 <0.0001 441.885 7.545E -23 error unrestricted (n = 0.038) power, unrestricted b 2 0.230 376.738 9.374E+01 2.088E+01 unrestricted (power = 0.418) a N on-con stant variance m odel selected (p = 0 .0 0 1 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3.44.2. Outputfor Selected Model: Power, Unrestricted 6 Sm ialow icz et al., 2008: PFC per Spleen 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 6 1 _ S m i a l _ 2 0 0 8 _ P F C s p l e e n _ P w r _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 6 1 _ S m i a l _ 2 0 0 8 _ P F C s p l e e n _ P w r _ U _ 1 . p l t 13 _ T u e F e b 16 1 9 7 5 6 : 2 6 2 0 1 0 14 15 16 A n t i R e s p o n s e t o S R B C s - P F C x 10 t o t h e 4 p e r s p l e e n , T a b l e 4 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 T h e p o w e r is n o t r e s t r i c t e d This document is a draftfor review purposes only and does not constitute Agency policy. E-479 DRAFT--DO NOT CITE OR QUOTE 1 The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i)) rho) 2 3 Total number of dose groups = 5 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 12 lalpha = 4.76607 13 rho = 0 14 control = 27.8 15 slope = -7.21601 16 power = 0.213905 17 18 19 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 20 21 lalpha rho control slope power 22 23 l a l p h a 1 -0.98 0.25 -0.27 -0.23 24 25 rho -0.98 1 -0.31 0.28 0.23 26 27 c o n t r o l 0.25 -0.31 1 -0.81 -0.74 28 29 slope -0.27 0.28 -0.81 1 0.99 30 31 power -0.23 0.23 -0.74 0.99 1 32 33 34 35 P a r a m e t e r E s t i m a t e s 36 37 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 38 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 39 lalpha 0.747155 1.0244 -1.26063 2.75494 40 rho 1.36972 0.357098 0.66982 2.06962 41 control 25.1733 2.93169 19.4273 30.9193 42 slope -1.98465 1.82113 -5.554 1.5847 43 power 0.417867 0.141932 0.139686 0.696048 44 45 46 47 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 48 49 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 50 51 52 0 15 27.8 25.2 13.4 13.2 0.769 53 1.07 14 21 23.1 13.6 12.5 -0.639 54 10.7 15 17.6 19.8 9.4 11.2 -0.768 55 107 15 12.6 11.2 8.7 7.59 0.721 56 321 8 3 3.04 3.1 3.11 -0.0353 57 58 59 60 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 61 62 63 M o d e l A1: Yij = Mu(i) + e(ij) 64 V a r j e i ;ij)} = S i g m a ^ 2 65 66 M o d e l A2: Yij = Mu(i) + e(ij) 67 V a r j e i :ij)} = S i g m a ( i ) ^2 68 69 M o d e l A3: Yij = Mu(i) + e(ij) 70 V a r j e i :ij)} = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) This document is a draftfor review purposes only and does not constitute Agency policy. E-480 DRAFT--DO NOT CITE OR QUOTE 1 Model A3 uses any fixed variance parameters that 2 were specified by the user 3 4 M o d e l R: Yi = M u + e(i) 5 Var{e(i)} = Sigma^2 6 7 8 Likelihoods of Interest 9 10 Model Log(likelihood) # Param's AIC 11 A1 -190.565019 6 393.130038 12 A2 -181.476284 10 3 8 2.952569 13 A3 -181.900030 7 377.800059 14 fitted -183.369059 5 376.738118 15 R -204.636496 2 413.272993 16 17 18 E x p l a n a t i o n o f T e s t s 19 20 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 21 (A2 vs. R) 22 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 23 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 24 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 25 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 26 27 T e s t s o f I n t e r e s t 28 29 Test -2*log(Likelihood Ratio) Test df p-value 30 31 T e s t 1 46.3204 8 <.0001 32 T e s t 2 18.1775 4 0.001139 33 T e s t 3 0.84749 3 0.8381 34 T e s t 4 2.93806 2 0.2301 35 36 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 37 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 38 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 39 40 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 41 m o d e l a p p e a r s t o b e a p p r o p r i a t e 42 43 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 44 t o b e a p p r o p r i a t e h e r e 45 46 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 47 t o a d e q u a t e l y d e s c r i b e t h e d a t a 48 49 50 B e n c h m a r k D o s e C o m p u t a t i o n 51 52 S p e c i f i e d e f f e c t = 1 53 54 R i s k T y p e = Estimated standard deviations from the control mean 55 56 C o n f i d e n c e l e v e l = 0.95 57 58 B M D = 9 3 . 7 4 1 6 59 60 61 B M D L = 2 0 . 8 7 5 8 62 This document is a draftfor review purposes only and does not constitute Agency policy. E-481 DRAFT--DO NOT CITE OR QUOTE 1 E .3.44.3. Figurefor Selected Model: Power, Unrestricted Power Model with 0.95 Confidence Level Mean Response 2 19:56 02/16 2010 3 dose 4 5 E .3 .4 4 .4 . Outputfor Additional Model Presented: Power 6 Sm ialow icz et al., 2008: PFC per Spleen 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 6 1 _ S m i a l _ 2 0 0 8 _ P F C s p l e e n _ P w r _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 6 1 _ S m i a l _ 2 0 0 8 _ P F C s p l e e n _ P w r _ 1 . p l t 13 T u e F e b 16 1 9 : 5 6 : 2 5 2 0 1 0 14 15 16 A n t i R e s p o n s e t o S R B C s - P F C x 10 t o t h e 4 p e r s p l e e n , T a b l e 4 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 T h e p o w e r is r e s t r i c t e d t o b e g r e a t e r t h a n o r e q u a l t o 1 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-482 DRAFT--DO NOT CITE OR QUOTE 1 Total number of dose groups = 5 2 Total number of records with missing values = 0 3 Maximum number of iterations = 250 4 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 5 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 6 7 8 9 Default Initial Parameter Values 10 lalpha = 4.76607 11 rho = 0 12 control = 27.8 13 slope = -54.5244 14 power = -0.136501 15 16 17 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 18 19 ( * * * T h e m o d e l p a r a m e t e r ( s ) - p o w e r 20 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 21 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 22 23 lalpha rho control slope 24 CO 'tr o CO 'tr o 25 l a l p h a 1 -0.98 0.16 26 27 rho -0.98 1 -0.25 0.54 28 o CO CO o CO CO 29 c o n t r o l 0.16 -0.25 1 30 31 s l o p e 0.54 1 32 33 34 35 P a r a m e t e r E s t i m a t e s 36 37 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 38 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limi 39 lalpha 0.474614 1.09569 -1.6729 2.62213 40 rho 1.48709 0.385029 0.732449 2.24173 41 control 21.3571 1.69233 18.0402 24.674 42 slope -0 .0574184 0.00632057 -0.0698064 -0.0450303 43 p o w e r 1 NA 44 45 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 46 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 47 h a s n o s t a n d a r d e r r o r . 48 49 50 51 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 52 53 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 54 55 56 0 15 27.8 21.4 13.4 12.3 2.02 57 1.07 14 21 21.3 13.6 12.3 -0.0898 58 10.7 15 17.6 20.7 9.4 12.1 -1.01 Lf) o \--1 59 107 15 12.6 15.2 8.7 9.6 60 321 8 3 2.93 3.1 2.82 0.0745 61 62 63 64 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 65 66 67 M o d e l A1: Yij Mu(i) + e(ij) 68 Var{e(ij)} Sigma^2 69 70 M o d e l A2: Yij = Mu(i) + e(ij) This document is a draftfor review purposes only and does not constitute Agency policy. E-483 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma(i)^2 2 3 Model A3: Yij = Mu(i) + e(ij) 4 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 5 Model A3 uses any fixed variance parameters that 6 were specified by the user 7 8 M o d e l R: Yi = M u + e(i) 9 Var{e(i)} = Sigma^2 10 11 12 L i k e l i h o o d s o f I n t e r e s t 13 14 Model Log(likelihood) # Param's AIC 15 A1 -190.565019 6 393.130038 16 A2 -181.476284 10 3 8 2.952569 17 A3 -181.900030 7 377.800059 18 fitted -185.947278 4 379.894555 19 R -204.636496 2 413.272993 20 21 22 E x p l a n a t i o n o f T e s t s 23 24 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 25 (A2 vs. R) 26 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 27 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 28 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 29 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 30 31 T e s t s o f I n t e r e s t 32 33 Test -2*log(Likelihood Ratio) Test df p-value 34 35 T e s t 1 46.3204 8 <.0001 36 T e s t 2 18.1775 4 0.001139 37 T e s t 3 0.84749 3 0.8381 38 T e s t 4 8.0945 3 0.0441 39 40 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 41 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 42 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 43 44 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 45 m o d e l a p p e a r s t o b e a p p r o p r i a t e 46 47 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 48 t o b e a p p r o p r i a t e h e r e 49 50 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 51 m o d e l 52 53 54 B e n c h m a r k D o s e C o m p u t a t i o n 55 56 S p e c i f i e d e f f e c t = 1 57 58 R i s k T y p e = Estimated standard deviations from the control mean 59 60 C o n f i d e n c e l e v e l = 0.95 61 62 B M D = 2 1 5 . 0 7 3 63 64 65 B M D L = 1 7 0 . 4 1 2 66 This document is a draftfor review purposes only and does not constitute Agency policy. E-484 DRAFT--DO NOT CITE OR QUOTE 1 E .3 .4 4 .5 . Figurefor Additional Model Presented: Power Power Model with 0.95 Confidence Level Mean Response 2 19:56 02/16 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-485 DRAFT--DO NOT CITE OR QUOTE 1 E .3.45. Toth et al., 1979: A m yloidosis 2 E .3 .4 5 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes gamma 2 0.022 1 50.666 2 .2 9 6 E + 0 2 1 .460E + 02 p o w er b oun d hit (p ow er = 1) logistic negative intercept (intercept = - 2 0.013 152.187 4.088E+02 3.125E+02 2.098) log-logistic a 2 0.028 149.984 1.759E+02 9.729E+01 slope bound hit (slope = 1) log-probit 2 0.007 1 53.479 4 .4 0 2 E + 0 2 2 .9 6 5 E + 0 2 slop e b ound hit (slop e = 1) m ultistage, 3degree probit 2 0.022 150.666 2.296E+02 1.460E+02 final h = 0 negative intercept (intercept = - 2 0.014 152.040 3.846E+02 2.911E+02 1.238) W eibull 2 0.022 1 50.666 2 .2 9 6 E + 0 2 1 .460E + 02 p o w er b oun d hit (p ow er = 1) gamma, unrestricted log-logistic, unrestricted b log-probit, unrestricted W eibull, unrestricted 2 0.917 140.208 7.687E-01 7.637E -04 unrestricted (power = 0.187) 2 0.847 140.370 8.465E-01 1.565E-03 unrestricted (slope = 0.238) 2 0.811 140.458 8.545E-01 2.334E-03 unrestricted (slope = 0.135) 2 0.882 140.287 8.179E-01 1.140E-03 unrestricted (power = 0.212) a Best-fitting m odel, BM D S output presented in this appendix b Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .4 5 .2 . Outputfor Selected Model: Log-Logistic 6 Toth et al., 1979: A m yloidosis 7 8 9 10 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 6 2 _ T o t h _ 1 9 7 9 _ A m y 1 y r _ L o g L o g i s t i c _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 6 2 _ T o t h _ 1 9 7 9 _ A m y 1 y r _ L o g L o g i s t i c _ 1 . p l t 13 T u e F e b 16 1 9 : 5 6 : 5 9 2 0 1 0 14 15 16 T a b l e 2 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e This document is a draftfor review purposes only and does not constitute Agency policy. E -486 D RAFT-- DO N O T CITE OR QUOTE 1 Slope pa r a m e t e r is r estricted as slope >= 1 2 3 Total number of observations = 4 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 12 13 14 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 15 background = 0 16 intercept = -6.90711 17 slope = 1 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( * * * T h e m o d e l p a r a m e t e r ( s ) - s l o p e 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 24 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 25 26 background intercept 27 28 b a c k g r o u n d 1 -0.47 29 30 i n t e r c e p t -0.47 1 31 32 33 34 P a r a m e t e r E s t i m a t e s 35 36 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 37 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf 38 b a c k g r o u n d 0 . 0 8 4 8 9 8 4 * * * 39 intercept -7.36716 * * * 40 s l o p e 1 41 42 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 43 44 45 46 A n a l y s i s o f D e v i a n c e T a b l e 47 48 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 49 F u l l m o d e l -68.017 4 50 F i t t e d m o d e l -72.9918 2 9.9496 2 0.00691 51 R e d u c e d m o d e l -82.0119 1 27.99 3 <.0001 52 53 AIC: 149.984 54 55 56 G o o d n e s s o f F i t 57 58 Dose Est. Prob. Expected Observed Size Residual 59 60 0.0000 0.0849 3.226 0.000 38 -1.878 61 1.0000 0.0855 3.761 5.000 44 0 . 6 6 8 62 100.0000 0.1393 6.128 10.000 44 1 . 6 8 6 63 1 0 0 0 . 0 0 0 0 0.4392 18.884 17.000 43 -0.579 64 65 C h i ^ 2 = 7 . 1 5 d.f. = 2 P-value = 0.0280 66 67 68 B e n c h m a r k D o s e C o m p u t a t i o n 69 70 S p e c i f i e d e f f e c t = 0.1 This document is a draftfor review purposes only and does not constitute Agency policy. E-487 DRAFT--DO NOT CITE OR QUOTE 1 2 Risk Type 3 4 Confidence level 5 6 BMD 7 8 BMDL 9 10 Extra risk 0.95 175.903 97.2899 11 E .3 .4 5 .3 . Figurefor Selected Model: Log-Logistic Log-Logistic Model with 0.95 Confidence Level 12 13 14 15 E .3.45.4. Outputfor Additional Model Presented: Log-Logistic, Unrestricted 16 Toth et al., 1979: A m y lo id o sis 17 18 19 20 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 21 I n p u t D a t a F i l e : C : \ 1 \ 6 2 _ T o t h _ 1 9 7 9 _ A m y 1 y r _ L o g L o g i s t i c _ U _ 1 . ( d ) 22 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 6 2 _ T o t h _ 1 9 7 9 _ A m y 1 y r _ L o g L o g i s t i c _ U _ 1 . p l t 23 ~ T u e F e b 16 1 9 : 5 7 : 0 0 ~ 2 0 1 0 24 25 26 T a b l e 2 27 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-488 DRAFT--DO NOT CITE OR QUOTE 1 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 2 3 P[response] = background+(1-background)/[1+EXP(-intercept-slope*Log(dose))] 4 5 6 Dependent variable = DichEff 7 Independent variable = Dose 8 Slope pa r a m e t e r is not restricted 9 10 T o t a l n u m b e r o f o b s e r v a t i o n s = 4 11 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 12 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 13 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 16 17 18 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 19 20 21 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 22 background = 0 23 intercept = -2.10894 24 slope = 0.227921 25 26 27 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 28 29 ( * * * T h e m o d e l p a r a m e t e r ( s ) - b a c k g r o u n d 30 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 31 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 32 33 intercept slope 34 g CO o g CO o 35 i n t e r c e p t 1 36 37 s l o p e 1 38 39 40 41 P a r a m e t e r E s t i m a t e s 42 43 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 44 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf 45 b a c k g r o u n d 0* * * 46 intercept -2.15753 * * * 47 slope 0.238304 48 49 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 50 51 52 53 A n a l y s i s o f D e v i a n c e T a b l e 54 55 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 56 F u l l m o d e l -68.017 4 57 F i t t e d m o d e l -68.1848 2 0.33571 2 0.8455 58 R e d u c e d m o d e l -82.0119 1 27.99 3 <.0001 59 60 AIC: 140.37 61 62 63 G o o d n e s s o f F i t 64 65 Dose Est. Prob. Expected Observed Size Residu. 66 67 0.0000 0.0000 0.000 0.000 38 0.000 68 1.0000 0.1036 4.560 5.000 44 0 . 2 1 8 69 100.0000 0.2573 11.321 10.000 44 - 0 . 4 5 6 70 1 0 0 0 . 0 0 0 0 0.3749 16.119 17.000 43 0.277 This document is a draftfor review purposes only and does not constitute Agency policy. E-489 DRAFT--DO NOT CITE OR QUOTE 1 2 Chi^2 = 0.33 d.f. = 2 P-value = 0.8471 3 4 5 Benchmark Dose Computation 6 7 Specified effect : 0.1 8 9 Risk Type Extra risk 10 11 C o n f i d e n c e l e v e l : 0.95 12 13 BMD ^ 0.846547 14 15 BMDL = 0.00156534 16 17 18 E .3.45.5. Figurefor Additional Model Presented: Log-Logistic, Unrestricted Log-Logistic Model with 0.95 Confidence Level 19 19:57 02/16 2010 20 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-490 DRAFT--DO NOT CITE OR QUOTE 1 E .3.46. Toth et al., 1979: Skin L esions 2 E .3 .4 6 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Notes gamma 2 0.009 159.223 1 .181E + 02 8.3 0 8 E + 0 1 p o w er b oun d hit (p ow er = 1) logistic a 2 0.002 162.974 2.709E+02 2.147E+02 negative intercept (intercept = -2.098) log-logistic 2 0.029 1 56.567 6 .7 5 0 E + 0 1 4 .0 5 7 E + 0 1 slop e b ound hit (slop e = 1) log-probit 2 0.001 1 64.598 2 .4 4 6 E + 0 2 1 .626E + 02 slop e b ound hit (slop e = 1) m ultistage, 3degree probit 2 0.009 159.223 1.181E+02 8.308E+01 final B = 0 negative intercept (intercept = - 2 0.003 162.684 2.522E+02 2.015E+02 1.238) W eibull 2 0.009 159.223 1 .181E + 02 8.3 0 8 E + 0 1 p o w er b oun d hit (p ow er = 1) gamma, unrestricted log-logistic, unrestricted b log-probit, unrestricted W eibull, unrestricted 2 0.882 147.287 error error unrestricted (power = 0.251) 2 0.630 147.969 1.137E+00 5.477E -02 unrestricted (slope = 0.351) 2 0.558 148.218 1.096E+00 6.847E -02 unrestricted (slope = 0.202) 2 0.762 147.581 1.077E+00 4.080E -02 unrestricted (power = 0.3) a Best-fitting m odel, BM D S output presented in this appendix b Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .4 6 .2 . Outputfor Selected Model: Logistic 6 Toth et al., 1979: Skin L esions 7 8 9 10 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 6 3 _ T o t h _ 1 9 7 9 _ S k i n L e s _ L o g i s t i c _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 6 3 _ T o t h _ 1 9 7 9 _ S k i n L e s _ L o g i s t i c _ 1 . p l t 13 ~ T u e F e b 16 1 9 : 5 7 : 2 9 2 0 1 0 14 15 16 T a b l e 2 17 18 19 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 20 21 P [ r e s p o n s e ] = 1 / [ 1 + E X P ( - i n t e r c e p t - s l o p e * d o s e ) ] 22 23 24 D e p e n d e n t v a r i a b l e = D i c h E f f 25 I n d e p e n d e n t v a r i a b l e = D o s e This document is a draftfor review purposes only and does not constitute Agency policy. E-491 D RAFT-- DO N O T CITE OR QUOTE 1 Slope pa r a m e t e r is not restricted 2 3 Total number of observations = 4 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 12 background = 0 Specified 13 intercept = -2.53484 14 slope = 0.00299511 15 16 17 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 18 19 ( * * * T h e m o d e l p a r a m e t e r ( s ) - b a c k g r o u n d 20 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 21 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 22 23 intercept slope 24 25 i n t e r c e p t 1 -0.67 26 27 slope -0.67 1 28 29 30 31 P a r a m e t e r E s t i m a t e s 32 33 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 34 Variable Estimate Std. Err. Lower Conf. Limit Uppe r Conf. Limit 35 intercept -1.91768 0.26892 -2.44475 -1.39061 36 slope 0.00230499 0.000419329 0.00148312 0.00312686 37 38 39 40 A n a l y s i s o f D e v i a n c e T a b l e 41 42 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 43 Full model -71.5177 4 44 F i t t e d m o d e l -79.487 2 15.9387 2 0.0003459 45 R e d u c e d m o d e l -95.8498 1 48.6642 3 <.0001 46 47 AIC: 162.974 48 49 50 G o o d n e s s o f F i t 51 S c a l e d 52 Dose Est. Prob. Expected Observed Size Residual 53 54 0.0000 0.1281 4.869 0.000 38 - 2 .363 55 1.0000 0.1284 5.649 5.000 44 - 0 . 2 9 2 56 100.0000 0.1561 6.870 13.000 44 2 . 5 4 6 57 1 0 0 0 . 0 0 0 0 0.5956 25.612 25.000 43 - 0 .190 58 59 C h i ^ 2 = 1 2 . 1 9 d.f. = 2 P-value = 0.0023 60 61 62 B e n c h m a r k D o s e C o m p u t a t i o n 63 64 S p e c i f i e d e f f e c t = 0.1 65 66 R i s k T y p e = Extra risk 67 68 C o n f i d e n c e l e v e l = 0.95 69 70 BMD = 270.917 This document is a draftfor review purposes only and does not constitute Agency policy. E-492 DRAFT--DO NOT CITE OR QUOTE 1 2 BMDL 214.66 3 4 5 E .3 .4 6 .3 . Figurefor Selected Model: Logistic Logistic Model with 0.95 Confidence Level 6 7 8 9 E .3.46.4. Outputfor Additional Model Presented: Log-Logistic, Unrestricted 10 Toth et al., 1979: Skin L esio n s 11 12 13 14 L o g i s t i c M o d e l . ( V e r s i o n : 2 . 1 2 ; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 15 I n p u t D a t a F i l e : C : \ 1 \ 6 3 _ T o t h _ 1 9 7 9 _ S k i n L e s _ L o g L o g i s t i c _ U _ 1 . ( d ) 16 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 6 3 _ T o t h _ 1 9 7 9 _ S k i n L e s _ L o g L o g i s t i c _ U _ 1 . p l t 17 ~ T u e F e b 16 2 0 : 0 1 : 5 6 2 0 1 0 18 19 20 T a b l e 2 21 22 23 T h e f o r m o f t h e p r o b a b i l i t y f u n c t i o n is: 24 25 P [ r e s p o n s e ] = b a c k g r o u n d + ( 1 - b a c k g r o u n d ) / [ 1 + E X P ( - i n t e r c e p t - s l o p e * L o g ( d o s e ) ) ] 26 27 28 D e p e n d e n t v a r i a b l e = D i c h E f f This document is a draftfor review purposes only and does not constitute Agency policy. E-493 DRAFT--DO NOT CITE OR QUOTE 1 Independent variable = Dose 2 Slope pa r a m e t e r is not restricted 3 4 Total number of observations = 4 5 Total number of records with missing values = 0 6 Maximum number of iterations = 250 7 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 8 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 9 10 11 12 U s e r h a s c h o s e n t h e l o g t r a n s f o r m e d m o d e l 13 14 15 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 16 background = 0 17 intercept = -2.14055 18 slope = 0.332409 19 20 21 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 22 23 ( * * * T h e m o d e l p a r a m e t e r ( s ) - b a c k g r o u n d 24 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 25 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 26 27 intercept slope 28 29 i n t e r c e p t 1 -0.9 30 31 s l o p e -0.9 1 32 33 34 35 P a r a m e t e r E s t i m a t e s 36 37 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 38 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 39 b a c k g r o u n d 0* * * 40 intercept -2.24241 * * * 41 slope 0.350932 * * * 42 43 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 44 45 46 47 A n a l y s i s o f D e v i a n c e T a b l e 48 49 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 50 Full model -71.5177 4 51 F i t t e d m o d e l -71.9844 2 0.93345 2 0.6271 52 R e d u c e d m o d e l -95.8498 1 48.6642 3 <.0001 53 54 AIC: 147.969 55 56 57 G o o d n e s s o f F i t 58 S c a l e d 59 Dose Est. Prob. Expected Observed Size Residu. 60 61 0.0000 0.0000 0.000 0.000 38 0.000 62 1.0000 0.0960 4.224 5.000 44 0 . 3 9 7 63 100.0000 0.3483 15.327 13.000 44 - 0 . 7 3 6 64 1 0 0 0 . 0 0 0 0 0.5453 23.448 25.000 43 0.475 65 66 C h i ^ 2 = 0 . 9 3 d.f. = 2 P-value = 0.6295 67 68 69 B e n c h m a r k D o s e C o m p u t a t i o n 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-494 DRAFT--DO NOT CITE OR QUOTE 1 Specified effect 2 3 Risk Type 4 5 Confidence level 6 7 BMD 8 9 BMDL 10 11 0.1 Extra risk 0.95 1.1374 0.0547689 12 E .3.46.5. Figurefor Additional Model Presented: Log-Logistic, Unrestricted Log-Logistic Model with 0.95 Confidence Level 13 20:01 02/16 2010 14 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-495 DRAFT--DO NOT CITE OR QUOTE 1 E .3.47. V an B irgelen et al., 1995a: H epatic R etinol 2 E .3 .4 7 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M 2) 4 <0.0001 164.340 2.912E +02 error Notes exponential (M 3) 4 <0.0001 164.340 2.912E +02 error p o w er hit boun d (d = 1) exponential (M4) b 3 <0.0001 148.052 1.151E+02 7.098E+01 exponential (M 5) 3 < 0 .0 0 0 1 1 4 8.052 1.151E + 02 7 .0 9 8 E + 0 1 p o w er hit boun d (d = 1) H ill 3 0.044 128.757 1.314E+01 error n lo w er bound hit (n = 1) linear 4 <0.0001 178.734 7.815E+02 5.997E+02 polynomial, 5-degree 0 N /A 283.606 2.481E +03 error power 4 < 0 .0 0 0 1 1 7 8.734 7 .8 1 5 E + 0 2 5 .9 9 7 E + 0 2 p o w er b ound hit (p ow er = 1) Hill, unrestricted 2 0.269 125.273 5.561E +00 error unrestricted (n = 0.571) power, unrestricted c 3 0.025 129.990 4.205E-01 8.504E-03 unrestricted (power = 0.118) a N on-con stant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, B M D S output also presented in this appendix 3 4 5 E .3 .4 7 .2 . Outputfor Selected Model: Exponential (M4) 6 Van B irgelen et al., 1995a: H epatic Retinol 7 8 9 10 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 6 5 _ V a n B _ 1 9 9 5 a _ H e p R e t _ E x p _ 1 . ( d ) 12 G n u p l o t P l o t t i n g Fil e : 13 T u e F e b 16 2 0 : 0 3 : 0 5 2 0 1 0 14 15 16 T b l 3 , h e p a t i c r e t i n o l 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 N o t e : Y [ d o s e ] is t h e m e d i a n r e s p o n s e f o r e x p o s u r e = d o s e ; 26 s i g n = +1 f o r i n c r e a s i n g t r e n d i n d a t a ; This document is a draftfor review purposes only and does not constitute Agency policy. E-496 DRAFT--DO NOT CITE OR QUOTE 1 sign = -1 for d e c r e a s i n g trend. 2 3 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 4 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 5 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 6 7 8 Dependent variable = Mean 9 Independent variable = Dose 10 D a t a a r e a s s u m e d t o b e d i s t r i b u t e d : n o r m a l l y 11 V a r i a n c e M o d e l : e x p ( l n a l p h a + r h o * l n ( Y [ d o s e ] ) ) 12 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 13 14 T o t a l n u m b e r o f d o s e g r o u p s = 6 15 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 16 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 17 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 18 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 19 20 M L E s o l u t i o n p r o v i d e d : E x a c t 21 22 23 I n i t i a l P a r a m e t e r V a l u e s 24 25 Variable Model 4 26 27 lnalpha -1.16065 28 rho 1.53688 29 a 1 5 . 6 4 5 30 b 0 . 0 0 6 2 5 1 1 7 31 c 0 . 0 3 6 5 2 4 7 32 d 1 33 34 35 36 P a r a m e t e r E s t i m a t e s 37 38 Variable Model 4 39 40 lnalpha -0.882225 41 rho 1.82707 42 a 1 0 . 5 2 9 4 43 b 0 . 0 0 7 2 0 3 4 6 44 c 0 . 0 6 8 8 6 6 1 45 d 1 46 47 48 T a b l e o f S t a t s F r o m I n p u t D a t a 49 50 D o s e Obs Mean Obs Std Dev 51 52 08 14.9 8.768 53 14 8 CO 3.394 54 26 8 8.2 2.263 55 47 8 5.1 0.8485 56 320 8 2.2 0.8485 57 1024 8 0.6 0.5657 58 59 60 E s t i m a t e d V a l u e s o f I n t e r e s t 61 62 Dose Est Mean Est Std Scaled Resid 63 64 0 10.53 5.526 2.237 65 14 9.589 5.073 -0.6628 66 26 8.855 4.717 -0.3926 67 47 7.714 4.159 -1.778 68 320 1.703 1.046 1.343 69 1024 0.7313 0.4833 -0.7681 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-497 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Other models for which likelihoods are calculated: 4 5 Model A1: Yij = Mu(i) + e(ij) 6 V a r { e i ;ij)} = S i g m a ^ 2 7 8 Model A2: Yij = Mu(i) + e(ij) 9 V a r { e i :ij)} = S i g m a ( i ) ^ 2 10 11 M o d e l A3: Yij = M u ( i ) + e(i j ) 12 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 13 14 M o d e l R: Yij = M u + e(i) 15 V a r { e i ij)} = S i g m a ^ 2 16 17 18 L i k e l i h o o d s o f I n t e r e s t 19 20 Model Log(likelihood) DF AIC 21 22 A1 -87.1567 7 188.3134 23 A2 -47.28742 12 11 8 . 5 7 4 8 24 A3 -55.32422 8 126.6484 25 R -109.967 2 223.934 26 4 -69.02619 5 148.0524 27 28 29 Additive constant for all log-likelihoods = -44.11. This constant added to the 30 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s n o t 31 d e p e n d o n t h e m o d e l p a r a m e t e r s . 32 33 34 E x p l a n a t i o n o f T e s t s 35 36 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 37 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 38 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 39 40 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 41 42 43 T e s t s o f I n t e r e s t 44 45 T e s t -2*log(Likelihood Ratio) D. F. p-value 46 47 T e s t 1 125.4 10 < 0.0001 48 T e s t 2 79.74 5 < 0.0001 49 T e s t 3 16.07 4 0.002922 50 T e s t 6a 27.4 3 < 0.0001 51 52 53 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 54 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e 55 l e v e l s , it s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a . 56 57 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 58 v a r i a n c e m o d e l a p p e a r s t o b e a p p r o p r i a t e . 59 60 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t to 61 c o n s i d e r a d i f f e r e n t v a r i a n c e m o d e l . 62 63 T h e p - v a l u e f o r T e s t 6a is l e s s t h a n .1. M o d e l 4 m a y n o t a d e q u a t e l y 64 d e s c r i b e t h e d a t a ; y o u m a y w a n t t o c o n s i d e r a n o t h e r m o d e l . 65 66 67 B e n c h m a r k D o s e C o m p u t a t i o n s : 68 69 S p e c i f i e d E f f e c t = 1 . 0 0 0 0 0 0 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-498 DRAFT--DO NOT CITE OR QUOTE 1 Risk Type Estimated standard deviations from control 2 3 Confidence Level = 0.950000 4 5 BMD = 115.128 6 7 BMDL = 70.981 8 9 10 E .3 .4 7 .3 . Figurefor Selected Model: Exponential (M4) Exponential_beta Model 4 with 0.95 Confidence Level Mean Response 11 12 13 14 E .3.47.4. Outputfor Additional Model Presented: Power, Unrestricted 15 V an B irg elen et al., 1995a: H epatic R etinol 16 17 18 19 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 20 I n p u t D a t a F i l e : C : \ 1 \ 6 5 _ V a n B _ 1 9 9 5 a _ H e p R e t _ P w r _ U _ 1 . ( d ) 21 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 6 5 _ V a n B _ 1 9 9 5 a _ H e p R e t _ P w r _ U _ 1 . p l t 22 _ T u e F e b 1 6 ~ 2 0 : 0 3 : 1 1 2 0 1 0 23 24 25 T b l 3 , h e p a t i c r e t i n o l 26 27 28 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: This document is a draftfor review purposes only and does not constitute Agency policy. E-499 DRAFT--DO NOT CITE OR QUOTE 1 2 Y[dose] = control + slope * doseApower 3 4 5 Dependent variable = Mean 6 Independent variable = Dose 7 The power is not restricted 8 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) * rho) 9 10 T o t a l n u m b e r o f d o s e g r o u p s = 6 11 T o t a l n u m b e r o f r e c o r d s w i t h m i s s i n g v a l u e s = 0 12 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 13 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 15 16 17 18 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 19 lalpha = 2.76506 20 rho = 0 21 control = 14.9 22 slope = -3.78637 23 power = 0.191713 24 25 26 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 27 28 lalpha rho control slope power 29 30 l a l p h a 1 -0.8 -0.047 0.042 0.065 31 32 rho -0.8 1 -0.085 -0.0029 -0.11 33 34 c o n t r o l -0.047 -0.085 1 -0.95 -0.81 35 36 slope 0.042 -0.0029 -0.95 1 0.96 37 38 p o w e r 0 . 0 6 5 - 0 . 1 1 - 0 . 8 1 0.96 1 39 40 41 42 P a r a m e t e r E s t i m a t e s 43 44 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 45 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 46 lalpha -1.02622 0.389164 -1.78897 -0.263475 47 rho 1.68421 0.199212 1.29376 2.07466 48 control 16.9577 2.21133 12.6235 21.2918 49 slope -7.19097 1.99708 -11.1052 -3.27676 50 power 0.117935 0.0225396 0.0737578 0.162111 51 52 53 54 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 55 56 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 57 58 59 08 14.9 17 8.77 6.49 - 0 . 8 9 6 60 14 8 8.4 7.14 3.39 3.13 1.14 61 26 8 8.2 6.4 2.26 2.86 1.78 62 47 8 5.1 5.63 0.849 2.57 -0.588 63 320 8 2.2 2.76 0.849 1.41 -1.12 64 1 0 2 4 8 0.6 0.672 0.566 0.428 -0.475 65 66 67 68 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-500 DRAFT--DO NOT CITE OR QUOTE 1 Model A1: Yij M u ( i ) + e(i j ) 2 Var{e(ij)} S i g m a ^ 2 3 4 Model A2: Yij Mu(i) + e(ij) 5 Var{e(ij)} Sigma(i)^2 6 7 Model A3: Yij = Mu(i) + e(ij) 8 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 9 Model A3 uses any fixed variance parameters that 10 w e r e s p e c i f i e d b y t h e u s e r 11 12 M o d e l R: Yi = M u + e(i) 13 V a r { e ( i ) } = S i g m a ^ 2 14 15 16 L i k e l i h o o d s o f I n t e r e s t 17 18 Model Log(likelihood) # Param's AIC 19 A1 -87.156698 7 188.313395 20 A2 -47.287416 12 1 1 8 . 5 7 4 8 3 3 21 A3 -55.324218 8 126.648436 22 fitted -59.994980 5 129.989960 23 R -109.967018 2 223.934036 24 25 26 E x p l a n a t i o n o f T e s t s 27 28 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 29 (A2 vs. R) 30 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 31 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 32 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 33 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 34 35 T e s t s o f I n t e r e s t 36 37 Test -2*log(Likelihood Ratio) Test df p-value 38 39 T e s t 1 125.359 10 <.0001 40 T e s t 2 79.7386 5 <.0001 41 T e s t 3 16.0736 4 0.002922 42 T e s t 4 9.34152 3 0.02508 43 44 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 45 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 46 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 47 48 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 49 m o d e l a p p e a r s t o b e a p p r o p r i a t e 50 51 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t t o c o n s i d e r a 52 d i f f e r e n t v a r i a n c e m o d e l 53 54 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 55 m o d e l 56 57 58 B e n c h m a r k D o s e C o m p u t a t i o n 59 60 S p e c i f i e d e f f e c t = 1 61 62 R i s k T y p e = Estimated standard deviations from the control mean 63 64 C o n f i d e n c e l e v e l = 0.95 65 66 B M D = 0 . 4 2 0 4 7 5 67 68 69 B M D L = 0 . 0 0 8 5 0 4 2 2 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-501 DRAFT--DO NOT CITE OR QUOTE 1 2 E .3.47.5. Figurefor Additional Model Presented: Power, Unrestricted Power Model with 0.95 Confidence Level Mean Response 3 20:03 02/16 2010 4 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-502 DRAFT--DO NOT CITE OR QUOTE 1 E.3.48. Van Birgelen et al., 1995a: Hepatic Retinol Palmitate 2 E.3.48.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M2) 4 <0.0001 467.446 error error Notes exponential (M3) 4 <0.0001 467.446 error error p o w er hit b ound (d = 1) exponential (M4) 3 <0.0001 454.087 error error exponential (M5) 3 <0.0001 454.087 error error p o w er hit b ound (d = 1) H ill 3 <0.0001 563.579 error error linear b polynom ial, 5degree power 4 <0.0001 488.446 1.420E+03 9.889E+02 0 N /A 573.977 error error 4 < 0 .0 0 0 1 4 8 8 .4 4 6 1.420E +03 9 .8 8 9 E + 0 2 p o w er b oun d hit (p ow er = 1) H ill, unrestricted 3 <0.0001 522.322 2.418E -12 2.418E -12 unrestricted (n = 0.452) power, unrestricted c 3 0.348 408.062 3.765E -02 1.208E-05 unrestricted (power = 0.054) a N on-con stant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .4 8 .2 . Outputfor Selected Model: Linear 6 Van B irgelen et al., 1995a: H epatic R etinol Palmitate 7 8 9 10 P o l y n o m i a l M o d e l . ( V e r s i o n : 2 . 1 3 ; D a t e : 0 4 / 0 8 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 6 6 _ V a n B _ 1 9 9 5 a _ H e p R e t P a l m _ L i n e a r _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 6 6 _ V a n B _ 1 9 9 5 a _ H e p R e t P a l m _ L i n e a r _ 1 . p l t 13 _ T u e F e b 1 6 ~ 2 0 : 0 3 : i 6 2 0 1 0 14 15 16 T b l 3 , h e p a t i c r e t i n o l p a l m i t a t e 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 S i g n s o f t h e p o l y n o m i a l c o e f f i c i e n t s a r e n o t r e s t r i c t e d This document is a draftfor review purposes only and does not constitute Agency policy. E-503 DRAFT--DO NOT CITE OR QUOTE 1 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) * rho) 2 3 Total number of dose groups = 6 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 12 lalpha = 9.57332 13 rho = 0 14 beta_0 = 177.506 15 b e t a 1 = - 0 . 2 0 4 7 7 5 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 19 20 lalpha rho beta_0 beta_1 21 Lf) g o Lf) g o 22 l a l p h a 1 -0.017 0.022 23 24 rh o 1 0.00019 -0.0048 25 26 beta 0 -0.017 0.00019 1 -1 27 28 b e t a 1 0.022 -0.0048 -1 1 29 30 31 32 P a r a m e t e r E s t i m a t e s 33 34 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 35 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 36 lalpha -0.723216 0.638291 -1.97424 0.527811 37 rho 2.26615 0.140196 1.99137 2.54093 38 beta_0 150.535 31.5457 88.7064 212.363 39 beta_1 -0.143931 0.0308317 -0.20436 -0.0835018 40 41 42 43 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 44 45 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res. 46 47 48 0 8 472 1 5 1 2 7 2 204 4.45 49 14 8 94 149 67.9 201 - 0.766 50 26 8 107 147 76.4 199 - 0.567 51 47 8 74 144 39.6 194 -1.02 52 320 8 22 104 22.6 135 -1.73 53 1 0 2 4 8 3 3.15 2.83 2.56 - 0.166 54 55 56 57 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 58 59 60 M o d e l A1: Yij Mu(i) + e(ij) 61 Var{e(ij)} Sigma^2 62 63 M o d e l A2: Yij Mu(i) + e(ij) 64 Var{e(ij)} Sigma(i)^2 65 66 M o d e l A3: Yij = Mu(i) + e(ij) 67 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 68 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 69 w e r e s p e c i f i e d b y t h e u s e r 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-504 DRAFT--DO NOT CITE OR QUOTE 1 Model R: Yi = Mu + e(i) 2 Var{e(i)} = Sigma^2 3 4 5 Likelihoods of Interest 6 7 Model Log(likelihood) # Param's AIC 8 A1 -250.554817 7 515.109634 9 A2 -196.755746 12 4 1 7 . 5 1 1 4 9 1 10 A3 -197.383174 8 410.766347 11 fitted -240.223107 4 488.446215 12 R -276.789644 2 557.579287 13 14 15 E x p l a n a t i o n o f T e s t s 16 17 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 18 (A2 vs. R) 19 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 20 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 21 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 22 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 23 24 T e s t s o f I n t e r e s t 25 26 Test -2*log(Likelihood Ratio) Test df p-value 27 28 T e s t 1 160.068 10 <.0001 29 T e s t 2 107.598 5 <.0001 30 T e s t 3 1.25486 4 0.869 31 T e s t 4 85.6799 4 <.0001 32 33 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 34 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 35 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 36 37 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 38 m o d e l a p p e a r s t o b e a p p r o p r i a t e 39 40 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 41 t o b e a p p r o p r i a t e h e r e 42 43 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 44 m o d e l 45 46 47 B e n c h m a r k D o s e C o m p u t a t i o n 48 49 S p e c i f i e d e f f e c t = 1 50 51 R i s k T y p e = Estimated standard deviations from the control mean 52 53 C o n f i d e n c e l e v e l = 0.95 54 55 BMD = 1419.81 56 57 58 BMDL = 988.945 59 This document is a draftfor review purposes only and does not constitute Agency policy. E-505 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E.3.48.3. F igure f o r Selected M odel: L in ea r Linear Model with 0.95 Confidence Level 2 3 4 5 E .3.48.4. Outputfor Additional Model Presented: Power, Unrestricted 6 Van B irgelen et al., 1995a: H epatic R etinol Palmitate 7 8 9 10 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 6 6 _ V a n B _ 1 9 9 5 a _ H e p R e t P a l m _ P w r _ U _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 6 6 _ V a n B _ 1 9 9 5 a _ H e p R e t P a l m _ P w r _ U _ 1 . p l t 13 _ T u e F e b 1 6 _ 2 0 : 0 3 i 5 0 2 0 1 0 14 15 16 T b l 3 , h e p a t i c r e t i n o l p a l m i t a t e 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e A p o w e r 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 T h e p o w e r is n o t r e s t r i c t e d 27 T h e v a r i a n c e is t o b e m o d e l e d as V a r ( i ) = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 28 This document is a draftfor review purposes only and does not constitute Agency policy. E-506 DRAFT--DO NOT CITE OR QUOTE 1 Total number of dose groups = 6 2 Total number of records with missing values = 0 3 Maximum number of iterations = 250 4 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 5 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 6 7 8 9 Default Initial Parameter Values 10 lalpha = 9.57332 11 rho = 0 12 control = 472 13 slope = -315.054 14 power = 0.0586881 15 16 17 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 18 19 lalpha rho control slope power 20 Lf) Gj o 21 l a l p h a 1 0.29 -0.31 -0.3 22 23 rho -0.95 1 -0.4 0.39 0.29 24 25 c o n t r o l 0.29 -0.4 1 -0.98 -0.82 26 27 slope -0.31 0.39 -0.98 1 0.91 28 29 p o w e r -0.3 0.29 -0.82 0.91 1 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 36 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 37 lalpha 0.0734958 0.849559 -1.59161 1.7386 38 rho 1.80632 0.194602 1.42491 2.18774 39 control 465.497 86.914 295.149 635.845 40 slope -318.06 82.4127 -479.586 -156.534 41 power 0.0540573 0.0117709 0.0309869 0.0771278 42 43 44 45 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 46 47 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res. 48 49 50 0 8 472 465 272 266 0.069 51 14 8 94 98.7 67.9 65.6 -0.201 52 26 8 107 86.2 76.4 58.1 1.01 53 47 8 74 73.8 39.6 50.5 0.0086 54 320 8 22 31.1 22.6 23.1 -1.11 55 1 0 2 4 8 3 2.86 2.83 2.68 0.145 56 57 58 59 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 60 61 62 M o d e l A1: Yij = Mu(i) + e(ij ) 63 V a r j e i ;i j ) } = S i g m a ^ 2 64 65 M o d e l A2: Yij = Mu(i) + e(ij ) 66 V a r j e i :i j ) } = S i g m a ( i ) ^2 67 68 M o d e l A3: Yij = Mu(i) + e(ij ) 69 V a r j e i :i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 70 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t This document is a draftfor review purposes only and does not constitute Agency policy. E-507 DRAFT--DO NOT CITE OR QUOTE 1 were specified by the user 2 3 M o d e l R: Yi = M u + e(i) 4 Var{e(i)} = Sigma^2 5 6 7 Likelihoods of Interest 8 9 Model Log(likelihood) # Param's AIC 10 A1 -250.554817 7 515.109634 11 A2 -196.755746 12 4 1 7 . 5 1 1 4 9 1 12 A3 -197.383174 8 410.766347 13 fitted -199.031154 5 408.062307 14 R -276.789644 2 557.579287 15 16 17 E x p l a n a t i o n o f T e s t s 18 19 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 20 (A2 vs. R) 21 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 22 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 23 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 24 (No t e W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e 25 26 T e s t s o f I n t e r e s t 27 28 Test -2*log(Likelihood Ratio) Test df p-value 29 30 T e s t 1 160.068 10 <.0001 31 T e s t 2 107.598 5 <.0001 32 T e s t 3 1.25486 4 0.869 33 T e s t 4 3.29596 3 0.3482 34 35 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 36 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 37 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 38 39 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 40 m o d e l a p p e a r s t o b e a p p r o p r i a t e 41 42 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 43 t o b e a p p r o p r i a t e h e r e 44 45 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 46 t o a d e q u a t e l y d e s c r i b e t h e d a t a 47 48 49 B e n c h m a r k D o s e C o m p u t a t i o n 50 51 S p e c i f i e d e f f e c t = 1 52 53 R i s k T y p e = Estimated standard deviations from the control mean 54 55 C o n f i d e n c e l e v e l = 0.95 56 57 B M D = 0 . 0 3 7 6 4 8 9 58 59 60 B M D L = 1 . 2 0 7 6 9 e - 0 0 5 61 This document is a draftfor review purposes only and does not constitute Agency policy. E-508 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 E.3.48.5. F igure f o r A d d itio n a l M o d el Presented: Power, Unrestricted Power Model with 0.95 Confidence Level 2 20:03 02/16 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. E-509 DRAFT--DO NOT CITE OR QUOTE 1 E.3.49. W hite et al., 1986: C H 50 2 E .3 .4 9 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2p Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) exponential (M2) 5 0.001 391.472 4.480E+02 2.844E+02 Notes exponential (M3) 5 0.001 3 9 1 .4 7 2 4 .4 8 0 E + 0 2 2 .8 4 4 E + 0 2 p o w er hit boun d (d = 1) exponential (M4) 4 0.001 392.128 3.126E+02 1.140E+02 exponential (M5) Hill b 4 0.001 3 9 2 .1 2 8 3 .1 2 6 E + 0 2 1 .140E + 02 p o w er hit boun d (d = 1) 4 0.001 391.223 2.042E+02 3.585E+01 n lower bound hit (n = 1) linear polynom ial, 6degree power 5 <0.0001 396.430 8.065E+02 5.899E+02 3 <0.0001 643.059 9.600E +02 error 5 < 0 .0 0 0 1 3 9 6 .4 3 0 8 .0 6 5 E + 0 2 5 .8 9 9 E + 0 2 p o w er bound hit (p ow er = 1) H ill, unrestricted c 3 0.058 381.943 9.677E-01 1.900E-01 unrestricted (n = 0.211) power, unrestricted 4 0.131 379.574 7.186E-01 1.157E-02 unrestricted (power = 0.188) a N on-con stant variance m odel selected (p = 0 .0 8 7 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 E .3 .4 9 .2 . Outputfor Selected Model: Hill 6 W hite et al., 1986: CH 50 7 8 9 10 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ 7 1 _ W h i t e _ 1 9 8 6 _ C H 5 0 _ H i l l _ 1 . ( d ) 12 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 7 1 _ W h i t e _ 1 9 8 6 _ C H 5 0 _ H i l l _ 1 . p l t 13 _ T u e F e b 16 2 0 : 0 6 : 4 5 2 0 1 0 14 15 16 [ i n s e r t s t u d y n o t e s ] 17 18 19 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 20 21 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 22 23 24 D e p e n d e n t v a r i a b l e = M e a n 25 I n d e p e n d e n t v a r i a b l e = D o s e 26 P o w e r p a r a m e t e r r e s t r i c t e d t o b e g r e a t e r t h a n 1 This document is a draftfor review purposes only and does not constitute Agency policy. E-510 DRAFT--DO NOT CITE OR QUOTE 1 The variance is to be modeled as Var(i) = exp(lalpha + rho ln(mean(i))) 2 3 Total number of dose groups = 7 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 10 11 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 12 lalpha 5.60999 13 rho 0 14 intercept 91 15 v -74 16 n 0 . 0 9 6 9 9 9 8 17 k 10 18 19 20 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 21 22 ( * * * T h e m o d e l p a r a m e t e r ( s ) -n 23 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 24 a n d d o n o t a p p e a r in t h e c o r r e l a t i o n m a t r i x ) 25 26 lalpha rho intercept v k 27 28 l a l p h a 1 -0.99 0.19 0.13 -0.22 29 30 rho -0.99 1 -0.2 -0.14 0.23 31 32 i n t e r c e p t 0.19 -0.2 1 0.33 -0.7 33 CO CO o co CO o 34 v 0.13 -0.14 0.33 1 35 36 k -0.22 0.23 -0.7 1 37 38 39 40 P a r a m e t e r E s t i m a t e s 41 42 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 43 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 44 lalpha 4.34761 1.59601 1.21948 7.47574 45 rho 0.381496 0.413764 -0.429467 1.19246 46 i n t e r c e p t 71.6585 5.38454 61.105 82.212 47 v -62.7464 14.9646 -92.0765 -33.4163 48 n 1 N A 49 k 441.016 460.151 -460.864 1342.9 50 51 N A - I n d i c a t e s t h a t t h i s p a r a m e t e r h a s h i t a b o u n d 52 i m p l i e d b y s o m e i n e q u a l i t y c o n s t r a i n t a n d t h u s 53 h a s n o s t a n d a r d e r r o r . 54 55 56 57 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 58 59 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 60 61 62 0 8 91 71.7 14.1 19.9 2.75 63 10 8 54 70.3 8.49 19.8 -2.33 64 50 8 63 6 5 . 3 1 1 . 3 1 9 . 5 - 0 . 3 2 9 65 100 8 56 60.1 25.5 19.2 -0.598 66 500 8 41 38.3 17 17.6 0.43 67 1 0 0 0 8 32 28.1 17 16.6 0.661 68 2 0 0 0 8 17 20.2 17 15.6 - 0 . 5 8 9 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-511 DRAFT--DO NOT CITE OR QUOTE 1 2 Model Descriptions for likelihoods calculated 3 4 5 Model A1: Yij Mu(i) + e(ij) 6 Var{e(ij)} Sigma^2 7 8 Model A2: Yij Mu(i) + e(ij) 9 Var{e(ij)} Sigma(i)^2 10 11 M o d e l A3: Yij = Mu(i) + e(ij) 12 V a r { e ( i j ) } = e x p ( l a l p h a + r h o * l n ( M u ( i ) ) ) 13 M o d e l A 3 u s e s a n y f i x e d v a r i a n c e p a r a m e t e r s t h a t 14 w e r e s p e c i f i e d b y t h e u s e r 15 16 M o d e l R: Yi = M u + e(i) 17 V a r { e ( i ) } = S i g m a ^ 2 18 19 20 L i k e l i h o o d s o f I n t e r e s t 21 22 Model Log(likelihood) # Param's AIC 23 A1 -181.340979 8 378.681959 24 A2 -175.820265 14 3 7 9 . 6 4 0 5 2 9 25 A3 -181.238690 9 380.477380 26 fitted -190.611743 5 391.223485 27 R -212.367055 2 428.734109 28 29 30 E x p l a n a t i o n o f T e s t s 31 32 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 33 (A2 vs. R) 34 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 35 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 36 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 37 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 38 39 T e s t s o f I n t e r e s t 40 41 Test -2*log(Likelihood Ratio) Test df p-value 42 43 T e s t 1 73.0936 12 <.0001 44 T e s t 2 11.0414 6 0.0871 45 T e s t 3 10.8369 5 0.05471 46 T e s t 4 18.7461 4 0.0008815 47 48 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 49 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 50 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 51 52 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 53 m o d e l a p p e a r s t o b e a p p r o p r i a t e 54 55 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t t o c o n s i d e r a 56 d i f f e r e n t v a r i a n c e m o d e l 57 58 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 59 m o d e l 60 61 62 B e n c h m a r k D o s e C o m p u t a t i o n 63 64 S p e c i f i e d e f f e c t = 1 65 66 R i s k T y p e = Estimated standard deviations from the control mean 67 68 C o n f i d e n c e l e v e l = 0.95 69 70 BMD = 204.214 This document is a draftfor review purposes only and does not constitute Agency policy. E-512 DRAFT--DO NOT CITE OR QUOTE 1 2 BMDL 35.8504 3 4 5 E .3 .4 9 .3 . Figurefor Selected Model: Hill Hill Model with 0.95 Confidence Level Mean Response 6 7 8 9 E .3.49.4. Outputfor Additional Model Presented: Hill, Unrestricted 10 W h ite et al., 1986: C H 50 11 12 13 14 H i l l M o d e l . ( V e r s i o n : 2 . 1 4 ; D a t e : 0 6 / 2 6 / 2 0 0 8 ) 15 I n p u t D a t a F i l e : C : \ 1 \ 7 1 _ W h i t e _ 1 9 8 6 _ C H 5 0 _ H i l l _ U _ 1 . ( d ) 16 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 7 1 _ W h i t e _ 1 9 8 6 _ C H 5 0 _ H i l l _ U _ 1 . p l t 17 T u e F e b 16 2 0 : 0 6 : 4 6 2 0 1 0 18 19 20 [ i n s e r t s t u d y n o t e s ] 21 22 23 T h e f o r m o f t h e r e s p o n s e f u n c t i o n is: 24 25 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 26 27 28 D e p e n d e n t v a r i a b l e = M e a n This document is a draftfor review purposes only and does not constitute Agency policy. E-513 DRAFT--DO NOT CITE OR QUOTE 1 Independent variable = Dose 2 Power pa r a m e t e r is not restricted 3 The v ariance is to be mode l e d as Var(i) = exp(lalpha + rho * ln(mean(i))) 4 5 Total number of dose groups = 7 6 Total number of records with missing values = 0 7 Maximum number of iterations = 250 8 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 9 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 10 11 12 13 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 14 lalpha 5.60999 15 rho 0 16 intercept 91 17 v -74 18 n= 0.0969998 19 k= 10 20 21 22 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 23 24 lalpha rho intercept v n k 25 26 l a l p h a 1 -1 0.17 0.22 -0.42 -0.022 27 28 rh o -1 1 -0.17 -0.22 0.42 0.019 29 Lf) CO o CO Lf) Lf) CO oo CO Lf) o 30 i n t e r c e p t 0.17 -0.17 1 0.16 0.0069 31 32 v 0.22 -0.22 0.16 1 -0.048 -0.91 33 34 n -0.42 0.42 -0.048 1 35 36 k -0.022 0.019 0.0069 -0.91 1 37 38 39 40 P a r a m e t e r E s t i m a t e s 41 42 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 43 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 44 lalpha 6.62767 2.14235 2.42875 10.8266 45 rho -0.266376 0.555274 -1.35469 0.821941 46 i n t e r c e p t 89.579 5.61106 78.5815 100.576 47 v -458.615 402.837 -1248.16 330.93 48 n 0.210614 0.0503369 0.111956 0.309273 49 k 9.00638e+006 4.61231e+007 -8.13933e+007 9.94061e+007 50 51 52 53 T a b l e o f D a t a a n d E s t i m a t e d V a l u e s o f I n t e r e s t 54 55 D o s e N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 56 57 58 0 8 91 8 9 . 6 1 4 . 1 1 5 . 1 0.266 59 10 8 54 65.4 8.49 15.8 -2.04 60 50 8 63 5 6 . 3 1 1 . 3 1 6 . 1 1.18 61 100 8 56 51.5 25.5 16.3 0.777 62 500 8 41 37.9 17 16.9 0.516 63 1 0 0 0 8 32 30.8 17 17.4 0.191 64 2 0 0 0 8 17 22.9 17 18.1 -0. 9 2 7 65 66 67 68 M o d e l D e s c r i p t i o n s f o r l i k e l i h o o d s c a l c u l a t e d 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. E-514 DRAFT--DO NOT CITE OR QUOTE 1 Model A1: Yij = Mu(i) + e(ij) 2 Var{e(ij)} = Sigma^2 3 4 Model A2: Yij = Mu(i) + e(ij) 5 Var{e(ij)} = Sigma(i)^2 6 7 Model A3: Yij = Mu(i) + e(ij) 8 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 9 Model A3 uses any fixed variance parameters that 10 w e r e s p e c i f i e d b y t h e u s e r 11 12 M o d e l R: Yi = M u + e(i) 13 V a r { e ( i ) } = S i g m a ^ 2 14 15 16 L i k e l i h o o d s o f I n t e r e s t 17 18 Model Log(likelihood) # Param's AIC 19 A1 -181.340979 8 378.681959 20 A2 -175.820265 14 3 7 9 . 6 4 0 5 2 9 21 A3 -181.238690 9 380.477380 22 fitted -184.971691 6 381.943382 23 R -212.367055 2 428.734109 24 25 26 E x p l a n a t i o n o f T e s t s 27 28 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 29 (A2 vs. R) 30 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 31 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 32 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 33 (Note: W h e n r h o = 0 t h e r e s u l t s o f T e s t 3 a n d T e s t 2 w i l l b e t h e s a m e . ) 34 35 T e s t s o f I n t e r e s t 36 37 Test -2*log(Likelihood Ratio) Test df p-value 38 39 T e s t 1 73.0936 12 <.0001 40 T e s t 2 11.0414 6 0.0871 41 T e s t 3 10.8369 5 0.05471 42 T e s t 4 7.466 3 0.05844 43 44 T h e p - v a l u e f o r T e s t 1 is l e s s t h a n .05. T h e r e a p p e a r s t o b e a 45 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g t h e d o s e l e v e l s 46 It s e e m s a p p r o p r i a t e t o m o d e l t h e d a t a 47 48 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 49 m o d e l a p p e a r s t o b e a p p r o p r i a t e 50 51 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t t o c o n s i d e r a 52 d i f f e r e n t v a r i a n c e m o d e l 53 54 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 55 m o d e l 56 57 58 B e n c h m a r k D o s e C o m p u t a t i o n 59 60 S p e c i f i e d e f f e c t = 1 61 62 R i s k T y p e = Estimated standard deviations from the control mean 63 64 C o n f i d e n c e l e v e l = 0.95 65 66 BMD = 0.967689 67 68 BMDL = 0.189992 This document is a draftfor review purposes only and does not constitute Agency policy. E-515 DRAFT--DO NOT CITE OR QUOTE 1 E.3.49.5. F igure f o r A d d itio n a l M o d el Presented: H ill, U nrestricted Hill Model with 0.95 Confidence Level Mean Response 2 20:06 02/16 2010 3 4 5 dose This document is a draftfor review purposes only and does not constitute Agency policy. E-516 DRAFT--DO NOT CITE OR QUOTE 1 E.4. R E FE R E N C E S 2 A m in, S; M oore, RW ; Peterson, RE; et al. (2 0 0 0 ) G estational and lactational exp osure to T C D D or coplanar P C B s 3 alters adult expression o f saccharin preference behavior in fem ale rats. N eurotoxicol Teratol 2 2 (5 ):6 7 5 -6 8 2 . 4 B ell, D R ; C lode, S; Fan, M Q ; et al. (2 0 0 7 a ) T oxicity o f 2,3,7,8-tetrach lorod ib en zo-p -d ioxin in the d ev elo p in g m ale 5 W istar(H an) rat. II: Chronic d o sin g cau ses d evelopm en tal delay. T o x ic o l S ci 9 9 (1 ):2 2 4 -2 3 3 . 6 B ell, D R ; C lode, S; Fan, M Q ; et al. (2 0 0 7 b ) R elation sh ips b etw e en tissu e le v e ls o f 2,3,7,8-tetrachlorodibenzo7 p -d io x in (T C D D ), m R N A s, and to x icity in the d ev elo p in g m ale W istar (Han) rat. T o x ico l S ci 9 9 (2 ):5 9 1 -6 0 4 . 8 Cantoni, L; Salmona, M; Rizzardini, M. (1981) Porphyrogenic effect o f chronic treatment with 9 2,3,7,8-tetrachlorodibenzo-p-dioxin in fem ale rates. D ose-effect relationship follow in g urinary excretion o f 10 porphyrins. T o x ico l A pp l Pharm acol 5 7 :1 5 6 -1 5 7 . 11 Crofton, KM ; Craft, ES; H ed ge, JM; et al. (2 0 0 5 ) T hyroid-horm one-disrupting chem icals: ev id en ce for d ose12 dependent additivity or synergism . E nviron H ealth P erspect 113(11): 1549--1554. 13 D eC aprio, AP; M cM artin, D N ; O 'K e efe, PE; et al. (1 9 8 6 ) Subchronic oral to x icity o f 2,3,7,8-tetrachlorodibenzo14 p -d ioxin in the gu inea pig: com parisons w ith a P C B -con tain in g transform er flu id pyrolysate. Fund A pp l T o x ico l 15 6 :4 5 4 -4 6 3 . 16 Franc, M A ; Pohjanvirta, R; T uom isto, J; et al. (2 0 0 1 ) Persistent, lo w -d o se 2,3,7,8-tetrach lorod ib en zo-p -d ioxin 17 exposure: effec t o n aryl hydrocarbon receptor ex p ressio n in a d ioxin -resistan ce m odel. T o x ico l A ppl Pharm acol 18 1 7 5 :4 3 -5 3 . 19 H ojo, R; Stern, S; Zareba, G; et al. (2 0 0 2 ) S exu ally dim orphic beh avioral resp onses to prenatal d io x in exposure. 20 Environ Health Perspect 110(3):247-254. 21 Kattainen, H; Tuukanan, J; Sim anainen, U; et al. (2 0 0 1 ) In utero/lactational 2,3,7,8-tetrach lorod ib en zo-p -d ioxin 22 exposure impairs m olar tooth developm ent in rats. T ox ico l A ppl Pharm acol 1 7 :2 1 6 -2 2 4 . 23 K eller, JM; H uet-H udson, YM ; Leam y, LJ. (2007) Q ualitative effects o f dioxin on molars vary am ong inbred m ouse 24 strains. Arch Oral B iol 52 :4 5 0 -4 5 4 . 25 K eller, JM; Zelditch, ML; Huet, YM ; et al. (2008a) G enetic differences in sensitivity to alterations o f m andible 26 structure caused by the teratogen 2,3,7,8-tetrachlorodibenzo-p-dioxin. T oxicol Pathol 36:1006-1013. 27 Keller, JM; H uet-H udson, Y; Leam y, LJ. (2008b) Effects o f 2,3,7,8-tetrachlorodibenzo-p-dioxin on molar 28 developm ent am ong non-resistant inbred strains o f mice: a geom etric morphometric analysis. Growth D evel A ging 29 7 1 :3 -1 6 . 30 K ociba, RJ; K ey es, D G ; B eyer, JE; et al. (1 9 7 8 ) R esu lts o f a tw o -y ea r chronic to x icity and on co g en icity study o f 31 2,3,7,8-tetrach lorod ib en zo-p -d ioxin in rats. T o x ico l A pp l Pharm acol 4 6 (2 ):2 7 9 -3 0 3 . 32 Latchoum ycandane, C; Mathur, PP. (2002) E ffects o f vitam in E on reactive o x y g en species-m ediated 33 2,3,7,8-tetrach lorod i-b en zo-p -d ioxin to x icity in rat testis. J A pp l T o x ic o l 2 2 (5 ):3 4 5 -3 5 1 . 34 Li, B; Liu, H -Y; D ai, L-J; et al. (2006) The early em bryo loss caused by 2,3,7,8-tetrachlorodibenzo-p-dioxin may be 35 related to the accum ulation o f this com p oun d in the uterus. R eprod T o x ic o l 2 1 :3 0 1 -3 0 6 . 36 M arkow ski, VP; Zareba, G; Stern, S; et al. (2001) A ltered operant responding for m otor reinforcem ent and the 37 determination o f benchmark doses follow in g perinatal exposure to low -level 2,3,7,8-tetrachlorodibenzo-p-dioxin. 38 Environ Health Perspect 109(6):621-627. This document is a draftfor review purposes only and does not constitute Agency policy. E-517 DRAFT--DO NOT CITE OR QUOTE 1 Miettinen, HM; Sorvari, R; Alaluusua, S; et al. (2006) The Effect of perinatal TCDD exposure on caries 2 susceptibility in rats. Toxicol Sci 91(2):568-575. 3 NTP (National Toxicology Program). (1982) NTP Technical Report on carcinogenesis bioassay of 4 2,3,7,8-tetrachlorodibenzo-p-dioxin in Osborne-Mendel rats and B6C3F1 mice (gavage study). Public Health 5 Service, U.S. Department of Health and Human Services; NTP TR 209. Available from the National Institute of 6 Environmental Health Sciences, Research Triangle Park, NC. 7 NTP (National Toxicology Program). (2006) NTP technical report on the toxicology and carcinogenesis studies of 8 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) (CAS No. 1746-01-6) in female Harlan Sprague-Dawley rats (Gavage 9 Studies). Natl Toxicol ProgramTech Rep 521. Public Health Service, National Institute of Health, U.S. Department 10 of Health and Human Services, Research Triangle Park, NC. 11 Ohsako, S; Miyabara, Y; Nishimura, N; et al. (2001) Maternal exposure to a low dose of 12 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) suppressed the development of reproductive organs of male rats: dose13 dependent increase of mRNA levels of 5a-reductase type 2 in contrast to decrease of androgren receptor in the 14 pubertal ventral prostate. Toxicol Sci 60:132-143. 15 Shi, Z; Valdez, KE; Ying, AY; et al. (2007) Ovarian endocrine disruption underlies premature reproduction 16 senescence following environmentally relevant chronic exposure to aryl hydrocarbon receptor agonist 17 2,3,7,8-tetrachlorodibenzo-p-dioxin. Biol Reprod 30(4):293-342. 18 Smialowicz, RJ; DeVito, MJ; Williams, WC; et al. (2008) Relative potency based on hepatic enzyme induction 19 predicts immunosuppressive effects of a mixture of PCDDS/PCDFS and PCBS. Toxicol Appl Pharmacol 20 227:477-484. 21 Toth, KJ; Sugar, S; Somfai-Relle, S; et al. (1978) Carcinogenic bioassay of the herbicide 2,4,5-trichlorphenoxy 22 ethanol (TCPE) with Swiss mice. Prog Biochem Pharmacol 14:82-93. 23 Toth, L; Somfai-Relle, S; Sugar, J; et al. (1979) Carcinogenicity testing of herbicide 2,4,5-trichlorophenoxyethanol 24 containing dioxin and of pure dioxin in Swiss mice. Nature 278:548-549. 25 Van Birgelen, AP; Van der Kolk, J; Fase, KM; et al. (1995) Subchronic dose-response study of 26 2,3,7,8-tetrachlorodibenzo-p-dioxin in female Sprague-Dawley rats. Toxicol Appl Pharmacol 132:1-13. 27 White, KL, Jr; Lysy, HH; McCay, JA; et al. (1986) Modulation of serum complement levels following exposure to 28 polychlorinated dibenzo-p-dioxins. Toxicol Appl Pharmacol 84:209-219. This document is a draftfor review purposes only and does not constitute Agency policy. E-518 DRAFT--DO NOT CITE OR QUOTE [This page intentionally left blank.] DRAFT DO N O T CITE OR QUOTE May 2010 External R eview Draft APPENDIX F Cancer Benchmark Dose Modeling NOTICE THIS D O C U M E N T IS A N E X T ER N A L REVIEW DRA FT. It has not been form ally released by the U .S. Environmental Protection A gency and should not at this stage be construed to represent A gency policy. It is being circulated for com m ent on its technical accuracy and policy im plications. National Center for Environmental A ssessm ent O ffice o f Research and D evelopm ent U .S. Environmental Protection A gency Cincinnati, OH C O N TEN T S-- A PPE N D IX F: C ancer B enchm ark D ose M odeling A P P E N D IX F. C A N C E R B E N C H M A R K D O S E M O D E L I N G .............................................................F-1 F.1.B L O O D B M D S R E S U L T S ...................................................................................................................... F-1 F.1.1. K ociba et al., 1978: Stratified squam ous cell carcinoma o f hard palate or nasal turbinates..........................................................................................................F-1 F .1 .1 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s .......................................... F-1 F .1 .1 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ....................F-1 F .1 .1 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................... F-3 F .1.2. K ocib a et al., 1978: Stratified squam ous cell carcinom a o f to n g u e ...................... F -4 F .1 .2 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................... F -4 F .1 .2 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ....................F -4 F .1 .2 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................... F -6 F .1.3. K ocib a et al., 1978: A d en om a o f adrenal c o r t e x ..........................................................F -7 F .1 .3 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................... F -7 F .1 .3 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ....................F -7 F .1 .3 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................... F -9 F .1.4. K ocib a et al., 1978: H epatocellu lar adenom a(s) or carcin om a(s)....................... F -1 0 F .1 .4 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F -1 0 F .1 .4 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................. F -1 0 F .1 .4 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e .................F -1 2 F.1.5. K ociba et al., 1978: Stratified squam ous cell carcinoma o f hard palate or nasal turbinates....................................................................................................... F -13 F .1 .5 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ....................................... F -13 F .1 .5 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................. F -13 F .1 .5 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e .................F -15 F .1.6. K ocib a et al., 1978: K eratinizing squam ous cell carcinom a o f lu n g ................. F -1 6 F .1 .6 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F -1 6 F .1 .6 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................. F -1 6 F .1 .6 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e .................F -1 8 F.1.7. National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcom a.............................................................................................................................. F -1 9 F .1 .7 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ....................................... F -1 9 F .1 .7 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................. F -1 9 F .1 .7 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e .................F-21 F.1.8. National T oxicology Program, 1982: Liver: N eoplastic N odule or H epatocellu lar C a r c in o m a ................................................................................................... F -2 2 F .1 .8 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F -2 2 F .1 .8 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................. F -2 2 F .1 .8 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e .................F -2 4 F.1.9. National T oxicology Program, 1982: Adrenal: Cortical A denom a, or C arcinom a or A denom a, N O S ............................................................................................F -25 F .1 .9 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F -25 F .1 .9 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................. F -25 F .1 .9 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e .................F -2 7 This document is a draftfor review purposes only and does not constitute Agency policy. F-ii DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) F.1.10. National T oxicology Program, 1982: Thyroid: Follicular-C ell A d e n o m a ......................................................................................................................................F -2 8 F .1 .1 0 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................F -2 8 F .1 .1 0 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................F -2 8 F .1 .1 0 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ............... F -3 0 F.1.11. National T oxicology Program, 1982: Liver: N eoplastic N odule or H epatocellu lar C a r c in o m a ................................................................................................... F-31 F .1 .1 1 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F-31 F .1 .1 1 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................F-31 F .1 .1 1 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ............... F -33 F.1.12. National T oxicology Program, 1982: Thyroid: Follicular-C ell A d en om a or C a rcin o m a ........................................................................................................ F -3 4 F .1 .1 2 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................F -3 4 F .1 .1 2 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................F -3 4 F .1 .1 2 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ............... F -3 6 F .1 .1 3 . N ation al T o x ic o lo g y Program, 1982: Adrenal cortex: A d e n o m a .......................F -3 7 F .1 .1 3 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................F -3 7 F .1 .1 3 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................F -3 7 F .1 .1 3 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ............... F -3 9 F.1.14. National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcom a.............................................................................................................................. F -4 0 F .1 .1 4 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F -4 0 F .1 .1 4 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................F -4 0 F .1 .1 4 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ............... F -4 2 F.1.15. National T oxicology Program, 1982: H em atopoietio System: L ym p h om a or L eu k em ia ....................................................................................................... F -43 F .1 .1 5 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................F -43 F .1 .1 5 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................F -43 F .1 .1 5 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ............... F -45 F.1.16. National T oxicology Program, 1982: Liver: Hepatooellular A d en om a or C a rcin o m a ........................................................................................................ F -4 6 F .1 .1 6 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F -4 6 F .1 .1 6 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................F -4 6 F .1 .1 6 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ............... F -4 8 F.1.17. National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcom a.............................................................................................................................. F -4 9 F .1 .1 7 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F -4 9 F .1 .1 7 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................F -4 9 F .1 .1 7 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ............... F-51 F.1.18. National T oxicology Program, 1982: Lung: Alveolar/Bronchiolar A d en om a or C a rcin o m a ........................................................................................................ F -5 2 F .1 .1 8 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F -5 2 F .1 .1 8 .2 . Output for S elected M odel: M u ltistage Cancer, 2 -D e g r e e ................ F -5 2 This document is a draftfor review purposes only and does not constitute Agency policy. F-iii DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) F .1 .1 8 .3 . Figure for S elected M odel: M u ltistage Cancer, 2 - D e g r e e ............... F -5 4 F.1.19. National T oxicology Program, 1982: Liver: H epatocellular A denom a or C arcinom a.............................................................................................................................. F -55 F .1 .1 9 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ....................................... F -55 F .1 .1 9 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -55 F .1 .1 9 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -5 7 F .1 .2 0 . N ation al T o x ic o lo g y Program, 2006: Liver: C h o la n g io ca rcin o m a ................... F -5 8 F .1 .2 0 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ....................................... F -5 8 F .1 .2 0 .2 . Output for S elected M odel: M u ltistage Cancer, 3 -D e g r e e ..............F -5 8 F .1 .2 0 .3 . Figure for S elected M odel: M u ltistage Cancer, 3 - D e g r e e ..............F -6 0 F .1 .2 1 . N ation al T o x ic o lo g y Program , 2006: Liver: H epatocellu lar ad en om a..............F-61 F .1 .2 1 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F-61 F .1 .2 1 .2 . Output For S elected M odel: M u ltistage Cancer, 3 -D e g r e e ..............F-61 F .1 .2 1 .3 . Figure For S elected M odel: M u ltistage Cancer, 3 -D e g r e e ..............F -63 F.1.22. National T oxicology Program, 2006: Oral mucosa: squamous cell carcin o m a .................................................................................................................................... F -6 4 F .1 .2 2 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F -6 4 F .1 .2 2 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -6 4 F .1 .2 2 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -6 6 F.1.23. National T oxicology Program, 2006: Pancreas: adenoma or carcin o m a .................................................................................................................................... F -6 7 F .1 .2 3 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F -6 7 F .1 .2 3 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -6 7 F .1 .2 3 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -6 9 F.1.24. National T oxicology Program, 2006: Lung: Cystic keratinizing ep ith elio m a ..................................................................................................................................F -7 0 F .1 .2 4 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F -7 0 F .1 .2 4 .2 . Output for S elected M odel: M u ltistage Cancer, 2 -D e g r e e ..............F -7 0 F .1 .2 4 .3 . Figure for S elected M odel: M u ltistage Cancer, 2 - D e g r e e ..............F -7 2 F .1 .2 5 . Toth et al., 1979: Liver: T u m o rs........................................................................................F -73 F .1 .2 5 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F -73 F .1 .2 5 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -73 F .1 .2 5 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -75 F.1.26. D ella Porta et al., 1987: Table 4, B 6C 3 m ice, m ale, hepatocellular carcin o m a .................................................................................................................................... F -7 6 F .1 .2 6 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F -7 6 F .1 .2 6 .2 . Output for S elected M odel: M u ltistage Cancer, 2 -D e g r e e ..............F -7 6 F .1 .2 6 .3 . Figure for S elected M odel: M u ltistage Cancer, 2 - D e g r e e ..............F -7 8 F.1.27. D ella Porta et al., 1987: Table 4, B 6C 3 m ice, fem ale, hepatocellular a d en o m a .......................................................................................................................................F -7 9 F .1 .2 7 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F -7 9 F .1 .2 7 .2 . Output for S elected M odel: M u ltistage Cancer, 2 -D e g r e e ..............F -7 9 F .1 .2 7 .3 . Figure for S elected M odel: M u ltistage Cancer, 2 - D e g r e e ..............F-81 This document is a draftfor review purposes only and does not constitute Agency policy. F-iv DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) F.1.28. D ella Porta et al., 1987: Table 4, B 6C 3 m ice, fem ale, hepatocellular carcin om a.................................................................................................................................... F -8 2 F .1 .2 8 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ......................................... F -8 2 F .1 .2 8 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e .................F -8 2 F .1 .2 8 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e .................F -8 4 F.2. A D M IN IS T E R E D D O S E B M D S R E S U L T S ............................................................................ F -85 F.2.1. K ociba et al., 1978: Stratified squam ous cell carcinoma o f hard palate or nasal turbinates....................................................................................................... F -85 F .2 .1 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ....................................... F -85 F .2 .1 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................ F -85 F .2 .1 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................ F -8 7 F .2.2. K ocib a et al., 1978: Stratified squam ous cell carcinom a o f to n g u e ...........F -8 8 F .2 .2 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ......................................... F -8 8 F .2 .2 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................ F -8 8 F .2 .2 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................ F -9 0 F .2.3. K ocib a et al., 1978: A d en om a o f adrenal c o r t e x .......................................................F-91 F .2 .3 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ........................................ F-91 F .2 .3 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................ F-91 F .2 .3 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................ F -93 F .2.4. K ocib a et al., 1978: H epatocellu lar adenom a(s) or carcin om a(s)...............F -9 4 F .2 .4 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ......................................... F -9 4 F .2 .4 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................ F -9 4 F .2 .4 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................ F -9 6 F.2.5. K ociba et al., 1978: Stratified squam ous cell carcinoma o f hard palate or nasal turbinates....................................................................................................... F -9 7 F .2 .5 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ....................................... F -9 7 F .2 .5 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................ F -9 7 F .2 .5 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ................ F -9 9 F .2.6. K ocib a et al., 1978: K eratinizing squam ous cell carcinom a o f lu n g .............. F -1 0 0 F .2 .6 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ...................................... F -1 0 0 F .2 .6 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 0 0 F .2 .6 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 0 2 F.2.7. National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcom a........................................................................................................................... F -103 F .2 .7 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ......................................F -103 F .2 .7 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -103 F .2 .7 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 0 5 F.2.8. National T oxicology Program, 1982: Liver: N eoplastic N odule or H epatocellu lar C a r c in o m a .................................................................................................F -1 0 6 F .2 .8 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ...................................... F -1 0 6 F .2 .8 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 0 6 F .2 .8 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 0 8 This document is a draftfor review purposes only and does not constitute Agency policy. F-v DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) F.2.9. National T oxicology Program, 1982: Adrenal: Cortical A denom a, or C arcinom a or A denom a, N O S ........................................................................................F -1 0 9 F .2 .9 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s .....................................F -1 0 9 F .2 .9 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 0 9 F .2 .9 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -111 F.2.10. National T oxicology Program, 1982: Thyroid: Follicular-C ell A d e n o m a ................................................................................................................................... F -1 1 2 F .2 .1 0 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s .....................................F -1 1 2 F .2 .1 0 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 1 2 F .2 .1 0 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 1 4 F.2.11. National T oxicology Program, 1982: Liver: N eoplastic N odule or H epatocellu lar C a r c in o m a .................................................................................................F -1 1 5 F .2 .1 1 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s .................................... F -1 1 5 F .2 .1 1 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ............. F -1 1 5 F .2 .1 1 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 1 7 F.2.12. National T oxicology Program, 1982: Thyroid: Follicular-C ell A d en om a or C a rcin o m a ......................................................................................................F -1 1 8 F .2 .1 2 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s .....................................F -1 1 8 F .2 .1 2 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ............. F -1 1 8 F .2 .1 2 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 2 0 F .2 .1 3 . N ation al T o x ic o lo g y Program, 1982: Adrenal cortex: A d e n o m a .................... F -121 F .2 .1 3 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s .....................................F -121 F .2 .1 3 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -121 F .2 .1 3 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ............. F -123 F.2.14. National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcom a............................................................................................................................F -1 2 4 F .2 .1 4 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s .................................... F -1 2 4 F .2 .1 4 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 2 4 F .2 .1 4 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 2 6 F.2.15. National T oxicology Program, 1982: H em atopoietio System: L ym p h om a or L eu k em ia .....................................................................................................F -1 2 7 F .2 .1 5 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s .................................... F -1 2 7 F .2 .1 5 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 2 7 F .2 .1 5 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 2 9 F.2.16. National T oxicology Program, 1982: Liver: Hepatooellular A d en om a or C a rcin o m a ......................................................................................................F -1 3 0 F .2 .1 6 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s .................................... F -1 3 0 F .2 .1 6 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 3 0 F .2 .1 6 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 3 2 F.2.17. National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcom a........................................................................................................................... F -133 F .2 .1 7 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s .................................... F -133 F .2 .1 7 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -133 This document is a draftfor review purposes only and does not constitute Agency policy. F-vi DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) F .2 .1 7 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e .............F -1 3 5 F.2.18. National T oxicology Program, 1982: Lung: Alveolar/Bronchiolar A d en om a or C a rcin o m a .....................................................................................................F -1 3 6 F .2 .1 8 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s .....................................F -1 3 6 F .2 .1 8 .2 . Output for S elected M odel: M u ltistage Cancer, 2 -D e g r e e .............F -1 3 6 F .2 .1 8 .3 . Figure for S elected M odel: M u ltistage Cancer, 2 - D e g r e e .............F -1 3 8 F.2.19. National T oxicology Program, 1982: Liver: H epatocellular A denom a or C arcinom a........................................................................................................................... F -1 3 9 F .2 .1 9 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s .....................................F -1 3 9 F .2 .1 9 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 3 9 F .2 .1 9 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -141 F .2 .2 0 . N ation al T o x ic o lo g y Program, 2006: Liver: C h o la n g io ca rcin o m a .................F -1 4 2 F .2 .2 0 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s .....................................F -1 4 2 F .2 .2 0 .2 . Output for S elected M odel: M u ltistage Cancer, 3 -D e g r e e ..............F -1 4 2 F .2 .2 0 .3 . Figure for S elected M odel: M u ltistage Cancer, 3 - D e g r e e ..............F -1 4 4 F .2 .2 1 . N ation al T o x ic o lo g y Program , 2006: Liver: H epatocellu lar ad en om a .......... F -1 4 5 F .2 .2 1 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s .....................................F -1 4 5 F .2 .2 1 .2 . Output for S elected M odel: M u ltistage Cancer, 3 -D e g r e e ..............F -1 4 5 F .2 .2 1 .3 . Figure for S elected M odel: M u ltistage Cancer, 3 - D e g r e e ..............F -1 4 7 F.2.22. National T oxicology Program, 2006: Oral mucosa: squam ous cell carcin o m a ..................................................................................................................................F -1 4 8 F .2 .2 2 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ......................................F -1 4 8 F .2 .2 2 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 4 8 F .2 .2 2 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 5 0 F.2.23. National T oxicology Program, 2006: Pancreas: adenoma or carcin o m a ..................................................................................................................................F -151 F .2 .2 3 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ...................................... F -151 F .2 .2 3 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -151 F .2 .2 3 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -153 F.2.24. National T oxicology Program, 2006: Lung: Cystic keratinizing ep ith elio m a ............................................................................................................................... F -1 5 4 F .2 .2 4 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ...................................... F -1 5 4 F .2 .2 4 .2 . Output for S elected M odel: M u ltistage Cancer, 2 -D e g r e e ..............F -1 5 4 F .2 .2 4 .3 . Figure for S elected M odel: M u ltistage Cancer, 2 - D e g r e e ..............F -1 5 6 F .2 .2 5 . Toth et al., 1979: Liver: T u m o rs..................................................................................... F -1 5 7 F .2 .2 5 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ...................................... F -1 5 7 F .2 .2 5 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 5 7 F .2 .2 5 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ..............F -1 5 9 F.2.26. D ella Porta et al., 1987: Table 4, B 6C 3 m ice, m ale, hepatocellular carcin o m a ..................................................................................................................................F -1 6 0 F .2 .2 6 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ...................................... F -1 6 0 F .2 .2 6 .2 . Output for S elected M odel: M u ltistage Cancer, 2 -D e g r e e ..............F -1 6 0 F .2 .2 6 .3 . Figure for S elected M odel: M u ltistage Cancer, 2 - D e g r e e ..............F -1 6 2 This document is a draftfor review purposes only and does not constitute Agency policy. F-vii DRAFT--DO NOT CITE OR QUOTE CONTENTS (continued) F.2.27. D ella Porta et al., 1987: Table 4, B 6C 3 m ice, fem ale, hepatocellular a d en o m a .................................................................................................................................... F -163 F .2 .2 7 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ......................................F -163 F .2 .2 7 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ...........F -163 F .2 .2 7 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ...........F -1 6 5 F.2.28. D ella Porta et al., 1987: Table 4, B 6C 3 m ice, fem ale, hepatocellular carcin om a..................................................................................................................................F -1 6 6 F .2 .2 8 .1 . Sum m ary T able o f B M D S M o d elin g R e s u lt s ...................................... F -1 6 6 F .2 .2 8 .2 . Output for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ...........F -1 6 6 F .2 .2 8 .3 . Figure for S elected M odel: M u ltistage Cancer, 1 -D e g r e e ...........F -1 6 8 F.3. R E F E R E N C E S ......................................................................................................................................F -1 6 9 This document is a draftfor review purposes only and does not constitute Agency policy. F-viii DRAFT--DO NOT CITE OR QUOTE 1 APPENDIX F. CANCER BENCH M ARK DOSE M O DELING 2 3 4 F.1. B L O O D B M D S R E SU L T S 5 F.1.1. K ociba et al., 1978: Stratified squam ous cell carcinom a o f hard palate or nasal 6 turbinates 7 F .1 .1 .1 . Summary Table o f BMDS Modeling Results Model Degrees of X2pFreedom Value Multistage Cancer, 1-Degree a 3 0.815 M ultistage Cancer, 2-D egree 3 0.985 M ultistage Cancer, 3-Degree 3 0.999 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 31.564 5.763E+00 2.795E+00 30.170 1.369E+01 3.416E +00 29.930 1.917E+01 3.578E +00 Notes a Best-fitting m odel, BM D S output presented in this appendix 8 9 10 F .1.1.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 11 K o cib a et al., 1978: Stratified squam ous cell carcinom a o f hard palate or nasal turbinates 12 13 14 15 16 M u l t i s t a g e C a n c e r M o d e l . (Version: 1.7; Date: 0 5 / 1 6 / 2 0 0 8 ) 17 I n p u t D a t a File: C : \ 4 \ B l o o d \ 1 m s c l I P e r c p a l a t e n a s a l . ( d ) 18 G n u p l o t P l o t t i n g File: C : \ 4 \ B l o o d \ 1 m s c l l P e r c p a l a t e n a s a l . p l t 19 T h u A p r 01 1 5 : 5 6 : 0 3 2010 20 21 22 Source - Table 4 23 24 25 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 26 27 P[response] = background + (1-background)*[1-EXP( 28 -betai*dose^1)] 29 30 The parameter betas are restricted to be positive 31 32 33 D e p e n d e n t v a r i a b l e = M e a n 34 Independent variable = Dose 35 36 Total number of observations = 4 37 Total number of records with missing values = 0 38 Total number of parameters in model = 2 39 Total number of specified parameters = 0 40 Degree of polynomial = 1 41 42 43 M a x i m u m number of iterations = 250 44 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 45 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 This document is a draftfor review purposes only and does not constitute Agency policy. F-1 DRAFT--DO NOT CITE OR QUOTE 1 Default Initial Parameter Values 2 Background = 0 3 Beta(1) = 0.00226154 4 5 6 Asymptotic Correlation Matrix of Parameter Estimates 7 8 ( *** The model parameter(s) -Background 9 have been estimated at a boundary point, or have been specified by the user, 10 an d do n ot a p p e a r in t he c o r r e l a t i o n m a t r i x ) 11 12 Beta(1) 13 14 Beta(1) 1 15 16 17 18 P a r a m e t e r E s t i m a t e s 19 20 95.0% Wald Confidence Interval 21 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 22 Background 0 23 Beta(1) 0.0017438 24 25 * - I n d i c a t e s that this v a l u e is not c a l culated. 26 27 28 29 Analysis of Deviance Table 30 31 M o d e l Log(likelihood)i # Param's Deviance Test d.f. P-value 32 Full model -13.9385 4 33 Fitted model -14.7819 1 1.68696 3 0.6398 34 Reduced model -20.2589 1 12.6409 3 0.005481 35 36 AIC: 31.5639 37 38 39 Goodness of Fit 40 Scaled 41 Dose Est. Prob. Expected Observed Size Residu 42 43 0.0000 0.0000 0.000 0.000 85 0.000 44 1.5617 0.0027 0.136 0.000 50 -0.369 45 7.1600 0.0124 0.620 0.000 50 -0.793 46 38.7212 0.0653 3.265 4.000 50 0.421 47 48 Chi^2 = 0.94 d.f. = 3 P-value = 0.8153 49 50 51 B e n c h m a r k D o s e C o m p u t a t i o n 52 53 S p e c i f i e d e f f e c t 0.01 54 55 Risk Type Extra risk 56 57 Confidence level 0.95 58 59 BMD 5.76347 60 61 BMDL 2.79485 62 63 BMDU 14.9396 64 65 Taken together, (2.79485, 14.9396) a 90 two-sided confidence 66 interval for the BMD 67 68 Multistage Cancer Slope Factor 0.003578 69 70 This document is a draftfor review purposes only and does not constitute Agency policy. F-2 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.1.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 14:56 04/01 2010 3 4 K ociba et al., 1978: Stratified squam ous cell carcinom a o f hard palate or nasal turbinates This document is a draftfor review purposes only and does not constitute Agency policy. F-3 DRAFT--DO NOT CITE OR QUOTE 1 F.1.2. Kociba et al., 1978: Stratified squamous cell carcinoma of tongue 2 F.1.2.1. S u m m a ry Table o f B M D S M odeling R esults Model Multistage Cancer, 1-Degree a Degrees of Freedom 2 x2pValue 0.472 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 47.933 6.091E+00 2.600E+00 Notes Multistage Cancer, 2-Degree 2 0.472 47.933 6.091E+00 2.600E+00 final =0 Multistage Cancer, 3-Degree 2 0.472 47.933 6.091E+00 2.600E+00 final =0 aBest-fitting model, BMDS output presented in this appendix 3 4 5 F .1.2.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 K ociba et al., 1978: Stratified squam ous cell carcinom a o f tongue 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 4 \ B l o o d \ 2 _ m s c 1 _ 1 P e r c _ t o n g u e . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 4 \ B l o o d \ 2 _ m s c 1 _ 1 P e r c _ t o n g u e . p l t 14 T h u A p r 01 1 5 : 5 6 : 3 5 2010 15 16 17 S o u r c e - T a b l e 4 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^i)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.0092514 46 Beta(1) = 0.00137224 47 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-4 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 Background Beta(1) 4 5 Background 1 -0.58 6 7 Beta(1) -0.58 1 8 9 10 11 P a r a m e t e r E s t i m a t e s 12 13 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 14 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 15 B a c k g r o u n d 0.00510501 16 Beta(1) 0.00165011 17 18 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 19 20 21 22 Analysis of Deviance Table 23 24 Model Log(likelihood) # Param's Deviance Test d.f. P-value 25 Full model -21.1523 4 26 Fitted model -21.9667 2 1.62881 2 0.4429 27 Reduced model -24.1972 1 6.08976 3 0.1073 28 29 AIC: 47.9334 30 31 32 Goodness of Fit 33 Scaled 34 Dose Est. Prob. Expected Observed Size Residual 35 36 0.0000 0.0051 0.434 0.000 85 - 0 .660 37 1.5617 0.0077 0.383 1.000 50 1.000 38 7.1600 0.0168 0.840 1.000 50 0.177 39 38.7212 0.0667 3.334 3.000 50 -0.189 40 41 Chi^2 = 1.50 d.f. = 2 P-value = 0.4716 42 43 44 Benchmark Dose Computation 45 46 Specified effect 0.01 47 48 Risk Type Extra risk 49 50 Confidence level 0.95 51 52 BMD 6.0907 53 54 BMDL 2.60049 55 56 BMDU 519124 57 58 T a k e n t o g e t h e r , (2.60049, 5 1 9 1 2 4 ) is a 90 two-sided confidence 59 interval for the BMD 60 61 M u l t i s t a g e C a n c e r S l o p e F a c t o r 0.00384542 62 This document is a draftfor review purposes only and does not constitute Agency policy. F-5 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.2.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 14:56 04/01 2010 3 4 K ociba et al., 1978: Stratified squam ous cell carcinom a o f tongue This document is a draftfor review purposes only and does not constitute Agency policy. F-6 DRAFT--DO NOT CITE OR QUOTE 1 F.1.3. Kociba et al., 1978: Adenoma of adrenal cortex 2 F.1.3.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 3 0.779 M ultistage Cancer, 2-D egree 3 0.779 M ultistage Cancer, 3-Degree 3 0.779 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 52.488 3.254E+00 1.852E+00 52.488 3.254E+00 1.852E+00 final =0 52.488 3.254E+00 1.852E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.3.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 K ociba et al., 1978: A denom a o f adrenal cortex 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 4 \ B l o o d \ 3 m s c l I P e r c a d r e a d e n o m a . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 4 \ B l o o d \ 3 m s c l l P e r c a d r e a d e n o m a . p l t 14 T h u A p r 01 1 5 : 5 7 : 0 7 2010 15 16 17 S o u r c e - T a b l e 5 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^i)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.00493756 46 Beta(1) = 0.0026639 47 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-7 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 ( *** The model parameter(s) -Background 4 have been estimated at a boundary point, or have been specified by the user, 5 and do not appear in the correlation matrix ) 6 7 Beta(1) 8 9 Beta(1) 1 10 11 12 13 P a r a m e t e r E s t i m a t e s 14 15 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 16 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 17 B a c k g r o u n d 0 18 Beta(1) 0.00308883 19 20 * - I n d i c a t e s that this v a l u e is not c a l culated. 21 22 23 24 Analysis of Deviance Table 25 26 Model Log(likelihood)i # Param's Deviance Test d.f. P-value 27 Full model -24.6514 4 28 Fitted model -25.2438 1 1.18487 3 0.7566 29 Reduced model -31.4904 1 13.6781 3 0.003378 30 31 AIC: 52.4876 32 33 34 Goodness of Fit 35 Scaled 36 Dose Est. Prob. Expected Observed Size Residu 37 38 0.0000 0.0000 0.000 0.000 85 0.000 39 1.5617 0.0048 0.241 0.000 50 -0.492 40 7.1600 0.0219 1.094 2.000 50 0.876 41 38.7212 0.1127 5.636 5.000 50 -0.285 42 43 Chi^2 = 1.09 d.f. = 3 P-value = 0.7793 44 45 46 Benchmark Dose Computation 47 48 Specified effect 0.01 49 50 Risk Type Extra risk 51 52 Confidence level 0.95 53 54 BMD 3.25376 55 56 BMDL 1.85162 57 58 BMDU 6.58595 59 60 T a k e n t o g e t h e r , (1.85162, 6.58595) is a 90 two-sided confidence 61 i n t e r v a l for t he B M D 62 63 M u l t i s t a g e C a n c e r Slope Factor 0.00540067 64 65 This document is a draftfor review purposes only and does not constitute Agency policy. F-8 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.3.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 14:57 04/01 2010 3 4 K ociba et al., 1978: A denom a o f adrenal cortex This document is a draftfor review purposes only and does not constitute Agency policy. F-9 DRAFT--DO NOT CITE OR QUOTE 1 F.1.4. Kociba et al., 1978: Hepatocellular adenoma(s) or carcinoma(s) 2 F.1.4.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom Multistage Cancer, 1-Degree a 2 x2pValue 0.245 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 143.261 7.010E-01 5.013E-01 Notes M ultistage Cancer, 2-D egree 2 0.245 143.261 7.010E-01 5.013E-01 final =0 M ultistage Cancer, 3-Degree 2 0.245 143.261 7.010E-01 5.013E-01 final =0 a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.4.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 K ociba et al., 1978: H epatocellular adenom a(s) or carcinom a(s) 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 4 \ B l o o d \ 4 m s c l I P e r c l i v e r ad carc.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ 4 \ B l o o d \ 4 m s c l l P e r c l i v e r a d c a r c . p l t 14 T h u A p r 01_ 1 5 i 5 7 : 4 1 2010 15 16 17 S o u r c e - T a b l e 1 in G o o d m a n an d S a u e r 1992 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -beta1*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.0400263 46 Beta(1) = 0.0124752 47 This document is a draftfor review purposes only and does not constitute Agency policy. F-10 DRAFT--DO NOT CITE OR QUOTE 1 2 Asymptotic Correlation Matrix of Parameter Estimates 3 4 Background Beta(1) 5 6 Background 1 -0.51 7 8 Beta(1) -0.51 1 9 10 11 12 P a r a m e t e r E s t i m a t e s 13 14 95.0% W a l d C o n f i d e n c e I n t e r v a l 15 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 16 Background 0.0221468 * 17 Beta(1) 0.0143372 * 18 19 * - I n d i c a t e s t h a t t h i s v a l u e is n ot c a l c u l a t e d . 20 21 22 23 A n alysis of Deviance Table 24 25 Model Log(likelihood) # Param's Deviance Test d.f. P-value 26 Full model -68.2561 4 27 Fitted model -69.6304 2 2.74857 2 0.253 28 Reduced model -89.1983 1 41.8843 3 <.0001 29 30 AIC: 143.261 31 32 33 G o o d n e s s of Fit 34 Scaled 35 Dose Est. Prob. Expected Observed Size Residual 36 37 0.0000 0.0221 1.905 2.000 86 0.070 38 1.5473 0.0436 2.180 1.000 50 -0.817 39 7.1546 0.1175 5.874 9.000 50 1.373 40 38.5608 0.4374 19.685 18.000 45 - 0 .506 41 42 Chi^2 = 2.81 d.f. = 2 P-value = 0.2449 43 44 45 Benchmark Dose Computation 46 47 Specified effect 0.01 48 49 Risk Type Extra risk 50 51 C o n f i d e n c e l e v e l 0.95 52 53 BMD 0.700996 54 55 BMDL 0.501345 56 57 BMDU 1.04839 58 59 T a k e n t o g e t h e r , (0.501345, 1.048 3 9 ) is a 90 two-sided confidence 60 interval for the BMD 61 62 Multistage Cancer Slope Factor 0.0199463 63 64 This document is a draftfor review purposes only and does not constitute Agency policy. F-11 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.4.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 14:57 04/01 2010 3 4 K ociba et al., 1978: H epatocellular adenom a(s) or carcinom a(s) This document is a draftfor review purposes only and does not constitute Agency policy. F-12 DRAFT--DO NOT CITE OR QUOTE 1 F.1.5. K ociba et al., 1978: Stratified squam ous cell carcinom a o f hard palate or nasal 2 turbinates 3 F .1 .5 .1 . Summary Table o f BMDS Modeling Results Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 3 0.815 M ultistage Cancer, 2-D egree 3 0.985 M ultistage Cancer, 3-Degree 3 0.999 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 31.564 5.763E+00 2.795E+00 30.170 1.369E+01 3.416E+00 29.930 1.917E+01 3.578E+00 Notes a Best-fitting m odel, BM D S output presented in this appendix 4 5 6 F .1.5.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 7 K ociba et al., 1978: Stratified squam ous cell carcinom a o f hard palate or nasal turbinates 8 9 10 11 12 M u l t i s t a g e C a n c e r M o d e l . (Version: 1.7; Date: 0 5 / 1 6 / 2 0 0 8 ) 13 I n p u t D a t a F i l e : C : \ 4 \ B l o o d \ 5 _ m s c 1 _ 1 P e r c _ n a s a l . ( d ) 14 G n u p l o t P l o t t i n g File: C : \ 4 \ B l o o d \ 5 _ m s c 1 _ 1 P e r c _ n a s a l . p l t 15 T h u A p r 01 1 5 : 5 8 : 1 4 2 0 1 0 16 17 18 S o u r c e - T a b l e 5 19 20 21 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 22 23 P[response] = b a c k g r o u n d + (1-background)*[1-EXP( 24 -betai*dose^i)] 25 26 The parameter betas are restricted to be positive 27 28 29 Dependent variable = Mean 30 Independent variable = Dose 31 32 Total number of observations = 4 33 Total n u m b e r of records w i t h m i s s i n g va l u e s = 0 34 Total number of parameters in model = 2 35 Total number of specified p arameters = 0 36 Degree of polynomial = 1 37 38 39 Maximum number of iterations = 250 40 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 41 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1e-008 42 43 44 45 Default Initial Parameter Values 46 Background = 7.10818e-005 47 Beta(1) = 0.00222324 This document is a draftfor review purposes only and does not constitute Agency policy. F-13 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Asymptotic Correlation Matrix of Parameter Estimates 4 5 ( *** The model parameter(s) -Background 6 have been estimated at a boundary point, or have been specified by the user, 7 and do not appear in the correlation matrix ) 8 9 Beta(1) 10 11 B e t a ( 1 ) 1 12 13 14 15 P a r a m e t e r E s t i m a t e s 16 17 95.0% W a l d C o n f i d e n c e I n t e r v a l 18 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 19 B a c k g r o u n d 0 20 Beta(1) 0.0022294 21 22 * - I n d i c a t e s that this v a l u e is not c a l culated. 23 24 25 26 Analysis of Deviance Table 27 28 Model Log(likelihood) # Param's Deviance Test d.f. P-value 29 Full model -18.7562 4 30 Fitted model -18.9547 1 0.397012 3 0.9409 31 R e d u c e d m o d e l -24.1972 1 10.882 3 0.01238 32 33 AIC: 39.9093 34 35 36 Goodness of Fit 37 Scaled 38 Dose Est. Prob. Expected Observed Size Residu. 39 40 0.0000 0.0000 0.000 0.000 86 0.000 41 1.5473 0.0034 0.172 0.000 50 -0.416 42 7.1546 0.0158 0.791 1.000 50 0.237 43 38.5608 0.0824 4.036 4.000 49 - 0 .019 44 45 Chi^2 = 0.23 d.f. = 3 P-value = 0.9728 46 47 48 Benchmark Dose Computation 49 50 Specified effect 0.01 51 52 Risk Type Extra risk 53 54 Confidence level 0.95 55 56 BMD 4.50809 57 58 BMDL 2.34012 59 60 BMDU 10.4588 61 62 T a k e n t o g e t h e r , (2.34012, 10.4588) is a 90 two-sided confidence 63 inter v a l for the BMD 64 65 M u l t i s t a g e Cancer Slope Factor 0.00427329 66 67 This document is a draftfor review purposes only and does not constitute Agency policy. F-14 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.5.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 14:58 04/01 2010 3 4 K ociba et al., 1978: Stratified squam ous cell carcinom a o f hard palate or nasal turbinates This document is a draftfor review purposes only and does not constitute Agency policy. F-15 DRAFT--DO NOT CITE OR QUOTE 1 F.1.6. Kociba et al., 1978: Keratinizing squamous cell carcinoma of lung 2 F.1.6.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 3 0.626 M ultistage Cancer, 2-D egree 3 0.964 M ultistage Cancer, 3-Degree 3 0.997 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 45.298 3.140E+00 1.786E+00 42.736 1.004E+01 2.707E+00 42.291 1.556E+01 3.135E+00 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.6.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 K ociba et al., 1978: K eratinizing squam ous cell carcinom a o f lung 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 4 \ B l o o d \ 6 m s c l I P e r c k e r a carc.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ 4 \ B l o o d \ 6 m s c l l P e r c k e r a c a r c . p l t 14 T h u A p r 01 1 5 : 5 8 : 4 9 2010 15 16 17 S o u r c e - T a b l e 5 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^i)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0 46 Beta(1) = 0.00419802 47 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-16 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 ( *** The model parameter(s) -Background 4 have been estimated at a boundary point, or have been specified by the user, 5 and do not appear in the correlation matrix ) 6 7 Beta(1) 8 9 Beta(1) 1 10 11 12 13 P a r a m e t e r E s t i m a t e s 14 15 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 16 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 17 B a c k g r o u n d 0 18 Beta(1) 0.00320098 19 20 * - I n d i c a t e s that this v a l u e is not c a l culated. 21 22 23 24 Analysis of Deviance Table 25 26 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 27 Full model -20.0957 4 28 Fitted model -21.6489 1 3.10639 3 0.3755 29 Reduced model -31.4904 1 22.7894 3 <.0001 30 31 AIC: 45.2978 32 33 34 Goodness of Fit 35 Scaled 36 Dose Est. Prob. Expected Observed Size Residu. 37 38 0.0000 0.0000 0.000 0.000 86 0.000 39 1.5473 0.0049 0.247 0.000 50 -0.498 40 7.1546 0.0226 1.132 0.000 50 -1.076 41 38.5608 0.1161 5.690 7.000 49 0.584 42 43 Chi^2 = 1.75 d.f. = 3 P-value = 0.6263 44 45 46 Benchmark Dose Computation 47 48 Specified effect 0.01 49 50 Risk Type Extra risk 51 52 Confidence level 0.95 53 54 BMD 3.13977 55 56 BMDL 1.78648 57 58 BMDU 6.28288 59 60 T a k e n t o g e t h e r , (1.78648, 6.28288) is a 90 two-sided confidence 61 i n t e r v a l for t he B M D 62 63 M u l t i s t a g e C a n c e r Slope Factor 0.0055976 64 65 This document is a draftfor review purposes only and does not constitute Agency policy. F-17 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.6.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 14:58 04/01 2010 3 4 K ociba et al., 1978: K eratinizing squam ous cell carcinom a o f lung This document is a draftfor review purposes only and does not constitute Agency policy. F-18 DRAFT--DO NOT CITE OR QUOTE 1 F.1.7. National Toxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma 2 F.1.7.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom Multistage Cancer, 1-Degree a 2 x2pValue 0.179 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 75.385 3.127E+00 1.380E+00 Notes M ultistage Cancer, 2 0.179 75.385 3.127E+00 1.380E+00 final =0 2-D egree M ultistage Cancer, 2 0.179 75.385 3.127E+00 1.380E+00 final =0 3-Degree a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.7.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 4 \ B l o o d \ 7 _ m s c 1 _ 1 P e r c _ s u b _ f i b r o . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 4 \ B l o o d \ 7 _ m s c 1 _ 1 P e r c _ s u b _ f i b r o . p l t 14 T h u A p r 01 1 5 : 5 9 : 2 5 2010 15 16 17 S o u r c e - T a b l e 10 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.0268183 46 Beta(1) = 0.00211524 47 This document is a draftfor review purposes only and does not constitute Agency policy. F-19 DRAFT--DO NOT CITE OR QUOTE 1 2 Asymptotic Correlation Matrix of Parameter Estimates 3 4 Background Beta(1) 5 6 Background 1 -0.63 7 8 Beta(1) -0.63 1 9 10 11 12 P a r a m e t e r E s t i m a t e s 13 14 95.0% W a l d C o n f i d e n c e I n t e r v a l 15 V a r i a b l e Estimate Std. Err. Lower Conf. Limit Upper Conf 16 Background 0.0149841 * * * 17 Beta(1) 0.00321423 * * * 18 19 - I n d i c a t e s t h a t t h i s v a l u e is not c a l c u l a t e d . 20 21 22 23 A n alysis of Deviance Table 24 25 Model Log(likelihood) # Param's Deviance Test d.f. P-value 26 Full model -33.5998 4 27 Fitted model -35.6923 2 4.18508 2 0.1234 28 Reduced model -37.7465 1 8.29346 3 0.04032 29 30 AIC: 75.3847 31 32 33 G o o d n e s s of Fit 34 Scaled 35 Dose Est. Prob. Expected Observed Size Residual 36 37 0.0000 0.0150 1.124 0.000 75 -1.068 38 1.9574 0.0212 1.058 2.000 50 0.926 39 5.6942 0.0328 1.642 3.000 50 1.077 40 29.7519 0.1048 5.136 4.000 49 - 0 .530 41 42 Chi^2 = 3.44 d.f. = 2 P-value = 0.1792 43 44 45 Benchmark Dose Computation 46 47 Specified effect = 0.01 48 49 Risk Type = Extra risk 50 51 C o n f i d e n c e l e v e l = 0.95 52 53 BMD = 3.12683 54 55 BMDL = 1.38047 56 57 BMDU = 2.18232e+006 58 59 T a k e n t ogether, (1.38047, 2 . 1 8 2 3 2 e + 0 0 6 ) is a 90 % two-sided confidence 60 interval for the BMD 61 62 Multistage Cancer Slope Factor = 0.00724391 63 64 This document is a draftfor review purposes only and does not constitute Agency policy. F-20 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.7.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 14:59 04/01 2010 3 4 National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma This document is a draftfor review purposes only and does not constitute Agency policy. F-21 DRAFT--DO NOT CITE OR QUOTE 1 F.1.8. National Toxicology Program, 1982: Liver: Neoplastic Nodule or Hepatocellular 2 Carcinoma 3 F .1 .8 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom Multistage Cancer, 1-Degree a 2 Multistage Cancer, 2-Degree 2 Multistage Cancer, 3-Degree 1 x2pValue 0.218 0.491 0.239 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 135.190 1.169E+00 7.375E-01 133.447 5.578E+00 8.771E-01 135.435 7.204E+00 8.786E-01 Notes aBest-fitting model, BMDS output presented in this appendix 4 5 6 F .1.8.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 7 National T oxicology Program, 1982: Liver: N eoplastic N odule or H epatocellular Carcinoma 8 9 10 11 12 M u l t i s t a g e C a n c e r M o d e l . (Version: 1.7; Date: 0 5 / 1 6 / 2 0 0 8 ) 13 I n p u t D a t a F i l e : C : \ 4 \ B l o o d \ 8 _ m s c 1 _ 1 P e r c _ l i v e r _ n o d . ( d ) 14 G n u p l o t P l o t t i n g File: C : \ 4 \ B l o o d \ 8 m s c 1 1 P e r c l i v e r n o d . p l t 15 T h u A p r 0 1 _ 1 6 : 0 0 : 0 0 2 0 1 0 16 17 18 S o u r c e - T a b l e 10 19 20 21 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 22 23 P[response] = b a c k g r o u n d + (1-background)*[1-EXP( 24 -betai*dose^i)] 25 26 The parameter betas are restricted to be positive 27 28 29 Dependent variable = Mean 30 Independent variable = Dose 31 32 Total number of observations = 4 33 Total n u m b e r of records w i t h m i s s i n g va l u e s = 0 34 Total number of parameters in model = 2 35 Total number of specified p arameters = 0 36 Degree of polynomial = 1 37 38 39 Maximum number of iterations = 250 40 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 41 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1e-008 42 43 44 45 Default Initial Parameter Values 46 Background = 0.0261097 47 Beta(1) = 0.0102165 This document is a draftfor review purposes only and does not constitute Agency policy. F-22 D RAFT-- DO N O T CITE OR QUOTE 1 2 3 Asymptotic Correlation Matrix of Parameter Estimates 4 5 Background Beta(1) 6 7 Background 1 -0.52 8 9 Beta(1) -0.52 1 10 11 12 13 P a r a m e t e r E s t i m a t e s 14 15 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 16 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 17 Background 0.0424738 18 Beta(1) 0.00859382 19 20 * - Indicates that this value is not calculated. 21 22 23 24 Analysis of Deviance Table 25 26 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 27 Full model -63.9149 4 28 Fitted model -65.5949 2 3.36005 2 0.1864 29 Reduced model -74.0195 1 20.2092 3 0.0001536 30 31 AIC: 135.19 32 33 34 Goodness of Fit 35 Scaled 36 Dose Est. Prob. Expected Observed Size Residu. 37 38 0.0000 0.0425 3.186 5.000 75 1.039 39 1.9574 0.0584 2.864 1.000 49 - 1 .135 40 5.6942 0.0882 4.410 3.000 50 -0.703 41 29.7519 0.2585 12.667 14.000 49 0.435 42 43 Chi^2 = 3.05 d.f. = 2 P-value = 0.2175 44 45 46 Benchmark Dose Computation 47 48 Specified effect 0.01 49 50 Risk Type Extra risk 51 52 Confidence level 0.95 53 54 BMD 1.16948 55 56 BMDL 0.737535 57 58 BMDU 2.17906 59 60 T a k e n t o g e t h e r , (0.737535, 2.17906) is a 90 two-sided confidence 61 i n t e r v a l for t he B M D 62 63 M u l t i s t a g e C a n c e r Slope Factor 0.0135587 64 65 This document is a draftfor review purposes only and does not constitute Agency policy. F-23 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.8.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 15:00 04/01 2010 3 4 National T oxicology Program, 1982: Liver: N eoplastic N odule or H epatocellular Carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-24 DRAFT--DO NOT CITE OR QUOTE 1 F.1.9. National Toxicology Program, 1982: Adrenal: Cortical Adenoma, or Carcinoma or 2 Adenoma, NOS 3 F .1 .9 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom Multistage Cancer, 1-Degree a 2 Multistage Cancer, 2-Degree 2 Multistage Cancer, 3-Degree 2 x2pValue 0.337 0.470 0.505 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 203.824 1.611E+00 8.140E-01 203.033 6.652E+00 8.904E-01 202.868 1.091E+01 9.100E-01 Notes aBest-fitting model, BMDS output presented in this appendix 4 5 6 F .1.9.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 7 National T oxicology Program, 1982: Adrenal: Cortical A denom a, or Carcinoma or Adenom a, 8 NOS 9 10 11 12 13 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 14 I n p u t D a t a File: C : \ 4 \ B l o o d \ 9 m s c l I P e r c a d r e c o r t ad carc.(d) 15 G n u p l o t P l o t t i n g F i l e : C : \ 4 \ B l o o d \ 9 m s c l l P e r c a d r e c o r t a d c a r c . p l t 16 T h u A p r 01 1 6 : 0 6 : 1 5 2010 17 18 19 S o u r c e - T a b l e 10 20 21 22 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 23 24 P[response] = background + (1-background)*[1-EXP( 25 -betai*dose^1)] 26 27 The parameter betas are restricted to be positive 28 29 30 Dependent variable = Mean 31 I n d e p e n d e n t v a r i a b l e = Dose 32 33 Total n u m b e r of o b s e r v a t i o n s = 4 34 Total number of records with missing values = 0 35 Total number of p arameters in model = 2 36 Total number of specified parameters = 0 37 Degree of polynomial = 1 38 39 40 Maximum number of iterations = 250 41 R e l a t i v e F u n c t i o n C o n v e r g e n c e has b e e n set to: 1e-008 42 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 43 44 45 46 Default Initial Parameter Values This document is a draftfor review purposes only and does not constitute Agency policy. F-25 DRAFT--DO NOT CITE OR QUOTE 1 Background = 0.134165 2 Beta(1) = 0.0069662 3 4 5 Asymptotic Correlation Matrix of Parameter Estimates 6 7 Background Beta(1) 8 9 Background 1 -0.54 10 11 B e t a ( 1 ) -0.54 1 12 13 14 15 P a r a m e t e r E s t i m a t e s 16 17 95.0% W a l d C o n f i d e n c e I n t e r v a l 18 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 19 B a c k g r o u n d 0.139854 20 Beta(1) 0.00623778 21 22 * - Indicates that this value is not calculated. 23 24 25 26 Analysis of Deviance Table 27 28 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 29 Full model -98.7282 4 30 Fitted model -99.912 2 2.36764 2 0.3061 31 R e d u c e d m o d e l -102.201 1 6.94636 3 0.07363 32 33 AIC: 203.824 34 35 36 Goodness of Fit 37 Scaled 38 Dose Est. Prob. Expected Observed Size Residu 39 40 0.0000 0.1399 10.209 11.000 73 0.267 41 1.9574 0.1503 7.364 9.000 49 0.654 42 5.6942 0.1699 8.324 5.000 49 -1.264 43 29.7519 0.2855 13.135 14.000 46 0.282 44 45 Chi^2 = 2.18 d.f. = 2 P-value = 0.3367 46 47 48 Benchmark Dose Computation 49 50 Specified effect 0.01 51 52 Risk Type Extra risk 53 54 Confidence level 0.95 55 56 BMD 1.6112 57 58 BMDL 0.81404 59 60 BMDU 370555 61 62 T a k e n t o g e t h e r , (0.81404, 3 7 0 5 5 5 ) is a 90 two-sided confidence 63 inter v a l for the BMD 64 65 M u l t i s t a g e Cancer Slope Factor 0.0122844 66 67 This document is a draftfor review purposes only and does not constitute Agency policy. F-26 DRAFT--DO NOT CITE OR QUOTE 1 F.1.9.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 15:06 04/01 2010 3 4 National T oxicology Program, 1982: Adrenal: Cortical A denom a, or Carcinoma or Adenom a, 5 NOS This document is a draftfor review purposes only and does not constitute Agency policy. F-27 DRAFT--DO NOT CITE OR QUOTE 1 F.1.10. N ational T oxicology Program , 1982: Thyroid: Follicular-C ell A denom a 2 F .1 .1 0 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom Multistage Cancer, 1-Degree a 2 x2pValue 0.568 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 92.411 3.376E+00 1.553E+00 Notes M ultistage Cancer, 2 0.735 91.749 9.526E+00 1.690E+00 2-D egree M ultistage Cancer, 2 0.773 91.626 1.385E+01 1.720E+00 3-Degree a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.10.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 1982: Thyroid: Follicular-C ell A denom a 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 4 \ B l o o d \ 1 0 _ m s c 1 _ 1 P e r c _ t h y _ a d . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 4 \ B l o o d \ 1 0 _ m s c 1 _ 1 P e r c _ t h y _ a d . p l t 14 _ T h u A p r 01 1 6 : 0 6 : 5 3 2010 15 16 17 S o u r c e - T a b l e 10 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.0283212 46 Beta(1) = 0.00346762 47 This document is a draftfor review purposes only and does not constitute Agency policy. F-28 DRAFT--DO NOT CITE OR QUOTE 1 2 Asymptotic Correlation Matrix of Parameter Estimates 3 4 Background Beta(1) 5 6 Background 1 -0.54 7 8 Beta(1) -0.54 1 9 10 11 12 P a r a m e t e r E s t i m a t e s 13 14 95.0% W a l d C o n f i d e n c e I n t e r v a l 15 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 16 Background 0.0332432 17 Beta(1) 0.00297726 18 19 * - I n d i c a t e s t h a t t h i s v a l u e is n ot c a l c u l a t e d . 20 21 22 23 Analysis of Deviance Table 24 25 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 26 Full model -43.5264 4 27 Fitted model -44.2053 2 1.35778 2 0.5072 28 Reduced model -46.2299 1 5.40699 3 0.1443 29 30 AIC: 92.4106 31 32 33 G o o d n e s s of Fit 34 Scaled 35 Dose Est. Prob. Expected Observed Size Residu. 36 37 0.0000 0.0332 2.427 3.000 73 0.374 38 1.9574 0.0389 1.749 2.000 45 0.194 39 5.6942 0.0495 2.425 1.000 49 - 0 .939 40 29.7519 0.1152 5.414 6.000 47 0 . 2 6 8 41 42 Chi^2 = 1.13 d.f. = 2 P-value = 0.5682 43 44 45 Benchmark Dose Computation 46 47 Specified effect 0.01 48 49 Risk Type Extra risk 50 51 C o n f i d e n c e l e v e l 0.95 52 53 BMD 3.3757 54 55 BMDL 1.55287 56 57 BMDU 306341 58 59 T a k e n t o g e t h e r , (1.55287, 3 0 6 3 4 1 ) is a 90 two-sided confidence 60 interval for the BMD 61 62 Multistage Cancer Slope Factor 0.00643967 63 64 This document is a draftfor review purposes only and does not constitute Agency policy. F-29 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.10.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 15:06 04/01 2010 3 4 National T oxicology Program, 1982: Thyroid: Follicular-C ell A denom a This document is a draftfor review purposes only and does not constitute Agency policy. F-30 DRAFT--DO NOT CITE OR QUOTE 1 F.1.11. National Toxicology Program, 1982: Liver: Neoplastic Nodule or Hepatocellular 2 Carcinoma 3 F .1 .1 1 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom x2pValue Multistage Cancer, 1-Degree a 2 0.218 M ultistage Cancer, 2 0.491 2-D egree M ultistage Cancer, 3-Degree 1 0.239 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 135.190 1.169E+00 7.375E-01 133.447 5.578E+00 8.771E-01 135.435 7.204E+00 8.786E-01 Notes a Best-fitting m odel, BM D S output presented in this appendix 4 5 6 F .1.11.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 7 National T oxicology Program, 1982: Liver: N eoplastic N odule or H epatocellular Carcinoma 8 9 10 11 12 M u l t i s t a g e C a n c e r M o d e l . (Version: 1.7; Date: 0 5 / 1 6 / 2 0 0 8 ) 13 I n p u t D a t a F i l e : C : \ 4 \ B l o o d \ 1 1 _ m s c 1 _ 1 P e r c _ l i v e r _ n o d . ( d ) 14 G n u p l o t P l o t t i n g File: C : \ 4 \ B l o o d \ 1 1 m s c 1 1 P e r c l i v e r n o d . p l t 15 _ T h u A p r 01 1 6 : 0 7 : 2 8 2 0 1 0 16 17 18 S o u r c e - T a b l e 9 19 20 21 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 22 23 P[response] = b a c k g r o u n d + (1-background)*[1-EXP( 24 -betai*dose^1)] 25 26 The parameter betas are restricted to be positive 27 28 29 Dependent variable = Mean 30 Independent variable = Dose 31 32 Total number of observations = 4 33 Total n u m b e r of records w i t h m i s s i n g va l u e s = 0 34 Total number of parameters in model = 2 35 Total number of specified p arameters = 0 36 Degree of polynomial = 1 37 38 39 Maximum number of iterations = 250 40 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 41 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1e-008 42 43 44 45 Default Initial Parameter Values 46 Background = 0 47 Beta(1) = 0.00219894 This document is a draftfor review purposes only and does not constitute Agency policy. F-31 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Asymptotic Correlation Matrix of Parameter Estimates 4 5 ( *** The model parameter(s) -Background 6 have been estimated at a boundary point, or have been specified by the user, 7 and do not appear in the correlation matrix ) 8 9 Beta(1) 10 11 B e t a ( 1 ) 1 12 13 14 15 P a r a m e t e r E s t i m a t e s 16 17 95.0% W a l d C o n f i d e n c e I n t e r v a l 18 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 19 B a c k g r o u n d 0 20 Beta(1) 0.00163808 21 22 * - Indicates that this value is not calculated. 23 24 25 26 Analysis of Deviance Table 27 28 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 29 Full model -11.3484 4 30 Fitted model -12.0522 1 1.40767 3 0.7037 31 R e d u c e d m o d e l -15.9189 1 9.14109 3 0.02747 32 33 AIC: 26.1044 34 35 36 Goodness of Fit 37 Scaled 38 Dose Est. Prob. Expected Observed Size Residu 39 40 0.0000 0.0000 0.000 0.000 74 0 . 000 41 1.9569 0.0032 0.160 0.000 50 -0.401 42 5.7027 0.0093 0.465 0.000 50 -0.685 43 29.8723 0.0478 2.388 3.000 50 0.406 44 45 Chi^2 = 0.79 d.f. = 3 P-value = 0.8507 46 47 48 Benchmark Dose Computation 49 50 Specified effect 0.01 51 52 Risk Type Extra risk 53 54 Confidence level 0.95 55 56 BMD 6.13543 57 58 BMDL 2.70101 59 60 BMDU 18.9354 61 62 T a k e n t o g e t h e r , (2.70101, 18.9354) is a 90 two-sided confidence 63 inter v a l for the BMD 64 65 M u l t i s t a g e Cancer Slope Factor 0.00370232 66 67 This document is a draftfor review purposes only and does not constitute Agency policy. F-32 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.11.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 15:07 04/01 2010 3 4 National T oxicology Program, 1982: Liver: N eoplastic N odule or H epatocellular Carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-33 DRAFT--DO NOT CITE OR QUOTE 1 F.1.12. National Toxicology Program, 1982: Thyroid: Follicular-Cell Adenoma or 2 Carcinoma 3 F .1 .1 2 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom Multistage Cancer, 1-Degree a 2 M ultistage Cancer, 2-D egree 2 M ultistage Cancer, 3-Degree 2 x2pValue 0.057 0.057 0.057 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 149.263 1.208E+00 6.984E-01 149.263 1.208E+00 6.984E-01 final =0 149.263 1.208E+00 6.984E-01 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 4 5 6 F .1.12.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 7 National T oxicology Program, 1982: Thyroid: Follicular-C ell A denom a or Carcinoma 8 9 10 11 12 M u l t i s t a g e C a n c e r M o d e l . (Version: 1.7; Date: 0 5 / 1 6 / 2 0 0 8 ) 13 I n p u t D a t a F i l e : C : \ 4 \ B l o o d \ 1 2 _ m s c 1 _ 1 P e r c _ t h y r o i d . ( d ) 14 G n u p l o t P l o t t i n g File: C : \ 4 \ B l o o d \ 1 2 _ m s c 1 _ 1 P e r c _ t h y r o i d . p l t 15 _ T h u A p r 01 1 6 : 0 8 : 0 3 2 0 1 0 16 17 18 S o u r c e - T a b l e 9 19 20 21 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 22 23 P[response] = b a c k g r o u n d + (1-background)*[1-EXP( 24 -betai*dose^1)] 25 26 The parameter betas are restricted to be positive 27 28 29 Dependent variable = Mean 30 Independent variable = Dose 31 32 Total number of observations = 4 33 Total n u m b e r of records w i t h m i s s i n g va l u e s = 0 34 Total number of parameters in model = 2 35 Total number of specified p arameters = 0 36 Degree of polynomial = 1 37 38 39 Maximum number of iterations = 250 40 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 41 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1e-008 42 43 44 45 Default Initial Parameter Values 46 Background = 0.0768555 47 Beta(1) = 0.00606248 This document is a draftfor review purposes only and does not constitute Agency policy. F-34 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Asymptotic Correlation Matrix of Parameter Estimates 4 5 Background Beta(1) 6 7 Background 1 -0.62 8 9 Beta(1) -0.62 1 10 11 12 13 P a r a m e t e r E s t i m a t e s 14 15 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 16 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 17 Background 0.0529006 18 Beta(1) 0.00831706 19 20 * - Indicates that this value is not calculated. 21 22 23 24 Analysis of Deviance Table 25 26 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 27 Full model -69.5946 4 28 Fitted model -72.6315 2 6.07383 2 0.04798 29 Reduced model -77.5267 1 15.8643 3 0.001209 30 31 AIC: 149.263 32 33 34 Goodness of Fit 35 Scaled 36 Dose Est. Prob. Expected Observed Size Residual 37 38 0.0000 0.0529 3.650 1.000 69 - 1 .425 39 1.9569 0.0682 3.273 5.000 48 0 . 989 40 5.7027 0.0968 4.839 8.000 50 1.512 41 29.8723 0.2613 13.063 11.000 50 -0.664 42 43 Chi^2 = 5.74 d.f. = 2 P-value = 0.0568 44 45 46 Benchmark Dose Computation 47 48 Specified effect 0.01 49 50 Risk Type Extra risk 51 52 Confidence level 0.95 53 54 BMD 1.2084 55 56 BMDL 0.698436 57 58 BMDU 2.89109 59 60 T a k e n t o g e t h e r , (0.698436, 2.89109) is a 90 two-sided confidence 61 i n t e r v a l for t he B M D 62 63 M u l t i s t a g e C a n c e r Slope Factor 0.0143177 64 65 This document is a draftfor review purposes only and does not constitute Agency policy. F-35 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.12.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 15:08 04/01 2010 3 4 National T oxicology Program, 1982: Thyroid: Follicular-C ell A denom a or Carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-36 DRAFT--DO NOT CITE OR QUOTE 1 F.1.13. National Toxicology Program, 1982: Adrenal cortex: Adenoma 2 F.1.13.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom Multistage Cancer, 1-Degree a 2 x2pValue 0.062 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 199.309 3.977E+00 1.223E+00 M ultistage Cancer, 2-D egree 2 0.062 199.309 3.977E+00 1.223E+00 final =0 M ultistage Cancer, 3-Degree 2 0.062 199.309 3.977E+00 1.223E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.13.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 1982: Adrenal cortex: Adenom a 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ B l o o d \ 1 3 _ m s c 1 _ 1 P e r c _ a d r e _ c o r t . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 1 3 m s c 1 1 P e r c a d r e c o r t . p l t 14 Fri A p r 02 1 0 : 5 3 : 1 6 2010 15 16 17 S o u r c e - T a b l e 9 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.163685 46 Beta(1) = 0.00144687 47 This document is a draftfor review purposes only and does not constitute Agency policy. F-37 DRAFT--DO NOT CITE OR QUOTE 1 2 Asymptotic Correlation Matrix of Parameter Estimates 3 4 Background Beta(1) 5 6 Background 1 -0.6 7 8 Beta(1) -0.6 1 9 10 11 12 P a r a m e t e r E s t i m a t e s 13 14 95.0% W a l d C o n f i d e n c e I n t e r v a l 15 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 16 B a c k g r o u n d 0.146079 17 Beta(1) 0.00252696 18 19 * - I n d i c a t e s t h a t t h i s v a l u e is n ot c a l c u l a t e d . 20 21 22 23 Analysis of Deviance Table 24 25 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 26 Full model -94.8672 4 27 Fitted model -97.6546 2 5.57468 2 0.06158 28 Reduced model -98.0432 1 6.35197 3 0.09569 29 30 AIC: 199.309 31 32 33 G o o d n e s s of Fit 34 Scaled 35 Dose Est. Prob. Expected Observed Size Residual 36 37 0.0000 0.1461 10.518 6.000 72 -1.507 38 1.9569 0.1503 7.515 9.000 50 0.588 39 5.7027 0.1583 7.756 12.000 49 1.661 40 29.8723 0.2082 10.200 9.000 49 -0.422 41 42 Chi^2 = 5.55 d.f. = 2 P-value = 0.0622 43 44 45 Benchmark Dose Computation 46 47 Specified effect 0.01 48 49 Risk Type : Extra risk 50 51 C o n f i d e n c e l e v e l 0.95 52 53 BMD 3.97724 54 55 BMDL 1.22286 56 57 58 BMDU did not converge for BMR = 0.010000 59 BMDU calculation failed 60 BMDU = Inf 61 62 This document is a draftfor review purposes only and does not constitute Agency policy. F-38 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.13.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 09:53 04/02 2010 3 4 National T oxicology Program, 1982: Adrenal cortex: Adenom a This document is a draftfor review purposes only and does not constitute Agency policy. F-39 DRAFT--DO NOT CITE OR QUOTE 1 F.1.14. National Toxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma 2 F.1.14.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom Multistage Cancer, 1-Degree a 2 x2pValue 0.179 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 75.385 3.127E+00 1.380E+00 Notes M ultistage Cancer, 2 0.179 75.385 3.127E+00 1.380E+00 final =0 2-D egree M ultistage Cancer, 2 0.179 75.385 3.127E+00 1.380E+00 final =0 3-Degree a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.14.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ B l o o d \ 1 4 _ m s c 1 _ 1 P e r c _ s u b c u _ f i b r o . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 1 4 m s c 1 1 P e r c s u b c u f i b r o . p l t 14 Fri A p r 02 1 0 : 5 9 : 3 8 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.010477 46 Beta(1) = 0.00314237 47 This document is a draftfor review purposes only and does not constitute Agency policy. F-40 DRAFT--DO NOT CITE OR QUOTE 1 2 Asymptotic Correlation Matrix of Parameter Estimates 3 4 Background Beta(1) 5 6 Background 1 -0.55 7 8 Beta(1) -0.55 1 9 10 11 12 P a r a m e t e r E s t i m a t e s 13 14 95.0% W a l d C o n f i d e n c e I n t e r v a l 15 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 16 Background 0.0124357 17 Beta(1) 0.0029518 18 19 * - I n d i c a t e s t h a t t h i s v a l u e is n ot c a l c u l a t e d . 20 21 22 23 A n alysis of Deviance Table 24 25 Model Log(likelihood) # Param's Deviance Test d.f. P-value 26 Full model -30.9876 4 27 Fitted model -31.0692 2 0.163345 2 0.9216 28 Reduced model -34.3291 1 6.68308 3 0.08272 29 30 AIC: 66.1385 31 32 33 G o o d n e s s of Fit 34 Scaled 35 Dose Est. Prob. Expected Observed Size Residu. 36 37 0.0000 0.0124 0.920 1.000 74 0.084 38 1.9460 0.0181 0.905 1.000 50 0.101 39 5.8440 0.0293 1.408 1.000 48 - 0 . 3 4 9 40 32.0560 0.1016 4.775 5.000 47 0 . 1 0 9 41 42 Chi^2 = 0.15 d.f. = 2 P-value = 0.9274 43 44 45 Benchmark Dose Computation 46 47 Specified effect 0.01 48 49 Risk Type Extra risk 50 51 C o n f i d e n c e l e v e l 0.95 52 53 BMD 3.40481 54 55 BMDL I. 68615 56 57 BMDU II. 3501 58 59 T a k e n t o g e t h e r , (1.68615, 11.3501) is a 90 two-sided confidence 60 interval for the BMD 61 62 Multistage Cancer Slope Factor 0.00593067 63 64 This document is a draftfor review purposes only and does not constitute Agency policy. F-41 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.14.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 09:59 04/02 2010 3 4 National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma This document is a draftfor review purposes only and does not constitute Agency policy. F-42 DRAFT--DO NOT CITE OR QUOTE 1 F.1.15. N ational Toxicology Program , 1982: H em atopoietio System : L ym phom a or 2 L eukem ia 3 F .1 .1 5 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom Multistage Cancer, 1-Degree a 2 M ultistage Cancer, 2-D egree 1 M ultistage Cancer, 3-Degree 1 x2pValue 0.977 0.869 0.869 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 261.445 1.145E+00 6.091E-01 263.426 1.704E+00 6.102E-01 263.426 1.704E+00 6.102E-01 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 4 5 6 F .1.15.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 7 National T oxicology Program, 1982: H em atopoietio System: Lym phom a or Leukem ia 8 9 10 11 12 M u l t i s t a g e C a n c e r M o d e l . (Version: 1.7; Date: 0 5 / 1 6 / 2 0 0 8 ) 13 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 1 5 m s c l I P e r c m i c e f l y m p h o m a . ( d ) 14 G n u p l o t P l o t t i n g File: C : \ 1 \ B l o o d \ 1 5 m s c 1 1 P e r c m i c e f l y m p h o m a . p l t 15 F r i A p r 02 1 1 : 0 0 : 0 7 2 0 1 0 16 17 18 T a b l e 15 p a g e 64 H e m a t o p o i e t i c S y s t e m L y m p h o m a o r L e u k e m i a 19 20 21 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 22 23 P[response] = b a c k g r o u n d + (1-background)*[1-EXP( 24 -betai*dose^1)] 25 26 The parameter betas are restricted to be positive 27 28 29 Dependent variable = Mean 30 Independent variable = Dose 31 32 Total number of observations = 4 33 Total n u m b e r of records w i t h m i s s i n g va l u e s = 0 34 Total number of parameters in model = 2 35 Total number of specified p arameters = 0 36 Degree of polynomial = 1 37 38 39 Maximum number of iterations = 250 40 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 41 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1e-008 42 43 44 45 Default Initial Parameter Values 46 Background = 0.23423 47 Beta(1) = 0.00892991 This document is a draftfor review purposes only and does not constitute Agency policy. F-43 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Asymptotic Correlation Matrix of Parameter Estimates 4 5 Background Beta(1) 6 7 Background 1 -0.54 8 9 Beta(1) -0.54 1 10 11 12 13 P a r a m e t e r E s t i m a t e s 14 15 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 16 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 17 B a c k g r o u n d 0.236159 * 18 Beta(1) 0.00877894 * 19 20 * - Indicates that this value is not calculated. 21 22 23 24 Analysis of Deviance Table 25 26 Model Log(likelihood) # Param's Deviance Test d.f. P-value 27 Full model -128.699 4 28 Fitted model -128.723 2 0.0465401 2 0.977 29 Reduced model -131.412 1 5.42487 3 0.1432 30 31 AIC: 261.445 32 33 34 Goodness of Fit 35 36 Dose Est. Prob. Expected Observed Size Residual 37 38 0.0000 0.2362 17.476 18.000 74 0 . 143 39 1.9460 0.2491 12.455 12.000 50 -0.149 40 5.8440 0.2744 13.169 13.000 48 - 0 . 0 5 5 41 32.0560 0.4235 19.905 20.000 47 0 . 0 2 8 42 43 Chi^2 = 0.05 d.f. = 2 P-value = 0.9770 44 45 46 Benchmark Dose Computation 47 48 Specified effect 0.01 49 50 Risk Type Extra risk 51 52 Confidence level 0.95 53 54 BMD 1.14482 55 56 BMDL 0.609084 57 58 BMDU 4.29581 59 60 T a k e n t o g e t h e r , (0.609084, 4.29581) is a 90 two-sided confidence 61 i n t e r v a l for t he B M D 62 63 M u l t i s t a g e C a n c e r Slope Factor 0.0164181 64 65 This document is a draftfor review purposes only and does not constitute Agency policy. F-44 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.15.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 10:00 04/02 2010 3 4 National T oxicology Program, 1982: H em atopoietio System: Lym phom a or Leukem ia This document is a draftfor review purposes only and does not constitute Agency policy. F-45 DRAFT--DO NOT CITE OR QUOTE 1 F.1.16. National Toxicology Program, 1982: Liver: Hepatooellular Adenoma or Carcinoma 2 F.1.16.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom x2pValue Multistage Cancer, 1-Degree a 2 0.340 AIC 155.213 BMD BMDL (ng/kg-d) (ng/kg-d) 1.488E+00 8.265E-01 Notes M ultistage Cancer, 2 2-D egree 0.340 155.213 1.488E+00 8.265E-01 final =0 M ultistage Cancer, 2 3-Degree 0.340 155.213 1.488E+00 8.265E-01 final =0 a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.16.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 1982: Liver: H epatooellular A denom a or Carcinoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ B l o o d \ 1 6 _ m s c 1 _ 1 P e r c _ m f _ L i v A d e n C a r c . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 1 6 _ m s c 1 _ 1 P e r c _ m f _ L i v A d e n C a r c . p l t 14 _ Fri A p r 02 1 1 : 0 4 : 1 1 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.080941 46 Beta(1) = 0.00583089 47 This document is a draftfor review purposes only and does not constitute Agency policy. F-46 DRAFT--DO NOT CITE OR QUOTE 1 2 Asymptotic Correlation Matrix of Parameter Estimates 3 4 Background Beta(1) 5 6 Background 1 -0.57 7 8 Beta(1) -0.57 1 9 10 11 12 P a r a m e t e r E s t i m a t e s 13 14 95.0% W a l d C o n f i d e n c e I n t e r v a l 15 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 16 Background 0.0692161 17 Beta(1) 0.00675636 18 19 * - I n d i c a t e s t h a t t h i s v a l u e is n ot c a l c u l a t e d . 20 21 22 23 Analysis of Deviance Table 24 25 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 26 Full model -74.5177 4 27 Fitted model -75.6063 2 2.17736 2 0.3367 28 Reduced model -79.6703 1 10.3053 3 0.01614 29 30 AIC: 155.213 31 32 33 G o o d n e s s of Fit 34 Scaled 35 Dose Est. Prob. Expected Observed Size Residual 36 37 0.0000 0.0692 5.053 3.000 73 -0.947 38 1.9460 0.0814 4.069 6.000 50 0.999 39 5.8440 0.1053 5.052 6.000 48 0 . 446 40 32.0560 0.2505 11.772 11.000 47 - 0 . 2 6 0 41 42 Chi^2 = 2.16 d.f. = 2 P-value = 0.3395 43 44 45 Benchmark Dose Computation 46 47 Specified effect 0.01 48 49 Risk Type Extra risk 50 51 C o n f i d e n c e l e v e l 0.95 52 53 BMD 1.48754 54 55 BMDL 0.826482 56 57 BMDU 3.9863 58 59 T a k e n t o g e t h e r , (0.826482, 3 . 9 8 6 3 ) is a 90 two-sided confidence 60 interval for the BMD 61 62 Multistage Cancer Slope Factor 0.0120995 63 64 This document is a draftfor review purposes only and does not constitute Agency policy. F-47 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.16.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 10:04 04/02 2010 3 4 National T oxicology Program, 1982: Liver: H epatooellular A denom a or Carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-48 DRAFT--DO NOT CITE OR QUOTE 1 F.1.17. National Toxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma 2 F.1.17.1. S u m m a ry Table o f B M D S M odeling R esults Model Degrees of Freedom Multistage Cancer, 1-Degree a 2 x2pValue 0.179 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 75.385 3.127E+00 1.380E+00 Notes M ultistage Cancer, 2 0.179 75.385 3.127E+00 1.380E+00 final =0 2-D egree M ultistage Cancer, 2 0.179 75.385 3.127E+00 1.380E+00 final =0 3-Degree a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.17.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ B l o o d \ 1 7 m s c l I P e r c m i c e f t h y r o i d aden.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 1 7 m s c 1 1 P e r c m i c e f t h y r o i d a d e n . p l t 14 Fri A p r 02 1 1 : 0 4 : 3 9 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.0202346 46 Beta(1) = 0.00292833 47 This document is a draftfor review purposes only and does not constitute Agency policy. F-49 DRAFT--DO NOT CITE OR QUOTE 1 2 Asymptotic Correlation Matrix of Parameter Estimates 3 4 Background Beta(1) 5 6 Background 1 -0.58 7 8 Beta(1) -0.58 1 9 10 11 12 P a r a m e t e r E s t i m a t e s 13 14 95.0% W a l d C o n f i d e n c e I n t e r v a l 15 V a r i a b l e Estimate Std. Err. Lower Conf. Limit Upper Conf 16 Background 0.0153082 * * * 17 Beta(1) 0.00329742 * * * 18 19 - I n d i c a t e s t h a t t h i s v a l u e is not c a l c u l a t e d . 20 21 22 23 A n alysis of Deviance Table 24 25 Model Log(likelihood) # Param's Deviance Test d.f. P-value 26 Full model -32.0017 4 27 Fitted model -34.3904 2 4.77738 2 0.09175 28 Reduced model -37.2405 1 10.4776 3 0.01491 29 30 AIC: 72.7807 31 32 33 G o o d n e s s of Fit 34 Scaled 35 Dose Est. Prob. Expected Observed Size Residual 36 37 0.0000 0.0153 1.056 0.000 69 - 1 .036 38 1.9460 0.0216 1.080 3.000 50 1.867 39 5.8440 0.0341 1.603 1.000 47 - 0 . 4 8 4 40 32.0560 0.1141 5.248 5.000 46 - 0 .115 41 42 Chi^2 = 4.81 d.f. = 2 P-value = 0.0904 43 44 45 Benchmark Dose Computation 46 47 Specified effect 0.01 48 49 Risk Type Extra risk 50 51 C o n f i d e n c e l e v e l 0.95 52 53 BMD 3.04794 54 55 BMDL 1.43569 56 57 BMDU 138876 58 59 T a k e n t o g e t h e r , (1.43569, 1 3 8 8 7 6 ) is a 90 two-sided confidence 60 interval for the BMD 61 62 Multistage Cancer Slope Factor 0.00696528 63 64 This document is a draftfor review purposes only and does not constitute Agency policy. F-50 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.17.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 10:04 04/02 2010 3 4 National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma This document is a draftfor review purposes only and does not constitute Agency policy. F-51 DRAFT--DO NOT CITE OR QUOTE 1 F.1.18. National Toxicology Program, 1982: Lung: Alveolar/Bronchiolar Adenoma or 2 Carcinoma 3 F .1 .1 8 .1 . Summary Table o f BMDS Modeling Results Model Degrees of Freedom M ultistage Cancer, 1-Degree 2 Multistage Cancer, 2-Degree a 2 M ultistage Cancer, 3-Degree 2 x2pValue 0.088 0.167 0.182 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 168.342 6.499E-01 3.512E-01 166.946 2.528E+00 4.135E-01 166.799 4.147E+00 4.230E-01 Notes a Best-fitting m odel, BM D S output presented in this appendix 4 5 6 F .1.18.2. Outputfor Selected Model: Multistage Cancer, 2-Degree 7 National T oxicology Program, 1982: Lung: Alveolar/Bronchiolar A denom a or Carcinoma 8 9 10 11 12 M u l t i s t a g e C a n c e r M o d e l . (Version: 1.7; Date: 0 5 / 1 6 / 2 0 0 8 ) 13 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 1 8 m s c 2 I P e r c l u n g a d e n c a r c . ( d ) 14 G n u p l o t P l o t t i n g File: C : \ 1 \ B l o o d \ 1 8 m s c 2 1 P e r c l u n g a d e n c a r c . p l t 15 F r i A p r 02 1 1 : 0 5 : 0 9 2 0 1 0 16 17 18 0 19 20 21 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 22 23 P[response] = b a c k g r o u n d + (1-background)*[1-EXP( 24 -betai*dose^1-beta2*dose^2)] 25 26 The parameter betas are restricted to be positive 27 28 29 Dependent variable = Mean 30 Independent variable = Dose 31 32 Total number of observations = 4 33 Total n u m b e r of records w i t h m i s s i n g va l u e s = 0 34 Total number of parameters in model = 3 35 Total number of specified p arameters = 0 36 Degree of polynomial = 2 37 38 39 Maximum number of iterations = 250 40 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 41 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1e-008 42 43 44 45 Default Initial Parameter Values 46 Background = 0.0868577 47 Beta(1) = 0 This document is a draftfor review purposes only and does not constitute Agency policy. F-52 D RAFT-- DO N O T CITE OR QUOTE 1 Beta(2) = 0.00165722 2 3 4 Asymptotic Correlation Matrix of Parameter Estimates 5 6 ( *** The model parameter(s) -Beta(1) 7 have been estimated at a boundary point, or have been specified by the user, 8 and do not appear in the correlation matrix ) 9 10 Background Beta(2) 11 12 B a c k g r o u n d 1 -0.46 13 14 Beta(2) -0.46 1 15 16 17 18 P a r a m e t e r E s t i m a t e s 19 20 95.0% Wald Confidence Interval 21 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 22 Background 0.0942466 23 Beta(1) 0 24 Beta(2) 0.00157255 25 26 Indicates that this value is not calculated. 27 28 29 30 Analysis of Deviance Table 31 32 Model Log(likelihood) # Param's Deviance Test d.f. P-value 33 Full model -79. 5959 4 34 Fitted model -81. 4729 2 3.754 2 0.153 35 Reduced model -85. 3351 1 11.4782 3 0.009402 36 37 AIC: 166.946 38 39 40 Goodness of Fit 41 Scaled 42 Dose Est. Prob. Expected Observed Size Residu. 43 44 0.0000 0.0942 6.692 10.000 71 1.344 45 0.7665 0.0951 4.564 2.000 48 - 1 . 2 6 2 46 2.2711 0.1016 4.875 4.000 48 - 0 . 4 1 8 47 11.2437 0.2575 12.877 13.000 50 0.040 48 49 Chi^2 = 3.57 d.f. = 2 P-value = 0.1674 50 51 52 Benchmark Dose Computation 53 54 Specified effect 0.01 55 56 Risk Type Extra risk 57 58 Confidence level 0.95 59 60 BMD 2.52806 61 62 BMDL 0.413504 63 64 BMDU 4.19905 65 66 T a k e n t o g e t h e r , (0.413504, 4.19905) is a 90 two-sided confidence 67 interval for the BMD 68 69 Multistage Cancer Slope Factor 0.0241835 70 This document is a draftfor review purposes only and does not constitute Agency policy. F-53 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.18.3. F igure f o r Selected M odel: M ultistage Cancer, 2-Degree Multistage Cancer Model with 0.95 Confidence Level 2 10:05 04/02 2010 3 4 National T oxicology Program, 1982: Lung: Alveolar/Bronchiolar A denom a or Carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-54 DRAFT--DO NOT CITE OR QUOTE 1 F.1.19. National Toxicology Program, 1982: Liver: Hepatocellular Adenoma or Carcinoma 2 F .1 .1 9 .1 . Summary Table o f BMDS Modeling Results Model Multistage Cancer, 1-Degree a Degrees of Freedom 2 x2p Value 0.928 AIC 258.548 BMD (ng/kg-d) 2.110E-01 BMDL (ng/kgd) 1.378E01 Notes M ultistage Cancer, 2-Degree M ultistage Cancer, 3-Degree 1 0.779 260.475 3.072E-01 1.385E01 1 0.790 260.468 2.934E-01 1.385E01 a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.19.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 1982: Liver: H epatocellular A denom a or Carcinoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ B l o o d \ 1 9 m s c l I P e r c m i c e m l i v e r a d e n carc.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 1 9 m s c 1 1 P e r c m i c e m l i v e r a d e n c a r c . p l t 14 Fri A p r 02 1 1 : 0 5 : 3 6 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.201679 46 Beta(1) = 0.0486492 47 48 49 Asymptotic Correlation Matrix of Parameter Estimates This document is a draftfor review purposes only and does not constitute Agency policy. F-55 DRAFT--DO NOT CITE OR QUOTE 1 2 Background Beta(1) 3 4 Background 1 -0.53 5 6 Beta(1) -0.53 1 7 8 9 10 P a r a m e t e r E s t i m a t e s 11 12 95.0% W a l d C o n f i d e n c e I n t e r v a l 13 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 14 B a c k g r o u n d 0.204258 * 15 Beta(1) 0.0476385 * 16 17 * - I n d i c a t e s t h a t t h i s v a l u e is n ot c a l c u l a t e d . 18 19 20 21 A n a l y s i s of D e v i a n c e Table 22 23 Model Log(likelihood) # Param's Deviance Test d.f. P-value 24 Full model -127.199 4 25 Fitted model -127.274 2 0.149955 2 0.9278 26 Reduced model -135.589 1 16.7801 3 0.0007843 27 28 AIC: 258.548 29 30 31 G o o d n e s s of Fit 32 Scaled 33 Dose Est. Prob. Expected Observed Size Residual 34 35 0.0000 0.2043 14.911 15.000 73 0.026 36 0.7665 0.2328 11.407 12.000 49 0.201 37 2.2711 0.2859 14.007 13.000 49 -0.318 38 11.2437 0.5343 26.713 27.000 50 0.081 39 40 Chi^2 = 0.15 d.f. = 2 P-value = 0.9283 41 42 43 B e n chmark Dose C o mputation 44 45 Specified effect 0.01 46 47 Risk Type Extra risk 48 49 Confidence level 0.95 50 51 BMD 0.210971 52 53 BMDL 0.137771 54 55 BMDU 0.383981 56 57 T a k e n t o g e t h e r , (0.137771, 0.38 3 9 8 1 ) is a 90 two-sided confidence 58 interval for the BMD 59 60 Multistage Cancer Slope Factor 0.0725843 61 62 This document is a draftfor review purposes only and does not constitute Agency policy. F-56 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.19.3. F igure f o r Selected M odel: M ultistage Cancer, 1-Degree Multistage Cancer Model with 0.95 Confidence Level 2 10:05 04/02 2010 3 4 National T oxicology Program, 1982: Liver: H epatocellular A denom a or Carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-57 DRAFT--DO NOT CITE OR QUOTE 1 F.1.20. National Toxicology Program, 2006: Liver: Cholangiocarcinoma 2 F.1.20.1. S u m m a ry Table o f B M D S M odeling R esults Model M ultistage Cancer, 1-Degree Degrees of Freedom 5 x2pValue 0.001 AIC 138.456 BMD (ng/kg-d) 9 .4 8 1 E -0 1 M ultistage 5 0.405 119.374 4.263E+00 Cancer, 2-Degree Multistage Cancer, 3Degree a 5 0.993 113.508 7.574E+00 a Best-fitting m odel, BM D S output presented in this appendix BMDL (ng/kg-d) 7.1 1 4 E -0 1 2 .9 5 9 E + 0 0 4.133E+00 Notes 3 4 5 F .1.20.2. Outputfor Selected Model: Multistage Cancer, 3-Degree 6 National T oxicology Program, 2006: Liver: Cholangiocarcinom a 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ B l o o d \ 2 0 _ m s c 3 _ 1 P e r c _ l i v _ c h o - c a r c . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 2 0 m s c 3 1 P e r c l i v c h o - c a r c . p l t 14 Fri A p r 02 1 1 : 0 6 : 0 3 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 - b e t a i * d o s e ^ 1 - b e t a 2*dose^2-beta3*dose^3) ] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 6 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 4 34 Total number of specified parameters = 0 35 Degree of p olynomial = 3 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0 46 Beta(1) = 0 47 Beta(2) = 0 48 Beta(3)= 2.44727e-005 49 This document is a draftfor review purposes only and does not constitute Agency policy. F-58 DRAFT--DO NOT CITE OR QUOTE 1 2 Asymptotic Correlation Matrix of Parameter Estimates 3 4 ( *** The model parameter(s) -Background -Beta(1) -Beta(2) 5 have been estimated at a boundary point, or have been specified by the user, 6 and do not appear in the correlation matrix ) 7 8 Beta(3) 9 10 Beta(3) 1 11 12 13 14 P a r a m e t e r E s t i m a t e s 15 16 95.0% W a l d C o n f i d e n c e I n t e r v a l 17 V a r i a b l e Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 18 B a c k g r o u n d 0 19 Beta(1) 0 20 Beta(2) 0 21 Beta(3) 2.31301e-005 22 23 * - Indicates that this value is not calculated. 24 25 26 27 Analysis of Deviance Table 28 29 Model Log(likelihood) # Param's Deviance Test d.f. P-value 30 Full model -55.408 6 31 F i t t e d m o d e l -55.7538 1 0.691671 5 0.9834 32 Reduced model -96.9934 1 83.1708 5 <.0001 33 34 AIC: 113.508 35 36 37 Goodness of Fit 38 Scaled 39 Est. Prob. Expected Observed Residual 40 41 0.0000 0. 0 0 0 0 0. 0 0 0 0. 0 0 0 49 0. 0 0 0 42 2.5565 0. 0 0 0 4 0. 0 1 9 0. 0 0 0 48 -0. 136 43 5.6937 0. 0 0 4 3 0. 1 9 6 0. 0 0 0 46 -0. 444 44 9.7882 0. 0 2 1 5 1. 0 7 3 1. 0 0 0 50 -0. 071 45 16.5688 0. 0 9 9 9 4. 8 9 3 4. 0 0 0 49 -0. 426 46 29.6953 0. 4 5 4 3 24. 078 25. 000 53 0. 2 54 47 48 Chi^2 = 0.47 d.f. = 5 P-value = 0.9933 49 50 51 B e n c h m a r k D o s e C o m p u t a t i o n 52 53 S p e c i f i e d e f f e c t 0.01 54 55 Risk Type Extra risk 56 57 Confidence level 0.95 58 59 BMD 7.57416 60 61 BMDL 4.13304 62 63 BMDU 8.42557 64 65 T a k e n t o g e t h e r , (4.13304, 8.42557) is a 90 two-sided confidence 66 interval for the BMD 67 68 Multistage Cancer Slope Factor 0.00241953 This document is a draftfor review purposes only and does not constitute Agency policy. F-59 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.20.3. F igure f o r Selected M odel: M ultistage Cancer, 3-Degree Multistage Cancer Model with 0.95 Confidence Level 2 10:06 04/02 2010 3 4 National T oxicology Program, 2006: Liver: Cholangiocarcinom a This document is a draftfor review purposes only and does not constitute Agency policy. F-60 DRAFT--DO NOT CITE OR QUOTE 1 F.1.21. National Toxicology Program, 2006: Liver: Hepatocellular adenoma 2 F.1.21.1. S u m m a ry Table o f B M D S M odeling R esults M odel D egrees o f x2p Freedom Value AIC BMD (ng/kg-d) BMDL (ng/kg-d) N otes M ultistage Cancer, 1-Degree 5 0.026 87.024 2.192E+00 1.455E+00 M ultistage Cancer, 2-Degree Multistage Cancer, 3-Degree a 5 5 0.509 76.982 6.602E+00 4.342E+00 0.933 72.782 1.022E+01 6.527E+00 a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.21.2. Output For Selected Model: Multistage Cancer, 3-Degree 6 National T oxicology Program, 2006: Liver: H epatocellular adenoma 7 8 9 10 M u l t i s t a g e C a n c e r M o d e l . (Version: 1.7; Date: 0 5 / 1 6 / 2 0 0 8 ) 11 I n p u t D a t a F i l e : C : \ 1 \ B l o o d \ 2 1 _ m s c 3 _ 1 P e r c _ l i v _ h e p a t _ a d . ( d ) 12 G n u p l o t P l o t t i n g File: C : \ 1 \ B l o o d \ 2 1 _ m s c 3 _ 1 P e r c _ l i v _ h e p a t _ a d . p l t 13 _ F r i A p r 02 1 1 : 0 6 : 3 2 2 0 1 0 14 15 16 0 17 18 19 T he f o r m of t he p r o b a b i l i t y f u n c t i o n is: 20 21 P[response] = b a c k g r o u n d + (1-backgroun d ) * [ 1 - E X P ( 22 -betai*dose^1-beta2*dose^2-beta3*dose^3)] 23 24 The parameter betas are restricted to be positive 25 26 27 Dependent variable = Mean 28 Independent variable = Dose 29 30 Total number of observations = 6 31 T o t a l n u m b e r of r e c o r d s w i t h m i s s i n g v a l u e s = 0 32 Total number of parameters in model = 4 33 Total n u m b e r of s p e c i f i e d p a r a m e t e r s = 0 34 Degree of polynomial = 3 35 36 37 Maximum number of iterations = 250 38 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 39 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 40 41 42 43 Default Initial Parameter Values 44 Background = 0 45 Beta(1) = 0 46 Beta(2) = 0 47 Beta(3)= 1.08896e-005 48 49 Asymptotic Correlation Matrix of Parameter Estimates 50 51 ( *** T h e m o d e l p a r a m e t e r ( s ) - B a c k g r o u n d -Beta(1) -Beta(2) This document is a draftfor review purposes only and does not constitute Agency policy. F-61 DRAFT--DO NOT CITE OR QUOTE 1 have been estimated at a boundary point, or have been specified by the user, 2 and do not appear in the correlation matrix ) 3 4 Beta(3) 5 6 Beta(3) 1 7 8 9 10 P a r a m e t e r E s t i m a t e s 11 12 95.0% W a l d C o n f i d e n c e I n t e r v a l 13 V a r i a b l e Estimate Std. Err. Lower Conf. Limit Upper Conf Limit 14 B a c k g r o u n d 0* * * 15 B e t a ( 1 ) 0* * * 16 Beta(2) 0* * * 17 Beta(3) 9.41228e-006 * * * 18 19 - I n d i c a t e s t h a t t h i s v a l u e is not c a l c u l a t e d . 20 21 22 23 Analysis of Deviance Table 24 25 Model Log(likelihood) # Param's Deviance Test d.f. P-value 26 Full model -34.4075 6 27 Fitted model -35.3907 1 1.96648 5 0.8538 28 Reduced model -56.3333 1 43.8515 5 <.0001 29 30 AIC: 72.7815 31 32 33 G o o d n e s s of Fit 34 Scaled 35 Dose Est. Prob. Expected Observed Residual 36 37 0.0000 0. 0 0 0 0 0. 0 0 0 0. 0 0 0 49 0. 0 0 0 38 2.5565 0. 0 0 0 2 0. 008 0. 0 0 0 48 -0. 087 39 5.6937 0. 0 0 1 7 0. 0 8 0 0. 0 0 0 46 -0. 283 40 9.7882 0. 0 0 8 8 0. 4 3 9 0. 0 0 0 50 -0. 666 41 16.5688 0. 0 4 1 9 2. 054 1. 0 0 0 49 -0. 751 42 29.6953 0. 2 1 8 4 11. 577 13. 000 53 0. 4 7 3 43 44 Chi^2 = 1.32 d.f. = 5 P-value = 0.9330 45 46 47 Benchmark Dose Computation 48 49 Specified effect 0.01 50 51 R i s k T y p e Extra risk 52 53 C o n f i d e n c e level 0.95 54 55 BMD 10.221 56 57 BMDL 6.52683 58 59 BMDU 11.9754 60 61 T a k e n t o g e t h e r , (6.52683, 11.97 5 4 ) is a 90 two-sided confidence 62 interval for the BMD 63 64 Multistage Cancer Slope Factor 0.00153214 65 66 This document is a draftfor review purposes only and does not constitute Agency policy. F-62 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.21.3. F ig u r e F o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 3 -D e g r e e Multistage Cancer Model with 0.95 Confidence Level 2 10:06 04/02 2010 3 4 National T oxicology Program, 2006: Liver: H epatocellular adenoma This document is a draftfor review purposes only and does not constitute Agency policy. F-63 DRAFT--DO NOT CITE OR QUOTE 1 F.1.22. National Toxicology Program, 2006: Oral mucosa: squamous cell carcinoma 2 F.1.22.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Multistage Cancer, 1Degree a Degrees of Freedom x2p Value AIC BMD (ng/kg-d) 4 0.270 126.963 2.204E+00 M ultistage Cancer, 2-Degree 4 0.538 123.896 7.108E+00 M ultistage Cancer, 3-Degree 4 0.565 123.295 1.103E+01 a Best-fitting m odel, BM D S output presented in this appendix BMDL (ng/kg-d) 1.389E+00 2 .1 5 8 E + 0 0 2 .2 9 8 E + 0 0 Notes 3 4 5 F .1.22.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 2006: Oral mucosa: squamous cell carcinoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ B l o o d \ 2 2 _ m s c 1 _ 1 P e r c _ o r a l _ c a r c . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 2 2 _ m s c 1 _ 1 P e r c _ o r a l _ c a r c . p l t 14 Fri A p r 02 1 1 : 0 7 : 0 0 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 6 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0 46 Beta(1) = 0.00629243 47 48 49 Asymptotic Correlation Matrix of Parameter Estimates This document is a draftfor review purposes only and does not constitute Agency policy. F-64 DRAFT--DO NOT CITE OR QUOTE 1 2 Background Beta(1) 3 4 Background 1 -0.67 5 6 Beta(1) -0.67 1 7 8 9 10 P a r a m e t e r E s t i m a t e s 11 12 95.0% W a l d C o n f i d e n c e I n t e r v a l 13 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 14 Background 0.0139169 15 Beta(1) 0.00456055 16 17 * - I n d i c a t e s t h a t t h i s v a l u e is n ot c a l c u l a t e d . 18 19 20 21 A n a l y s i s of D e v i a n c e Table 22 23 Model Log(likelihood) # Param's Deviance Test d.f. P-value 24 Full model -57.5353 6 25 Fitted model -61.4815 2 7.89233 4 0.0956 26 Reduced model -67.7782 1 20.4858 5 0.001013 27 28 AIC: 126.963 29 30 31 G o o d n e s s of Fit 32 Scaled 33 Dose Est. Prob. Expected Observed Size Residu 34 35 0.0000 0.0139 0. 682 1.000 49 0.388 36 2.5565 0.0253 1. 217 2.000 48 0 . 719 37 5.6937 0.0392 1. 8 0 3 1.000 46 - 0 .610 38 9.7882 0.0570 2. 848 0.000 50 -1.738 39 16.5688 0.0857 4. 1 9 8 4.000 49 -0.101 40 29.6953 0.1388 7. 357 10.000 53 1.050 41 42 i^2 = 5.17 d.f. = 4 P-value = 0.2700 43 44 45 Benchmark Dose Computation 46 47 Specified effect 0.01 48 49 Risk Type Extra risk 50 51 C o n f i d e n c e l e v e l 0.95 52 53 BMD 2.20376 54 55 BMDL 1.38901 56 57 BMDU 4.3103 58 59 T a k e n t o g e t h e r , (1.38901, 4 . 3 1 0 3 ) is a 90 two-sided confidence 60 interval for the BMD 61 62 Multistage Cancer Slope Factor 0.00719939 63 64 This document is a draftfor review purposes only and does not constitute Agency policy. F-65 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.22.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 10:07 04/02 2010 3 4 National T oxicology Program, 2006: Oral mucosa: squamous cell carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-66 DRAFT--DO NOT CITE OR QUOTE 1 F.1.23. National Toxicology Program, 2006: Pancreas: adenoma or carcinoma 2 F.1.23.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Multistage Cancer, 1Degree a Degrees of Freedom x2p Value 5 0.640 AIC BMD (ng/kg-d) 29.373 1.052E+01 M ultistage Cancer, 2-Degree 5 0.929 27.061 1.458E+01 M ultistage Cancer, 3-Degree 5 0.986 25.972 1.739E+01 a Best-fitting m odel, BM D S output presented in this appendix BMDL (ng/kg-d) 4.630E+00 7 .2 2 7 E + 0 0 9 .3 7 3 E + 0 0 Notes 3 4 5 F .1.23.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 2006: Pancreas: adenoma or carcinoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ B l o o d \ 2 3 m s c l I P e r c p a n e ad carc.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 2 3 m s c l l P e r c p a n c a d c a r c . p l t 14 Fri A p r 02 1 1 : 0 7 : 2 9 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 6 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0 46 Beta(1) = 0.00191132 47 48 49 Asymptotic Correlation Matrix of Parameter Estimates This document is a draftfor review purposes only and does not constitute Agency policy. F-67 DRAFT--DO NOT CITE OR QUOTE 1 2 ( *** The model parameter(s) -Background 3 have been estimated at a boundary point, or have been specified by the user, 4 and do not appear in the correlation matrix ) 5 6 Beta(1) 7 8 Beta(1) 1 9 10 11 12 P a r a m e t e r E s t i m a t e s 13 14 95.0% W a l d C o n f i d e n c e I n t e r v a l 15 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 16 B a c k g r o u n d 0 17 B eta(1) 0 . 0 0 0 9 5 5 6 6 2 18 19 * - I n d i c a t e s t h a t t h i s v a l u e is not c a l c u l a t e d . 20 21 22 23 A n alysis of Deviance Table 24 25 Model Log(likelihood) # Param's Deviance Test d.f. P-value 26 Full model -11.4096 6 27 Fitted model -13.6865 1 4.55375 5 0.4727 28 Reduced model -16.7086 1 10.598 5 0.05996 29 30 AIC: 29.373 31 32 33 G o o d n e s s of Fit 34 Scaled 35 Dose Est. Prob. Expected Observed Size Residual 36 37 0. 0 0 0 0 0. 0 0 0 0 0. 0 0 0 0.000 48 -0. 0 0 7 38 2. 5 5 6 5 0. 0 0 2 4 0. 117 0.000 48 - 0 . 3 4 3 39 5. 6 9 3 7 0. 0 0 5 4 0. 2 5 0 0.000 46 -0.501 40 9. 7 8 8 2 0. 0 0 9 3 0. 4 6 6 0.000 50 -0.686 41 16. 5688 0. 0 1 5 7 0. 754 0.000 48 - 0 . 8 7 5 42 29. 6953 0. 0 2 8 0 1. 427 3.000 51 1.336 43 44 Chi^2 = 3.39 d.f. = 5 P-value = 0.6403 45 46 47 Benchmark Dose Computation 48 49 Specified effect 0.01 50 51 R i s k T y p e Extra risk 52 53 C o n f i d e n c e level 0.95 54 55 BMD 10.5166 56 57 BMDL 4.62967 58 59 BMDU 32.8573 60 61 T a k e n t o g e t h e r , (4.62967, 32.8573) is a 90 two-sided confidence 62 interval for the BMD 63 64 Multistage Cancer Slope Factor 0.00215998 65 66 This document is a draftfor review purposes only and does not constitute Agency policy. F-68 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.23.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 10:07 04/02 2010 3 4 National T oxicology Program, 2006: Pancreas: adenoma or carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-69 DRAFT--DO NOT CITE OR QUOTE 1 F.1.24. National Toxicology Program, 2006: Lung: Cystic keratinizing epithelioma 2 F.1.24.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model M ultistage Cancer, 1-Degree Multistage Cancer, 2-Degree a Degrees of Freedom 5 5 x2p Value 0.062 0.507 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 64.034 3.445E+00 2.084E+00 56.943 8.304E+00 5.245E+00 Notes M ultistage Cancer, 3-Degree 5 0.845 53.558 1.193E+01 7.765E+00 a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.24.2. Outputfor Selected Model: Multistage Cancer, 2-Degree 6 National T oxicology Program, 2006: Lung: Cystic keratinizing epitheliom a 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ B l o o d \ 2 4 _ m s c 2 _ 1 P e r c _ l u n g _ e p i t h . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 2 4 _ m s c 2 _ 1 P e r c _ l u n g _ e p i t h . p l t 14 Fri A p r 02 1 1 : 0 7 : 5 7 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1-beta2*dose^2)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 6 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 3 34 Total number of specified parameters = 0 35 Degree of p olynomial = 2 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0 46 Beta(1) = 0 47 Beta(2) = 0.000216412 48 49 This document is a draftfor review purposes only and does not constitute Agency policy. F-70 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 ( *** The model parameter(s) -Background -Beta(1) 4 have been estimated at a boundary point, or have been specified by the user, 5 and do not appear in the correlation matrix ) 6 7 Beta(2) 8 9 Beta(2) 1 10 11 12 13 P a r a m e t e r E s t i m a t e s 14 15 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 16 V a r i a b l e Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 17 B a c k g r o u n d 0 18 B e t a(1) 0 19 Beta(2) 0.000145744 20 21 * - In d i c a t e s that this v a l u e is not calculated. 22 23 24 25 Analysis of Deviance Table 26 27 Model Log(likelihood) # Param's Deviance Test d.f. P-value 28 Full model -23.958 6 29 Fitted model -27.4714 1 7.02662 5 0.2187 30 Reduced model -40.2069 1 32.4976 5 <.0001 31 32 AIC: 56.9427 33 34 35 Goodness of Fit 36 Scaled 37 Dose Est. Prob. Expected Observed Size Residual 38 39 0.0000 0.0000 0. 0 0 0 0. 0 0 0 49 0 000 40 2.5565 0.0010 0. 0 4 6 0. 0 0 0 48 -0 214 41 5.6937 0.0047 0. 217 0. 0 0 0 46 -0 467 42 9.7882 0.0139 0. 6 7 9 0. 0 0 0 49 -0 830 43 16.5688 0.0392 1. 922 0. 0 0 0 49 -1 414 44 29.6953 0.1206 6. 2 7 1 9. 0 0 0 52 1. 1 6 2 45 46 Chi^2 = 4.30 d.f. = 5 P-value = 0.5067 47 48 49 Benchmark Dose Computation 50 51 S p e c i f i e d e f f e c t 0.01 52 53 R i s k Type Extra risk 54 55 C o n f i d e n c e level 0.95 56 57 BMD 8.30415 58 59 BMDL 5.24499 60 61 BMDU 11.2298 62 63 T a k e n t o g e t h e r , (5.24499, 11.22 9 8 ) is a 90 two-sided confidence 64 interval for the BMD 65 66 Multistage Cancer Slope Factor 0.00190658 67 68 This document is a draftfor review purposes only and does not constitute Agency policy. F-71 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.24.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 2 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 10:07 04/02 2010 3 4 National T oxicology Program, 2006: Lung: Cystic keratinizing epitheliom a This document is a draftfor review purposes only and does not constitute Agency policy. F-72 DRAFT--DO NOT CITE OR QUOTE 1 F.1.25. Toth et al., 1979: Liver: Tumors 2 F.1.25.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of Freedom Multistage Cancer, 1-Degree a 1 x2pValue 0.293 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 155.740 3.684E-01 2.096E-01 Notes M ultistage Cancer, 2-D egree 1 0.293 155.740 3.684E-01 2.096E-01 final =0 a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.25.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 Toth et al., 1979: Liver: Tumors 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ B l o o d \ 2 5 m s c l I P e r c ad r co r 1yr.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 2 5 m s c 1 1 P e r c a d r c o r 1 y r . p l t 14 Fri A p r 02 1 1 : 0 8 : 2 6 2010 15 16 17 T a b l e 1 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -beta1*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 1 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.234952 46 Beta(1) = 0.0269892 47 48 49 Asymptotic Correlation Matrix of Parameter Estimates 50 51 Background Beta(1) This document is a draftfor review purposes only and does not constitute Agency policy. F-73 D RAFT-- DO N O T CITE OR QUOTE 1 2 Background 1 -0.55 3 4 Beta(1) -0.55 1 5 6 7 8 Parameter Estimates 9 10 95.0% W a l d C o n f i d e n c e I n t e r v a l 11 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 12 B a c k g r o u n d 0.235297 13 Beta(1) 0.0272796 14 15 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 16 17 18 19 A n a l y s i s of D e v i a n c e T a b l e 20 21 Model Log(likelihood) # Param's Deviance Test d.f. P-value 22 Full model -75.3127 3 23 Fitted model -75.8702 2 1.11506 1 0.291 24 Reduced model -79.4897 1 8.35401 2 0.01534 25 26 AIC: 155.74 27 28 29 Goodness of Fit 30 Scaled 31 Dose Est. Prob. Expected Observed Size Residual 32 33 0.0000 0.2353 8.941 7.000 38 -0.742 34 0.5732 0.2472 10.875 13.000 44 0 . 7 4 3 35 14.2123 0.4811 21.167 21.000 44 - 0 . 0 5 0 36 37 Chi^2 = 1.11 d.f. = 1 P-value = 0.2931 38 39 40 Benchmark Dose Computation 41 42 Specified effect 0.01 43 44 Risk Type Extra risk 45 46 Confidence level 0.95 47 48 BMD 0.368419 49 50 BMDL 0.209642 51 52 BMDU 1.01064 53 54 T a k e n t o g e t h e r , (0.209642, 1.01064) is a 90 two-sided confidence 55 inter v a l for the BMD 56 57 Multistage Cancer Slope Factor 0.0477004 58 59 This document is a draftfor review purposes only and does not constitute Agency policy. F-74 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.25.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 10:08 04/02 2010 3 4 Toth et al., 1979: Liver: Tumors This document is a draftfor review purposes only and does not constitute Agency policy. F-75 DRAFT--DO NOT CITE OR QUOTE 1 F.1.26. Della Porta et al., 1987: Table 4, B6C3 mice, male, hepatocellular carcinoma 2 F.1.26.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) M ultistage Cancer, 1-Degree Multistage Cancer, 2-Degree a 1 1 0.036 165.333 9.239E-01 6.933E-01 0.525 161.217 7.143E+00 1.170E+00 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.26.2. Outputfor Selected Model: Multistage Cancer, 2-Degree 6 D ella Porta et al., 1987: Table 4, B6C 3 m ice, m ale, hepatocellular carcinoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ B l o o d \ 9 4 _ D P o r t a _ 1 9 8 7 _ M a l e _ H e p _ C a r c _ M u l t i C a n c 2 _ 1 . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 9 4 _ D P o r t a _ 1 9 8 7 _ M a l e _ H e p _ C a r c _ M u l t i C a n c 2 _ 1 . p l t 14 Fri A p r 02 1 3 : 5 2 : 2 1 2010 15 16 17 T a b l e 4, B 6 C 3 m i c e , M a l e , H e p a t o c e l l u l a r c a r c i n o m a 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1-beta2*dose^2)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = DichEff 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 3 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 3 34 Total number of specified parameters = 0 35 Degree of p olynomial = 2 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.0865895 46 Beta(1) = 0 47 Beta(2) = 0.000211877 48 49 50 Asymptotic Correlation Matrix of Parameter Estimates 51 This document is a draftfor review purposes only and does not constitute Agency policy. F-76 DRAFT--DO NOT CITE OR QUOTE 1 ( *** The model parameter(s) -Beta(1) 2 have been estimated at a boundary point, or have been specified by the user, 3 and do not appear in the correlation matrix ) 4 5 Background Beta(2) 6 7 Background 1 -0.64 8 9 Beta(2) -0.64 1 10 11 12 13 P a r a m e t e r E s t i m a t e s 14 15 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 16 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 17 B a c k g r o u n d 0.107218 18 B e t a(1) 0 19 Beta(2) 0.00019698 20 21 In d i c a t e s that this v a l u e is not calculated. 22 23 24 25 Analysis of Deviance Table 26 27 Model Log(likelihood) # Param's Deviance Test d.f. P-value 28 Full model -78.4036 3 29 Fitted model -78.6083 2 0.409345 1 0.5223 30 Reduced model -94.7394 1 32.6717 2 <.0001 31 32 AIC: 161.217 33 34 35 Goodness of Fit 36 Scaled 37 Dose Est. Prob. Expected Observed Size Residual 38 39 0.0000 0.1072 4.610 5.000 43 0.192 40 37.9990 0.3282 16.740 15.000 51 - 0 .519 41 67.7695 0.6387 31.936 33.000 50 0.313 42 43 Chi^2 = 0.40 d.f. = 1 P-value = 0.5249 44 45 46 Benchmark Dose Computation 47 48 Specified effect 0.01 49 50 Risk Type Extra risk 51 52 Confidence level 0.95 53 54 BMD 7.14298 55 56 BMDL 1.16991 57 58 BMDU 8.58118 59 60 T a k e n t o g e t h e r , (1.16991, 8.58118) is a 90 two-sided confidence 61 i n t e r v a l for t he B M D 62 63 M u l t i s t a g e C a n c e r Slope Factor 0.0085477 64 65 This document is a draftfor review purposes only and does not constitute Agency policy. F-77 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.26.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 2 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 12:52 04/02 2010 3 4 D ella Porta et al., 1987: Table 4, B6C 3 m ice, m ale, hepatocellular carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-78 DRAFT--DO NOT CITE OR QUOTE 1 F.1.27. Della Porta et al., 1987: Table 4, B6C3 mice, female, hepatocellular adenoma 2 F.1.27.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value M ultistage Cancer, 1-Degree Multistage Cancer, 2-Degree a 1 1 0.380 0.863 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 99.614 3.599E+00 2.186E+00 98.833 1.449E+01 2.342E+00 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.27.2. Outputfor Selected Model: Multistage Cancer, 2-Degree 6 D ella Porta et al., 1987: Table 4, B6C 3 m ice, fem ale, hepatocellular adenoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ B l o o d \ 9 5 _ D P o r t a _ 1 9 8 7 _ F e m a l e _ H e p _ A d e n _ M u l t i C a n c 2 _ 1 . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 9 5 _ D P o r t a _ 1 9 8 7 _ F e m a l e _ H e p _ A d e n _ M u l t i C a n c 2 _ 1 . p l t 14 Fri A p r 02 1 3 : 5 2 : 5 1 2010 15 16 17 T a b l e 4, B 6 C 3 m i c e , F e m a l e , H e p a t o c e l l u l a r a d e n o m a 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1-beta2*dose^2)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = DichEff 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 3 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 3 34 Total number of specified parameters = 0 35 Degree of p olynomial = 2 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.0364319 46 Beta(1) = 0 47 Beta(2) = 4.92861e-005 48 49 50 Asymptotic Correlation Matrix of Parameter Estimates 51 This document is a draftfor review purposes only and does not constitute Agency policy. F-79 DRAFT--DO NOT CITE OR QUOTE 1 ( *** The model parameter(s) -Beta(1) 2 have been estimated at a boundary point, or have been specified by the user, 3 and do not appear in the correlation matrix ) 4 5 Background Beta(2) 6 7 Background 1 -0.69 8 9 Beta(2) -0.69 1 10 11 12 13 P a r a m e t e r E s t i m a t e s 14 15 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 16 V a r i a b l e Estimate Std. Err. Lower Conf. Limit Upper Conf 17 Background 0.0392633 * * * 18 B e t a(1) 0* * * 19 Beta(2) 4.78928e-005 20 21 - In d i c a t e s that this v a l u e is not calculated. 22 23 24 25 Analysis of Deviance Table 26 27 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 28 Full model -47.4015 3 29 Fitted model -47.4165 2 0.0299957 1 0.8625 30 Reduced model -51.6367 1 8.47042 2 0.01448 31 32 AIC: 98.8329 33 34 35 Goodness of Fit 36 Scaled 37 Dose Est. Prob. Expected Observed Size Residual 38 39 0.0000 0.0393 1.924 2.000 49 0.056 40 37.5865 0.1021 4.289 4.000 42 -0. 1 4 7 41 66.9741 0.2250 10.800 11.000 48 0 . 069 42 43 Chi^2 = 0.03 d.f. = 1 P-value = 0.8634 44 45 46 Benchmark Dose Computation 47 48 Specified effect 0.01 49 50 Risk Type Extra risk 51 52 Confidence level 0.95 53 54 BMD 14.4862 55 56 BMDL 2.3421 57 58 BMDU 22.1663 59 60 T a k e n t o g e t h e r , (2.3421 , 22.1663) is a 90 two-sided confidence 61 i n t e r v a l for t he B M D 62 63 M u l t i s t a g e C a n c e r Slope Factor 0.00426967 64 65 This document is a draftfor review purposes only and does not constitute Agency policy. F-80 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.27.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 2 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 12:52 04/02 2010 3 4 D ella Porta et al., 1987: Table 4, B6C 3 m ice, fem ale, hepatocellular adenoma This document is a draftfor review purposes only and does not constitute Agency policy. F-81 DRAFT--DO NOT CITE OR QUOTE 1 F.1.28. Della Porta et al., 1987: Table 4, B6C3 mice, female, hepatocellular carcinoma 2 F.1.28.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 1 0.019 M ultistage Cancer, 2-D egree 1 0.019 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 115.539 2.302E+00 1.545E+00 115.539 2.302E+00 1.545E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .1.28.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 D ella Porta et al., 1987: Table 4, B6C 3 m ice, fem ale, hepatocellular carcinoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ B l o o d \ 9 6 _ D P o r t a _ 1 9 8 7 _ F e m a l e _ H e p _ C a r c _ M u l t i C a n c 1 _ 1 . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ B l o o d \ 9 6 _ D P o r t a _ 1 9 8 7 _ F e m a l e _ H e p _ C a r c _ M u l t i C a n c 1 _ 1 . p l t 14 Fri A p r 02 1 3 : 5 3 : 2 0 2010 15 16 17 T a b l e 4, B 6 C 3 m i c e , F e m a l e , H e p a t o c e l l u l a r c a r c i n o m a 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -beta1*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = DichEff 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 3 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.0787329 46 Beta(1) = 0.00304814 47 48 49 Asymptotic Correlation Matrix of Parameter Estimates 50 51 Background Beta(1) This document is a draftfor review purposes only and does not constitute Agency policy. F-82 DRAFT--DO NOT CITE OR QUOTE 1 2 Background 1 -0.8 3 4 Beta(1) -0.8 1 5 6 7 8 Parameter Estimates 9 10 95.0% W a l d C o n f i d e n c e I n t e r v a l 11 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 12 Background 0.0268873 13 Beta(1) 0.00436529 14 15 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 16 17 18 19 A n a l y s i s of D e v i a n c e T a b l e 20 21 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 22 Full model -53.1726 3 23 Fitted model -55.7697 2 5.19425 1 0.02266 24 Reduced model -60.7146 1 15.084 2 0.0005303 25 26 AIC: 115.539 27 28 29 Goodness of Fit 30 Scaled 31 Dose Est. Prob. Expected Observed Size Residual 32 33 0.0000 0.0269 1.317 1.000 49 - 0 .280 34 37.5865 0.1741 7.314 12.000 42 1.907 35 66.9741 0.2736 13.131 9.000 48 - 1 . 3 3 8 36 37 Chi^2 = 5.50 d.f. = 1 P-value = 0.0190 38 39 40 Benchmark Dose Computation 41 42 Specified effect 0.01 43 44 Risk Type Extra risk 45 46 Confidence level 0.95 47 48 BMD 2.30233 49 50 BMDL 1.54479 51 52 BMDU 4.37768 53 54 T a k e n t o g e t h e r , (1.54479, 4.37768) is a 90 two-sided confidence 55 inter v a l for the BMD 56 57 Multistage Cancer Slope Factor 0.00647339 58 59 This document is a draftfor review purposes only and does not constitute Agency policy. F-83 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.1.28.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 12:53 04/02 2010 3 4 D ella Porta et al., 1987: Table 4, B6C 3 m ice, fem ale, hepatocellular carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-84 DRAFT--DO NOT CITE OR QUOTE F.2. ADMINISTERED DOSE BMDS RESULTS 1 F.2.1. K ociba et al., 1978: Stratified squam ous cell carcinom a o f hard palate or nasal 2 turbinates 3 F .2 .1 .1 . Summary Table o f BMDS Modeling Results Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 3 0.928 M ultistage Cancer, 2-D egree 3 0.998 M ultistage Cancer, 3-Degree 3 1.000 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 30.745 1.344E+01 6.515E+00 29.961 3.490E+01 7.216E+00 29.885 4.941E+01 7.297E+00 Notes a Best-fitting m odel, BM D S output presented in this appendix 4 5 6 F .2.1.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 7 K ociba et al., 1978: Stratified squam ous cell carcinom a o f hard palate or nasal turbinates 8 9 10 11 12 M u l t i s t a g e C a n c e r M o d e l . (Version: 1.7; Date: 0 5 / 1 6 / 2 0 0 8 ) 13 I n p u t D a t a F i l e : C : \ C a n c \ 1 m s c l I P e r c p a l a t e n a s a l . ( d ) 14 G n u p l o t P l o t t i n g File: C : \ C a n c \ 1 m s c l l P e r c p a l a t e n a s a l . p l t 15 T h u A p r 01 1 2 : 4 7 : 4 0 2 0 1 0 16 17 18 S o u r c e - T a b l e 4 19 20 21 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 22 23 P[response] = b a c k g r o u n d + (1-background)*[1-EXP( 24 -betai*dose^i)] 25 26 The parameter betas are restricted to be positive 27 28 29 Dependent variable = Mean 30 Independent variable = Dose 31 32 Total number of observations = 4 33 Total n u m b e r of records w i t h m i s s i n g va l u e s = 0 34 Total number of parameters in model = 2 35 Total number of specified p arameters = 0 36 Degree of polynomial = 1 37 38 39 Maximum number of iterations = 250 40 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 41 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1e-008 42 43 44 45 Default Initial Parameter Values This document is a draftfor review purposes only and does not constitute Agency policy. F-85 DRAFT--DO NOT CITE OR QUOTE 1 Background 0 2 Beta(1) 0.000858074 3 4 5 Asymptotic Correlation Matrix of Parameter Estimates 6 7 ( *** The model parameter(s) -Background 8 have been estimated at a boundary point, or have been specified by the user, 9 and do not appear in the correlation matrix ) 10 11 B e t a ( 1 ) 12 13 B e t a ( 1 ) 1 14 15 16 17 P a r a m e t e r E s t i m a t e s 18 19 95.0% W a l d C o n f i d e n c e I n t e r v a l 20 Variable Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 21 B a c k g r o u n d 0 22 Beta(1) 0.00074801 23 24 * - I n d i c a t e s that this v a l u e is not c a l culated. 25 26 27 28 Analysis of Deviance Table 29 30 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 31 Full model -13.9385 4 32 Fitted model -14.3726 1 0.868297 3 0.8331 33 R e d u c e d model -20.2589 1 12.6409 3 0.005481 34 35 AIC: 30.7452 36 37 38 Goodness of Fit 39 Scaled 40 Dose Est. Prob. Expected Observed Size Residu 41 42 0.0000 0.0000 0.000 0.000 85 0.000 43 1.0000 0.0007 0.037 0.000 50 -0.193 44 10.0000 0.0075 0.373 0.000 50 -0.613 45 100.0000 0.0721 3.604 4.000 50 0.217 46 47 Chi^2 = 0.46 d.f. = 3 P-value = 0.9276 48 49 50 Benchmark Dose Computation 51 52 Specified effect 0.01 53 54 Risk Type Extra risk 55 56 Confidence level 0.95 57 58 BMD 13.4361 59 60 BMDL 6.51522 61 62 BMDU 34.829 63 64 T a k e n t o g e t h e r , (6.51522, 3 4 . 8 2 9 ) is a 90 two-sided confidence 65 interval for the BMD 66 67 Multistage Cancer Slope Factor 0.00153487 68 This document is a draftfor review purposes only and does not constitute Agency policy. F-86 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.1.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:47 04/01 2010 3 4 K ociba et al., 1978: Stratified squam ous cell carcinom a o f hard palate or nasal turbinates This document is a draftfor review purposes only and does not constitute Agency policy. F-87 DRAFT--DO NOT CITE OR QUOTE 1 F.2.2. Kociba et al., 1978: Stratified squamous cell carcinoma of tongue 2 F.2.2.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 2 0.451 M ultistage Cancer, 2-D egree 2 0.451 M ultistage Cancer, 3-Degree 2 0.451 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 48.368 1.742E+01 7.146E+00 48.368 1.742E+01 7.146E+00 final =0 48.368 1.742E+01 7.146E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.2.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 K ociba et al., 1978: Stratified squam ous cell carcinom a o f tongue 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 2 m s c l I P e r c t o n g u e . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 2 m s c l l P e r c t o n g u e . p l t 14 T h u A p r 01 1 2 : 4 8 : 1 6 2010 15 16 17 S o u r c e - T a b l e 4 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.0113883 46 Beta(1) = 0.000508703 47 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-88 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 Background Beta(1) 4 5 Background 1 -0.52 6 7 Beta(1) -0.52 1 8 9 10 11 P a r a m e t e r E s t i m a t e s 12 13 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 14 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 15 B a c k g r o u n d 0.00809154 16 Beta(1) 0.000576915 17 18 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 19 20 21 22 Analysis of Deviance Table 23 24 Model Log(likelihood) # Param's Deviance Test d.f. P-value 25 Full model -21.1523 4 26 Fitted model -22.1838 2 2.06309 2 0.3565 27 Reduced model -24.1972 1 6.08976 3 0.1073 28 29 AIC: 48.3677 30 31 32 Goodness of Fit 33 Scaled 34 Dose Est. Prob. Expected Observed Size Residual 35 36 0.0000 0.0081 0.688 0.000 85 - 0 .833 37 1.0000 0.0087 0.433 1.000 50 0.865 38 10.0000 0.0138 0.690 1.000 50 0.376 39 100.0000 0.0637 3.185 3.000 50 -0.107 40 41 Chi^2 = 1.59 d.f. = 2 P-value = 0.4506 42 43 44 Benchmark Dose Computation 45 46 Specified effect = 0.01 47 48 Risk Type = Extra risk 49 50 Confidence level = 0.95 51 52 BMD = 17.4208 53 54 BMDL = 7.14637 55 56 BMDU = 3.20359e+006 57 58 T a k e n t ogether, (7.14637, 3 . 2 0 3 5 9 e + 0 0 6 ) is a 90 % two-sided confidence 59 interval for the BMD 60 61 M u l t i s t a g e C a n c e r S l o p e F a c t o r = 0.00139931 62 63 This document is a draftfor review purposes only and does not constitute Agency policy. F-89 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.2.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:48 04/01 2010 3 4 K ociba et al., 1978: Stratified squam ous cell carcinom a o f tongue This document is a draftfor review purposes only and does not constitute Agency policy. F-90 DRAFT--DO NOT CITE OR QUOTE 1 F.2.3. Kociba et al., 1978: Adenoma of adrenal cortex 2 F.2.3.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 3 0.376 M ultistage Cancer, 2-D egree 3 0.376 M ultistage Cancer, 3-Degree 3 0.376 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 53.518 7.587E+00 4.317E+00 53.518 7.587E+00 4.317E+00 final =0 53.518 7.587E+00 4.317E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.3.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 K ociba et al., 1978: A denom a o f adrenal cortex 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 3 m s c l I P e r c a d r e a d e n o m a . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 3 m s c l l P e r c a d r e a d e n o m a . p l t 14 T h u A p r 01 1 2 : 4 8 : 5 2 2010 15 16 17 S o u r c e - T a b l e 5 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.00927818 46 Beta(1) = 0.00098105 47 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-91 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 ( *** The model parameter(s) -Background 4 have been estimated at a boundary point, or have been specified by the user, 5 and do not appear in the correlation matrix ) 6 7 Beta(1) 8 9 Beta(1) 1 10 11 12 13 P a r a m e t e r E s t i m a t e s 14 15 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 16 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 17 B a c k g r o u n d 0 18 Beta(1) 0.00132464 19 20 * - I n d i c a t e s that this v a l u e is not c a l culated. 21 22 23 24 Analysis of Deviance Table 25 26 Model Log(likelihood) # Param's Deviance Test d.f. P-value 27 Full model -24.6514 4 28 Fitted model -25.759 1 2.2152 3 0.529 29 Reduced model -31.4904 1 13.6781 3 0.003378 30 31 AIC: 53.5179 32 33 34 Goodness of Fit 35 Scaled 36 Dose Est. Prob. Expected Observed Size Residu. 37 38 0.0000 0.0000 0.000 0.000 85 0.000 39 1.0000 0.0013 0.066 0.000 50 -0.257 40 10.0000 0.0132 0.658 2.000 50 1.666 41 100.0000 0.1241 6.203 5.000 50 -0.516 42 43 Chi^2 = 3.11 d.f. = 3 P-value = 0.3755 44 45 46 Benchmark Dose Computation 47 48 Specified effect 0.01 49 50 Risk Type Extra risk 51 52 Confidence level 0.95 53 54 BMD 7.58722 55 56 BMDL 4.31737 57 58 BMDU 17.638 59 60 T a k e n t o g e t h e r , (4.31737, 1 7 . 6 3 8 ) is a 90 two-sided confidence 61 i n t e r v a l for t he B M D 62 63 M u l t i s t a g e C a n c e r Slope Factor 0.00231623 64 65 This document is a draftfor review purposes only and does not constitute Agency policy. F-92 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.3.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:48 04/01 2010 3 4 K ociba et al., 1978: A denom a o f adrenal cortex This document is a draftfor review purposes only and does not constitute Agency policy. F-93 DRAFT--DO NOT CITE OR QUOTE 1 F.2.4. Kociba et al., 1978: Hepatocellular adenoma(s) or carcinoma(s) 2 F.2.4.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 2 0.034 M ultistage Cancer, 2-D egree 2 0.034 M ultistage Cancer, 3-Degree 2 0.034 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 146.199 1.769E+00 1.225E+00 146.199 1.768E+00 1.225E+00 final =0 146.199 1.768E+00 1.225E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.4.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 K ociba et al., 1978: H epatocellular adenom a(s) or carcinom a(s) 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 4 m s c l I P e r c l i v e r ad carc.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 4 m s c l l P e r c l i v e r a d c a r c . p l t 14 T h u A p r _ 01_ 1 2 : 4 9 : 2 5 2010 15 16 17 S o u r c e - T a b l e 1 in G o o d m a n an d S a u e r 1992 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -beta1*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.0591902 46 Beta(1) = 0.00458516 47 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-94 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 Background Beta(1) 4 5 Background 1 -0.47 6 7 Beta(1) -0.47 1 8 9 10 11 P a r a m e t e r E s t i m a t e s 12 13 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 14 V a r i a b l e Estimate Std. Err. Lower Conf. Limit Upper Conf 15 Background 0.0328755 * * * 16 Beta(1) 0.00568299 * * * 17 18 - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 19 20 21 22 Analysis of Deviance Table 23 24 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 25 Full model -68.2561 4 26 Fitted model -71.0993 2 5.68634 2 0.05824 27 Reduced model -89.1983 1 41.8843 3 <.0001 28 29 AIC: 146.199 30 31 32 Goodness of Fit 33 Scaled 34 Dose Est. Prob. Expected Observed Size Residual 35 36 0.0000 0.0329 2.827 2.000 86 - 0 .500 37 1.0000 0.0384 1.918 1.000 50 -0.676 38 10.0000 0.0863 4.315 9.000 50 2.359 39 100.0000 0.4521 20.346 18.000 45 - 0 .703 40 41 Chi^2 = 6.77 d.f. = 2 P-value = 0.0339 42 43 44 Benchmark Dose Computation 45 46 Specified effect 0.01 47 48 Risk Type Extra risk 49 50 Confidence level 0.95 51 52 BMD 1.7685 53 54 BMDL 1.22517 55 56 BMDU 2.77641 57 58 T a k e n t o g e t h e r , (1.22517, 2.77641) is a 90 two-sided confidence 59 interval for the BMD 60 61 M u l t i s t a g e C a n c e r S l o p e F a c t o r 0.00816214 62 63 This document is a draftfor review purposes only and does not constitute Agency policy. F-95 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.4.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:49 04/01 2010 3 4 K ociba et al., 1978: H epatocellular adenom a(s) or carcinom a(s) This document is a draftfor review purposes only and does not constitute Agency policy. F-96 DRAFT--DO NOT CITE OR QUOTE 1 F.2.5. K ociba et al., 1978: Stratified squam ous cell carcinom a o f hard palate or nasal 2 turbinates 3 F .2 .5 .1 . Summary Table o f BMDS Modeling Results Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 3 0.928 M ultistage Cancer, 2-D egree 3 0.998 M ultistage Cancer, 3-Degree 3 1.000 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 30.745 1.344E+01 6.515E+00 29.961 3.490E+01 7.216E+00 29.885 4.941E+01 7.297E+00 Notes a Best-fitting m odel, BM D S output presented in this appendix 4 5 6 F .2.5.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 7 K ociba et al., 1978: Stratified squam ous cell carcinom a o f hard palate or nasal turbinates 8 9 10 11 12 M u l t i s t a g e C a n c e r M o d e l . (Version: 1.7; Date: 0 5 / 1 6 / 2 0 0 8 ) 13 I n p u t D a t a F i l e : C : \ C a n c \ 5 m s c l I P e r c n a s a l . ( d ) 14 G n u p l o t P l o t t i n g File: C : \ C a n c \ 5 m s c l l P e r c n a s a l . p l t 15 T h u A p r 01 1 2 : 4 9 : 5 9 2 0 1 0 16 17 18 S o u r c e - T a b l e 5 19 20 21 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 22 23 P[response] = b a c k g r o u n d + (1-background)*[1-EXP( 24 -betai*dose^i)] 25 26 The parameter betas are restricted to be positive 27 28 29 Dependent variable = Mean 30 Independent variable = Dose 31 32 Total number of observations = 4 33 Total n u m b e r of records w i t h m i s s i n g va l u e s = 0 34 Total number of parameters in model = 2 35 Total number of specified p arameters = 0 36 Degree of polynomial = 1 37 38 39 Maximum number of iterations = 250 40 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 41 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1e-008 42 43 44 45 Default Initial Parameter Values 46 Background = 0.00343283 47 Beta(1) = 0.000825276 This document is a draftfor review purposes only and does not constitute Agency policy. F-97 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Asymptotic Correlation Matrix of Parameter Estimates 4 5 ( *** The model parameter(s) -Background 6 have been estimated at a boundary point, or have been specified by the user, 7 and do not appear in the correlation matrix ) 8 9 Beta(1) 10 11 B e t a ( 1 ) 1 12 13 14 15 P a r a m e t e r E s t i m a t e s 16 17 95.0% W a l d C o n f i d e n c e I n t e r v a l 18 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 19 B a c k g r o u n d 0 20 Beta(1) 0.000953868 21 22 * - Indicates that this value is not calculated. 23 24 25 26 Analysis of Deviance Table 27 28 Model Log(likelihood) # Param' s Deviance Test d.f. P-value 29 Full model -18.7562 4 30 Fitted model -19.0532 1 0.594034 3 0.8978 31 R e d u c e d m o d e l -24.1972 1 10.882 3 0.01238 32 33 AIC: 40.1064 34 35 36 Goodness of Fit 37 38 Dose Est. Prob. Expected Observed Size Residu 39 40 0.0000 0.0000 0.000 0.000 86 0.000 41 1.0000 0.0010 0.048 0.000 50 -0.218 42 10.0000 0.0095 0.475 1.000 50 0.766 43 100.0000 0.0910 4.458 4.000 49 -0.227 44 45 Chi^2 = 0 .69 d.f. = 3 P- v a l u e = 0.8764 46 47 48 Benchmark Dose Computation 49 5 0 S p e c i f i e d <e f f e c t = 0.01 51 52 Risk Type = Extra risk 53 54 Confidence level = 0.95 55 56 BMD = 10.5364 57 58 BMDL = 5.46907 59 60 BMDU = 25.864 61 62 T a k e n t ogether, (5.46907 , 25.864 ) is a 90 % two-- s i d e d c o n f i d e n c e 63 inter v a l for the BMD 64 65 Multi s t a g e Cancer Slope Factor = 0 .00182846 66 67 This document is a draftfor review purposes only and does not constitute Agency policy. F-98 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.5.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:49 04/01 2010 3 4 K ociba et al., 1978: Stratified squam ous cell carcinom a o f hard palate or nasal turbinates This document is a draftfor review purposes only and does not constitute Agency policy. F-99 DRAFT--DO NOT CITE OR QUOTE 1 F.2.6. Kociba et al., 1978: Keratinizing squamous cell carcinoma of lung 2 F.2.6.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 3 0.837 M ultistage Cancer, 2-D egree 3 0.994 M ultistage Cancer, 3-Degree 3 1.000 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 43.792 7.311E+00 4.159E+00 42.346 2.568E+01 4.917E+00 42.207 4.026E+01 5.022E+00 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.6.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 K ociba et al., 1978: K eratinizing squam ous cell carcinom a o f lung 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 6 m s c l I P e r c k e r a carc.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 6 m s c l l P e r c k e r a c a r c . p l t 14 T h u A p r 01 1 2 : 5 0 : 3 4 2010 15 16 17 S o u r c e - T a b l e 5 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^i)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0 46 Beta(1) = 0.00158635 47 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-100 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 ( *** The model parameter(s) -Background 4 have been estimated at a boundary point, or have been specified by the user, 5 and do not appear in the correlation matrix ) 6 7 Beta(1) 8 9 Beta(1) 1 10 11 12 13 P a r a m e t e r E s t i m a t e s 14 15 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 16 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 17 B a c k g r o u n d 0 18 Beta(1) 0.0013747 19 20 * - I n d i c a t e s that this v a l u e is not c a l culated. 21 22 23 24 Analysis of Deviance Table 25 26 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 27 Full model -20.0957 4 28 Fitted model -20.8959 1 1.60041 3 0.6593 29 Reduced model -31.4904 1 22.7894 3 <.0001 30 31 AIC: 43.7918 32 33 34 Goodness of Fit 35 Scaled 36 Dose Est. Prob. Expected Observed Size Residu. 37 38 0.0000 0.0000 0.000 0.000 86 0.000 39 1.0000 0.0014 0.069 0.000 50 -0.262 40 10.0000 0.0137 0.683 0.000 50 -0.832 41 100.0000 0.1284 6.294 7.000 49 0.302 42 43 Chi^2 = 0.85 d.f. = 3 P-value = 0.8370 44 45 46 Benchmark Dose Computation 47 48 Specified effect 0.01 49 50 Risk Type Extra risk 51 52 Confidence level 0.95 53 54 BMD 7.31091 55 56 BMDL 4.15929 57 58 BMDU 14.6306 59 60 T a k e n t o g e t h e r , (4.15929, 14.6306) is a 90 two-sided confidence 61 i n t e r v a l for t he B M D 62 63 M u l t i s t a g e C a n c e r Slope Factor 0.00240426 64 65 This document is a draftfor review purposes only and does not constitute Agency policy. F-101 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.6.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:50 04/01 2010 3 4 K ociba et al., 1978: K eratinizing squam ous cell carcinom a o f lung This document is a draftfor review purposes only and does not constitute Agency policy. F-102 DRAFT--DO NOT CITE OR QUOTE 1 F.2.7. National Toxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma 2 F.2.7.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 2 0.146 M ultistage Cancer, 2-D egree 2 0.146 M ultistage Cancer, 3-Degree 2 0.146 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 76.377 9.761E+00 3.964E+00 76.377 9.761E+00 3.964E+00 final =0 76.377 9.761E+00 3.964E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.7.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 7 m s c l I P e r c su b f ibro.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 7 m s c l l P e r c s u b f i b r o . p l t 14 T h u A p r 01 1 2 : 5 1 : 0 7 2010 15 16 17 S o u r c e - T a b l e 10 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.030595 46 Beta(1) = 0.000799545 47 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-103 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 Background Beta(1) 4 5 Background 1 -0.54 6 7 Beta(1) -0.54 1 8 9 10 11 P a r a m e t e r E s t i m a t e s 12 13 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 14 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 15 Background 0.0231556 16 Beta(1) 0.00102962 17 18 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 19 20 21 22 Analysis of Deviance Table 23 24 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 25 Full model -33.5998 4 26 Fitted model -36.1883 2 5.17698 2 0.07513 27 Reduced model -37.7465 1 8.29346 3 0.04032 28 29 AIC: 76.3766 30 31 32 Goodness of Fit 33 Scaled 34 Dose Est. Prob. Expected Observed Size Residual 35 36 0.0000 0.0232 1.737 0.000 75 -1.333 37 1.4000 0.0246 1.228 2.000 50 0.705 38 7.1000 0.0303 1.514 3.000 50 1.227 39 71.0000 0.0920 4.509 4.000 49 -0.252 40 41 Chi^2 = 3.84 d.f. = 2 P-value = 0.1463 42 43 44 Benchmark Dose Computation 45 46 Specified effect = 0.01 47 48 Risk Type = Extra risk 49 50 Confidence level = 0.95 51 52 BMD = 9.76124 53 54 BMDL = 3.96354 55 56 BMDU = 1.03301e+006 57 58 T a k e n t ogether, (3.96354, 1 . 0 3 3 0 1 e + 0 0 6 ) is a 90 % two-sided confidence 59 interval for the BMD 60 61 M u l t i s t a g e C a n c e r S l o p e F a c t o r = 0.002523 62 63 This document is a draftfor review purposes only and does not constitute Agency policy. F-104 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.7.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:51 04/01 2010 3 4 National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma This document is a draftfor review purposes only and does not constitute Agency policy. F-105 DRAFT--DO NOT CITE OR QUOTE 1 F.2.8. National Toxicology Program, 1982: Liver: Neoplastic Nodule or Hepatocellular 2 Carcinoma 3 F .2 .8 .1 . Summary Table o f BMDS Modeling Results Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 2 0.398 M ultistage Cancer, 2-D egree 2 0.503 M ultistage Cancer, 3-Degree 2 0.503 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 133.832 2.554E+00 1.600E+00 133.436 1.334E+01 1.652E+00 133.436 1.334E+01 1.652E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 4 5 6 F .2.8.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 7 National T oxicology Program, 1982: Liver: N eoplastic N odule or H epatocellular Carcinoma 8 9 10 11 12 M u l t i s t a g e C a n c e r M o d e l . (Version: 1.7; Date: 0 5 / 1 6 / 2 0 0 8 ) 13 I n p u t D a t a F i l e : C : \ C a n c \ 8 m s c l I P e r c l i v e r n o d . ( d ) 14 G n u p l o t P l o t t i n g File: C : \ C a n c \ 8 m s c l l P e r c l i v e r n o d . p l t 15 T h u A p r _ 01 1 2 : 5 1 : 4 1 2 0 1 0 16 17 18 S o u r c e - T a b l e 10 19 20 21 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 22 23 P[response] = b a c k g r o u n d + (1-background)*[1-EXP( 24 -betai*dose^i)] 25 26 The parameter betas are restricted to be positive 27 28 29 Dependent variable = Mean 30 Independent variable = Dose 31 32 Total number of observations = 4 33 Total n u m b e r of records w i t h m i s s i n g va l u e s = 0 34 Total number of parameters in model = 2 35 Total number of specified p arameters = 0 36 Degree of polynomial = 1 37 38 39 Maximum number of iterations = 250 40 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 41 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1e-008 42 43 44 45 Default Initial Parameter Values 46 Background = 0.0383072 47 Beta(1) = 0.00417257 This document is a draftfor review purposes only and does not constitute Agency policy. F-106 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Asymptotic Correlation Matrix of Parameter Estimates 4 5 Background Beta(1) 6 7 Background 1 -0.47 8 9 Beta(1) -0.47 1 10 11 12 13 P a r a m e t e r E s t i m a t e s 14 15 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 16 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 17 Background 0.0451327 18 Beta(1) 0.00393556 19 20 * - Indicates that this value is not calculated. 21 22 23 24 Analysis of Deviance Table 25 26 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 27 Full model -63.9149 4 28 Fitted model -64.916 2 2.00214 2 0.3675 29 Reduced model -74.0195 1 20.2092 3 0.0001536 30 31 AIC: 133.832 32 33 34 Goodness of Fit 35 Scaled 36 Dose Est. Prob. Expected Observed Size Residu. 37 38 0.0000 0.0451 3.385 5.000 75 0.898 39 1.4000 0.0504 2.469 1.000 49 - 0 .959 40 7.1000 0.0714 3.572 3.000 50 -0.314 41 71.0000 0.2779 13.618 14.000 49 0.122 42 43 Chi^2 = 1.84 d.f. = 2 P-value = 0.3984 44 45 46 Benchmark Dose Computation 47 48 Specified effect 0.01 49 50 Risk Type Extra risk 51 52 Confidence level 0.95 53 54 BMD 2.55373 55 56 BMDL 1.59983 57 58 BMDU 4.74206 59 60 T a k e n t o g e t h e r , (1.59983, 4.74206) is a 90 two-sided confidence 61 i n t e r v a l for t he B M D 62 63 M u l t i s t a g e C a n c e r Slope Factor 0.00625067 64 65 This document is a draftfor review purposes only and does not constitute Agency policy. F-107 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.8.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:51 04/01 2010 3 4 National T oxicology Program, 1982: Liver: N eoplastic N odule or H epatocellular Carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-108 DRAFT--DO NOT CITE OR QUOTE 1 F.2.9. National Toxicology Program, 1982: Adrenal: Cortical Adenoma, or Carcinoma or 2 Adenoma, NOS 3 F .2 .9 .1 . Summary Table o f BMDS Modeling Results Model Degrees of x2pFreedom Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Multistage Cancer, 1-Degree a 2 0.405 203.380 3.672E+00 1.871E+00 Multistage Cancer, 2-Degree 2 0.501 202.885 1.577E+01 1.974E+00 Multistage Cancer, 3-Degree 2 0.513 202.832 2.600E+01 1.986E+00 Notes aBest-fitting model, BMDS output presented in this appendix 4 5 6 F .2.9.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 7 National T oxicology Program, 1982: Adrenal: Cortical A denom a, or Carcinoma or Adenom a, 8 NOS 9 10 11 12 13 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 14 I n p u t D a t a File: C : \ C a n c \ 9 m s c l I P e r c a d r e c o r t ad carc.(d) 15 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 9 m s c l l P e r c a d r e c o r t a d c a r c . p l t 16 T h u A p r 01 1 2 : 5 3 : 5 7 2010 17 18 19 S o u r c e - T a b l e 10 20 21 22 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 23 24 P[response] = background + (1-background)*[1-EXP( 25 -betai*dose^1)] 26 27 The parameter betas are restricted to be positive 28 29 30 Dependent variable = Mean 31 I n d e p e n d e n t v a r i a b l e = Dose 32 33 Total n u m b e r of o b s e r v a t i o n s = 4 34 Total number of records with missing values = 0 35 Total number of p arameters in model = 2 36 Total number of specified parameters = 0 37 Degree of polynomial = 1 38 39 40 Maximum number of iterations = 250 41 R e l a t i v e F u n c t i o n C o n v e r g e n c e has b e e n set to: 1e-008 42 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 43 44 45 46 Default Initial Parameter Values 47 Background = 0.140663 This document is a draftfor review purposes only and does not constitute Agency policy. F-109 DRAFT--DO NOT CITE OR QUOTE 1 Beta(1) 0.00289845 2 3 4 Asymptotic Correlation Matrix of Parameter Estimates 5 6 Background Beta(1) 7 8 Background 1 -0.48 9 10 Beta(1) -0.48 1 11 12 13 14 P a r a m e t e r E s t i m a t e s 15 16 95.0% W a l d C o n f i d e n c e I n t e r v a l 17 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 18 B a c k g r o u n d 0.143284 * 19 Beta(1) 0.00273674 * 20 21 * - In d i c a t e s that this v a l u e is not calculated. 22 23 24 25 Analysis of Deviance Table 26 27 Model Log(likelihood) # Param's Deviance Test d.f. P-value 28 Full model -98.7282 4 29 Fitted model -99.6898 2 1.92318 2 0.3823 30 Reduced model -102.201 1 6.94636 3 0.07363 31 32 AIC: 203.38 33 34 35 Goodness of Fit 36 Scaled 37 Dose Est. Prob. Expected Observed Size Residual 38 39 0.0000 0.1433 10.460 11.000 73 0.180 40 1.4000 0.1466 7.181 9.000 49 0.735 41 7.1000 0.1598 7.829 5.000 49 - 1 .103 42 71.0000 0.2946 13.551 14.000 46 0.145 43 44 Chi^2 = 1.81 d.f. = 2 P-value = 0.4046 45 46 47 Benchmark Dose Computation 48 49 Specified effect 0.01 50 51 R i s k T y p e Extra risk 52 53 C o n f i d e n c e level 0.95 54 55 BMD 3.67237 56 57 BMDL 1.87133 58 59 BMDU 15.4002 60 61 T a k e n t o g e t h e r , (1.87133, 15.40 0 2 ) is a 90 two-sided confidence 62 interval for the BMD 63 64 Multistage Cancer Slope Factor 0.00534381 65 66 This document is a draftfor review purposes only and does not constitute Agency policy. F-110 DRAFT--DO NOT CITE OR QUOTE 1 F.2.9.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:53 04/01 2010 3 4 National T oxicology Program, 1982: Adrenal: Cortical A denom a, or Carcinoma or Adenom a, 5 NOS This document is a draftfor review purposes only and does not constitute Agency policy. F-111 DRAFT--DO NOT CITE OR QUOTE 1 F.2.10. N ational T oxicology Program , 1982: Thyroid: Follicular-C ell A denom a 2 F .2 .1 0 .1 . Summary Table o f BMDS Modeling Results Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 2 0.661 M ultistage Cancer, 2-D egree 2 0.769 M ultistage Cancer, 3-Degree 2 0.781 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 92.020 7.571E+00 3.488E+00 91.639 2.257E+01 3.656E+00 91.601 3.302E+01 3.675E+00 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.10.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 1982: Thyroid: Follicular-C ell A denom a 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 1 0 _ m s c 1 _ 1 P e r c _ t h y _ a d . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 1 0 _ m s c 1 _ 1 P e r c _ t h y _ a d . p l t 14 T h u A p r 01 1 2 : 5 4 : 3 1 2010 15 16 17 S o u r c e - T a b l e 10 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^i)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.032089 46 Beta(1) = 0.00143599 47 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-112 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 Background Beta(1) 4 5 Background 1 -0.5 6 7 Beta(1) -0.5 1 8 9 10 11 P a r a m e t e r E s t i m a t e s 12 13 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 14 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 15 Background 0.0345958 16 Beta(1) 0.00132742 17 18 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 19 20 21 22 Analysis of Deviance Table 23 24 Model Log(likelihood) # Param' s Deviance Test d.f. P-value 25 Full model -43.5264 4 26 Fitted model -44.0098 2 0.966786 2 0.6167 27 Reduced model -46.2299 1 5.40699 3 0.1443 28 29 AIC: 92.0196 30 31 32 Goodness of Fit 33 34 Dose Est. Prob. Expected Observed Size Residu; 35 36 0.0000 0.0346 2.525 3.000 73 0.304 37 1.4000 0.0364 1.637 2.000 45 0.289 38 7.1000 0.0437 2.139 1.000 49 - 0 .796 39 71.0000 0.1214 5.707 6.000 47 0 . 1 3 1 40 41 Chi^2 = 0 .83 d.f. = 2 P- v a l u e = 0.6614 42 43 44 Benchmark Dose Computation 45 4 6 S p e c i f i e d <e f f e c t = 0.01 47 48 Risk Type = Extra risk 49 50 Confidence level = 0.95 51 52 BMD = 7.57131 53 54 BMDL = 3.48815 55 56 BMDU = 964541 57 58 T a k e n t ogether, (3.48815 , 964541 ) is a 90 % two -sided confidence 59 interval for the BMD 60 61 M u l t i s t a g e C a n c e r S l o p e F a c t o r = 0 .00286685 62 63 This document is a draftfor review purposes only and does not constitute Agency policy. F-113 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.10.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:54 04/01 2010 3 4 National T oxicology Program, 1982: Thyroid: Follicular-C ell A denom a This document is a draftfor review purposes only and does not constitute Agency policy. F-114 DRAFT--DO NOT CITE OR QUOTE 1 F.2.11. National Toxicology Program, 1982: Liver: Neoplastic Nodule or Hepatocellular 2 Carcinoma 3 F .2 .1 1 .1 . Summary Table o f BMDS Modeling Results Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 2 0.398 M ultistage Cancer, 2-D egree 2 0.503 M ultistage Cancer, 3-Degree 2 0.503 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 133.832 2.554E+00 1.600E+00 133.436 1.334E+01 1.652E+00 133.436 1.334E+01 1.652E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 4 5 6 F .2.11.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 7 National T oxicology Program, 1982: Liver: N eoplastic N odule or H epatocellular Carcinoma 8 9 10 11 12 M u l t i s t a g e C a n c e r M o d e l . (Version: 1.7; Date: 0 5 / 1 6 / 2 0 0 8 ) 13 I n p u t D a t a F i l e : C : \ C a n c \ 1 1 m s c l I P e r c l i v e r n o d . ( d ) 14 G n u p l o t P l o t t i n g File: C : \ C a n c \ 1 1 m s c 1 1 P e r c l i v e r n o d . p l t 15 T h u A p r 01 1 2 : 5 5 : 0 5 2 0 1 0 16 17 18 S o u r c e - T a b l e 9 19 20 21 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 22 23 P[response] = b a c k g r o u n d + (1-background)*[1-EXP( 24 -betai*dose^1)] 25 26 The parameter betas are restricted to be positive 27 28 29 Dependent variable = Mean 30 Independent variable = Dose 31 32 Total number of observations = 4 33 Total n u m b e r of records w i t h m i s s i n g va l u e s = 0 34 Total number of parameters in model = 2 35 Total number of specified p arameters = 0 36 Degree of polynomial = 1 37 38 39 Maximum number of iterations = 250 40 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 41 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1e-008 42 43 44 45 Default Initial Parameter Values 46 Background = 0 47 Beta(1) = 0.000900399 This document is a draftfor review purposes only and does not constitute Agency policy. F-115 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Asymptotic Correlation Matrix of Parameter Estimates 4 5 ( *** The model parameter(s) -Background 6 have been estimated at a boundary point, or have been specified by the user, 7 and do not appear in the correlation matrix ) 8 9 Beta(1) 10 11 B e t a ( 1 ) 1 12 13 14 15 P a r a m e t e r E s t i m a t e s 16 17 95.0% W a l d C o n f i d e n c e I n t e r v a l 18 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 19 B a c k g r o u n d 0 20 Beta(1) 0.000775683 21 22 * - Indicates that this value is not calculated. 23 24 25 26 Analysis of Deviance Table 27 28 Model Log(likelihood) # Param's Deviance Test d.f. P-value 29 Full model -11.3484 4 30 Fitted model -11.6976 1 0.698469 3 0.8736 31 R e d u c e d m o d e l -15.9189 1 9.14109 3 0.02747 32 33 AIC: 25.3952 34 35 36 Goodness of Fit 37 Scaled 38 Dose Est. Prob. Expected Observed Size Residu 39 40 0.0000 0.0000 0.000 0.000 74 0 . 000 41 1.4000 0.0011 0.054 0.000 50 -0.233 42 7.1000 0.0055 0.275 0.000 50 -0.525 43 71.0000 0.0536 2.679 3.000 50 0.201 44 45 Chi^2 = 0.37 d.f. = 3 P-value = 0.9462 46 47 48 Benchmark Dose Computation 49 50 Specified effect 0.01 51 52 Risk Type Extra risk 53 54 Confidence level 0.95 55 56 BMD 12.9568 57 58 BMDL 5.70369 59 60 BMDU 39.9878 61 62 T a k e n t o g e t h e r , (5.70369, 39.9878) is a 90 two-sided confidence 63 inter v a l for the BMD 64 65 M u l t i s t a g e Cancer Slope Factor 0.00175325 66 67 This document is a draftfor review purposes only and does not constitute Agency policy. F-116 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.11.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:55 04/01 2010 3 4 National T oxicology Program, 1982: Liver: N eoplastic N odule or H epatocellular Carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-117 DRAFT--DO NOT CITE OR QUOTE 1 F.2.12. National Toxicology Program, 1982: Thyroid: Follicular-Cell Adenoma or 2 Carcinoma 3 F .2 .1 2 .1 . Summary Table o f BMDS Modeling Results Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 2 0.028 M ultistage Cancer, 2-D egree 2 0.028 M ultistage Cancer, 3-Degree 2 0.028 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 151.224 3.521E+00 1.916E+00 151.224 3.521E+00 1.916E+00 final =0 151.224 3.521E+00 1.916E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 4 5 6 F .2.12.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 7 National T oxicology Program, 1982: Thyroid: Follicular-C ell A denom a or Carcinoma 8 9 10 11 12 M u l t i s t a g e C a n c e r M o d e l . (Version: 1.7; Date: 0 5 / 1 6 / 2 0 0 8 ) 13 I n p u t D a t a F i l e : C : \ C a n c \ 1 2 _ m s c 1 _ 1 P e r c _ t h y r o i d . ( d ) 14 G n u p l o t P l o t t i n g File: C : \ C a n c \ 1 2 m s c 1 1 P e r c t h y r o i d . p l t 15 T h u A p r 01 1 2 : 5 5 : 3 8 2 0 1 0 16 17 18 S o u r c e - T a b l e 9 19 20 21 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 22 23 P[response] = b a c k g r o u n d + (1-background)*[1-EXP( 24 -betai*dose^1)] 25 26 The parameter betas are restricted to be positive 27 28 29 Dependent variable = Mean 30 Independent variable = Dose 31 32 Total number of observations = 4 33 Total n u m b e r of records w i t h m i s s i n g va l u e s = 0 34 Total number of parameters in model = 2 35 Total number of specified p arameters = 0 36 Degree of polynomial = 1 37 38 39 Maximum number of iterations = 250 40 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 41 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1e-008 42 43 44 45 Default Initial Parameter Values 46 Background = 0.0867382 47 Beta(1) = 0.00232055 This document is a draftfor review purposes only and does not constitute Agency policy. F-118 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Asymptotic Correlation Matrix of Parameter Estimates 4 5 Background Beta(1) 6 7 Background 1 -0.53 8 9 Beta(1) -0.53 1 10 11 12 13 P a r a m e t e r E s t i m a t e s 14 15 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 16 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 17 Background 0.0704713 18 Beta(1) 0.00285481 19 20 * - Indicates that this value is not calculated. 21 22 23 24 Analysis of Deviance Table 25 26 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 27 Full model -69.5946 4 28 Fitted model -73.6119 2 8.03468 2 0.018 29 Reduced model -77.5267 1 15.8643 3 0.001209 30 31 AIC: 151.224 32 33 34 Goodness of Fit 35 Scaled 36 Dose Est. Prob. Expected Observed Size Residual 37 38 0.0000 0.0705 4.863 1.000 69 -1.817 39 1.4000 0.0742 3.561 5.000 48 0 . 793 40 7.1000 0.0891 4.456 8.000 50 1.759 41 71.0000 0.2410 12.051 11.000 50 -0.347 42 43 Chi^2 = 7.14 d.f. = 2 P-value = 0.0281 44 45 46 Benchmark Dose Computation 47 48 Specified effect 0.01 49 50 Risk Type Extra risk 51 52 Confidence level 0.95 53 54 BMD 3.5205 55 56 BMDL 1.91558 57 58 BMDU 9.76663 59 60 T a k e n t o g e t h e r , (1.91558, 9.76663) is a 90 two-sided confidence 61 i n t e r v a l for t he B M D 62 63 M u l t i s t a g e C a n c e r Slope Factor 0.00522034 64 65 This document is a draftfor review purposes only and does not constitute Agency policy. F-119 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.12.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:55 04/01 2010 3 4 National T oxicology Program, 1982: Thyroid: Follicular-C ell A denom a or Carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-120 DRAFT--DO NOT CITE OR QUOTE 1 F.2.13. National Toxicology Program, 1982: Adrenal cortex: Adenoma 2 F.2.13.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 2 0.054 M ultistage Cancer, 2-D egree 2 0.054 M ultistage Cancer, 3-Degree 2 0.054 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 199.672 1.400E+01 3.444E+00 199.672 1.400E+01 3.444E+00 final =0 199.672 1.400E+01 3.444E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.13.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 1982: Adrenal cortex: Adenom a 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 1 3 m s c l I P e r c a d r e cort.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 1 3 m s c l l P e r c a d r e c o r t . p l t 14 T h u A p r _ 01 1 2 : 5 6 : 1 0 2010 15 16 17 S o u r c e - T a b l e 9 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.168444 46 Beta(1) = 0.000395949 47 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-121 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 Background Beta(1) 4 5 Background 1 -0.53 6 7 Beta(1) -0.53 1 8 9 10 11 P a r a m e t e r E s t i m a t e s 12 13 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 14 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 15 B a c k g r o u n d 0.153096 16 Beta(1) 0.000718012 17 18 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 19 20 21 22 Analysis of Deviance Table 23 24 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 25 Full model -94.8672 4 26 Fitted model -97.8359 2 5.93732 2 0.05137 27 Reduced model -98.0432 1 6.35197 3 0.09569 28 29 AIC: 199.672 30 31 32 Goodness of Fit 33 Scaled 34 Dose Est. Prob. Expected Observed Size Residual 35 36 0.0000 0.1531 11.023 6.000 72 -1.644 37 1.4000 0.1539 7.697 9.000 50 0.510 38 7.1000 0.1574 7.713 12.000 49 1.682 39 71.0000 0.1952 9.564 9.000 49 - 0 .203 40 41 Chi^2 = 5.83 d.f. = 2 P-value = 0.0541 42 43 44 Benchmark Dose Computation 45 46 Specified effect 0.01 47 48 Risk Type : Extra risk 49 50 Confidence level 0.95 51 52 BMD 13.9974 53 54 BMDL 3.4443 55 56 57 BMDU did not converge for BMR = 0.010000 58 BMDU calculation failed 59 BMDU = Inf 60 61 This document is a draftfor review purposes only and does not constitute Agency policy. F-122 DRAFT--DO NOT CITE OR QUOTE 1 F.2.13.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level Fraction Affected 2 11:56 04/01 2010 3 4 National T oxicology Program, 1982: Adrenal cortex: Adenom a This document is a draftfor review purposes only and does not constitute Agency policy. F-123 DRAFT--DO NOT CITE OR QUOTE 1 F.2.14. National Toxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma 2 F.2.14.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 2 0.146 M ultistage Cancer, 2-D egree 2 0.146 M ultistage Cancer, 3-Degree 2 0.146 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 76.377 9.761E+00 3.964E+00 76.377 9.761E+00 3.964E+00 final =0 76.377 9.761E+00 3.964E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.14.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 1 4 m s c l I P e r c s u b c u f ibro.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 1 4 m s c l l P e r c s u b c u f i b r o . p l t 14 T h u A p r 01 1 2 : 5 6 : 4 1 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.0143554 46 Beta(1) = 0.000341874 47 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-124 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 Background Beta(1) 4 5 Background 1 -0.5 6 7 Beta(1) -0.5 1 8 9 10 11 P a r a m e t e r E s t i m a t e s 12 13 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 14 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 15 Background 0.0145028 16 Beta(1) 0.000338561 17 18 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 19 20 21 22 Analysis of Deviance Table 23 24 Model Log(likelihood) # Param's Deviance Test d.f. P-value 25 Full model -30.9876 4 26 Fitted model -31.0199 2 0.0645971 2 0.9682 27 Reduced model -34.3291 1 6.68308 3 0.08272 28 29 AIC: 66.0397 30 31 32 Goodness of Fit 33 Scaled 34 Dose Est. Prob. Expected Observed Size Residual 35 36 0.0000 0.0145 1.073 1.000 74 - 0 . 0 7 1 37 5.7000 0.0164 0.820 1.000 50 0.200 38 28.6000 0.0240 1.152 1.000 48 - 0 . 1 4 3 39 286.0000 0.1055 4.956 5.000 47 0 . 0 2 1 40 41 Chi^2 = 0.07 d.f. = 2 P-value = 0.9675 42 43 44 Benchmark Dose Computation 45 46 Specified effect 0.01 47 48 Risk Type Extra risk 49 50 Confidence level 0.95 51 52 BMD 29.6855 53 54 BMDL 14.3524 55 56 BMDU 100.382 57 58 T a k e n t o g e t h e r , (14.3524, 100.382) is a 90 two-sided confidence 59 interval for the BMD 60 61 M u l t i s t a g e C a n c e r S l o p e F a c t o r 0.000696747 62 63 This document is a draftfor review purposes only and does not constitute Agency policy. F-125 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.14.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:56 04/01 2010 3 4 National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma This document is a draftfor review purposes only and does not constitute Agency policy. F-126 DRAFT--DO NOT CITE OR QUOTE 1 F.2.15. N ational Toxicology Program , 1982: H em atopoietio System : L ym phom a or 2 L eukem ia 3 F .2 .1 5 .1 . Summary Table o f BMDS Modeling Results Model Degrees of x2pFreedom Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Multistage Cancer, 1-Degree a 2 0.987 261.425 1.034E+01 5.456E+00 M ultistage Cancer, 2-D egree 2 0.987 261.425 1.034E+01 5.456E+00 final =0 M ultistage Cancer, 3-Degree 2 0.987 261.425 1.034E+01 5.456E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 4 5 6 F .2.15.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 7 National T oxicology Program, 1982: H em atopoietio System: Lym phom a or Leukem ia 8 9 10 11 12 M u l t i s t a g e C a n c e r M o d e l . (Version: 1.7; Date: 0 5 / 1 6 / 2 0 0 8 ) 13 I n p u t D a t a F i l e : C : \ C a n c \ 1 5 m s c l I P e r c m i c e f l y m p h o m a . ( d ) 14 G n u p l o t P l o t t i n g File: C : \ C a n c \ 1 5 m s c l l P e r c m i c e f l y m p h o m a . p l t 15 T h u A p r _ 01 1 2 : 5 7 : 1 4 2 0 1 0 16 17 18 T a b l e 15 p a g e 64 H e m a t o p o i e t i c S y s t e m L y m p h o m a o r L e u k e m i a 19 20 21 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 22 23 P[response] = b a c k g r o u n d + (1-background)*[1-EXP( 24 -betai*dose^1)] 25 26 The parameter betas are restricted to be positive 27 28 29 Dependent variable = Mean 30 Independent variable = Dose 31 32 Total number of observations = 4 33 Total n u m b e r of records w i t h m i s s i n g va l u e s = 0 34 Total number of parameters in model = 2 35 Total number of specified p arameters = 0 36 Degree of polynomial = 1 37 38 39 Maximum number of iterations = 250 40 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 41 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1e-008 42 43 44 45 Default Initial Parameter Values 46 Background = 0.242959 47 Beta(1) = 0.000967723 This document is a draftfor review purposes only and does not constitute Agency policy. F-127 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Asymptotic Correlation Matrix of Parameter Estimates 4 5 Background Beta(1) 6 7 Background 1 -0.48 8 9 Beta(1) -0.48 1 10 11 12 13 P a r a m e t e r E s t i m a t e s 14 15 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 16 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 17 B a c k g r o u n d 0.242712 * 18 Beta(1) 0.000971954 * 19 20 * - Indicates that this value is not calculated. 21 22 23 24 Analysis of Deviance Table 25 26 Model Log(likelihood) # Param's Deviance Test d.f. P-value 27 Full model -128.699 4 28 Fitted model -128.712 2 0.0264819 2 0.9868 29 Reduced model -131.412 1 5.42487 3 0.1432 30 31 AIC: 261.425 32 33 34 Goodness of Fit 35 Scaled 36 Dose Est. Prob. Expected Observed Size Residual 37 38 0.0000 0.2427 17.961 18.000 74 0.011 39 5.7000 0.2469 12.345 12.000 50 -0.113 40 28.6000 0.2635 12.647 13.000 48 0 . 116 41 286.0000 0.4265 20.045 20.000 47 - 0 . 0 1 3 42 43 Chi^2 = 0.03 d.f. = 2 P-value = 0.9868 44 45 46 Benchmark Dose Computation 47 48 Specified effect 0.01 49 50 Risk Type Extra risk 51 52 Confidence level 0.95 53 54 BMD 10.3403 55 56 BMDL 5.45599 57 58 BMDU 38.9139 59 60 T a k e n t o g e t h e r , (5.45599, 38.91 3 9 ) is a 90 two-sided confidence 61 i n t e r v a l for t he B M D 62 63 M u l t i s t a g e C a n c e r Slope Factor 0.00183285 64 65 This document is a draftfor review purposes only and does not constitute Agency policy. F-128 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.15.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:57 04/01 2010 3 4 National T oxicology Program, 1982: H em atopoietio System: Lym phom a or Leukem ia This document is a draftfor review purposes only and does not constitute Agency policy. F-129 DRAFT--DO NOT CITE OR QUOTE 1 F.2.16. National Toxicology Program, 1982: Liver: Hepatooellular Adenoma or Carcinoma 2 F.2.16.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 2 0.244 M ultistage Cancer, 2-D egree 2 0.244 M ultistage Cancer, 3-Degree 2 0.244 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 156.001 1.458E+01 7.829E+00 156.001 1.458E+01 7.829E +00 final =0 156.001 1.458E+01 7.829E +00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.16.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 1982: Liver: H epatooellular A denom a or Carcinoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 1 6 m s c l I P e r c m i c e f l i v a d e n carc.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 1 6 m s c l l P e r c m i c e f l i v a d e n c a r c . p l t 14 T h u A p r _ 01 1 2 7 5 7 : 4 7 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.0888873 46 Beta(1) = 0.000616931 47 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-130 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 Background Beta(1) 4 5 Background 1 -0.5 6 7 Beta(1) -0.5 1 8 9 10 11 P a r a m e t e r E s t i m a t e s 12 13 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 14 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 15 Background 0.0788077 16 Beta(1) 0.000689385 17 18 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 19 20 21 22 Analysis of Deviance Table 23 24 Model Log(likelihood) # Param's Deviance Test d.f. P-value 25 Full model -74.5177 4 26 Fitted model -76.0006 2 2.96597 2 0.227 27 Reduced model -79.6703 1 10.3053 3 0.01614 28 29 AIC: 156.001 30 31 32 Goodness of Fit 33 Scaled 34 Dose Est. Prob. Expected Observed Size Residual 35 36 0.0000 0.0788 5.753 3.000 73 -1.196 37 5.7000 0.0824 4.121 6.000 50 0.966 38 28.6000 0.0968 4.646 6.000 48 0.661 39 286.0000 0.2436 11.452 11.000 47 - 0 . 1 5 3 40 41 Chi^2 = 2.82 d.f. = 2 P-value = 0.2436 42 43 44 Benchmark Dose Computation 45 46 Specified effect 0.01 47 48 Risk Type Extra risk 49 50 Confidence level 0.95 51 52 BMD 14.5787 53 54 BMDL 7.82902 55 56 BMDU 42.4536 57 58 T a k e n t o g e t h e r , (7.82902, 42.4536) is a 90 two-sided confidence 59 interval for the BMD 60 61 M u l t i s t a g e C a n c e r S l o p e F a c t o r 0.0012773 62 63 This document is a draftfor review purposes only and does not constitute Agency policy. F-131 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.16.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:57 04/01 2010 3 4 National T oxicology Program, 1982: Liver: H epatooellular A denom a or Carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-132 DRAFT--DO NOT CITE OR QUOTE 1 F.2.17. National Toxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma 2 F.2.17.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 2 0.146 M ultistage Cancer, 2-D egree 2 0.146 M ultistage Cancer, 3-Degree 2 0.146 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 76.377 9.761E+00 3.964E+00 76.377 9.761E+00 3.964E+00 final =0 76.377 9.761E+00 3.964E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.17.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 1 7 m s c l I P e r c m i c e f t h y r o i d aden.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 1 7 m s c l l P e r c m i c e f t h y r o i d a d e n . p l t 14 T h u A p r _ 01 1 2 : 5 8 : 2 0 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.02405 46 Beta(1) = 0.000315564 47 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-133 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 Background Beta(1) 4 5 Background 1 -0.51 6 7 Beta(1) -0.51 1 8 9 10 11 P a r a m e t e r E s t i m a t e s 12 13 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 14 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 15 Background 0.0207192 16 Beta(1) 0.000331835 17 18 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 19 20 21 22 Analysis of Deviance Table 23 24 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 25 Full model -32.0017 4 26 Fitted model -34.6122 2 5.22112 2 0.07349 27 Reduced model -37.2405 1 10.4776 3 0.01491 28 29 AIC: 73.2245 30 31 32 Goodness of Fit 33 Scaled 34 Dose Est. Prob. Expected Observed Size Residual 35 36 0.0000 0.0207 1.430 0.000 69 -1.208 37 5.7000 0.0226 1.128 3.000 50 1.782 38 28.6000 0.0300 1.409 1.000 47 - 0 . 3 5 0 39 286.0000 0.1094 5.032 5.000 46 - 0 .015 40 41 Chi^2 = 4.76 d.f. = 2 P-value = 0.0927 42 43 44 Benchmark Dose Computation 45 46 Specified effect 0.01 47 48 Risk Type Extra risk 49 50 Confidence level 0.95 51 52 BMD 30.2871 53 54 BMDL 13.993 55 56 BMDU 130.014 57 58 T a k e n t o g e t h e r , (13.993 , 130.014) is a 90 two-sided confidence 59 interval for the BMD 60 61 M u l t i s t a g e C a n c e r S l o p e F a c t o r 0.000714641 62 63 This document is a draftfor review purposes only and does not constitute Agency policy. F-134 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.17.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:58 04/01 2010 3 4 National T oxicology Program, 1982: Subcutaneous Tissue: Fibrosarcoma This document is a draftfor review purposes only and does not constitute Agency policy. F-135 DRAFT--DO NOT CITE OR QUOTE 1 F.2.18. N ational Toxicology Program , 1982: Lung: A lveolar/B ronchiolar A denom a or 2 C arcinom a 3 F .2 .1 8 .1 . Summary Table o f BMDS Modeling Results Model Degrees of x2pFreedom Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) M ultistage Cancer, 1-Degree Multistage Cancer, 2-Degree a 2 2 0.138 167.341 3.706E+00 2.026E+00 0.181 166.805 1.590E+01 2.139E+00 M ultistage Cancer, 3-Degree 2 0.185 166.777 2.618E+01 2.145E+00 Notes a Best-fitting m odel, BM D S output presented in this appendix 4 5 6 F .2.18.2. Outputfor Selected Model: Multistage Cancer, 2-Degree 7 National T oxicology Program, 1982: Lung: Alveolar/Bronchiolar A denom a or Carcinoma 8 9 10 11 12 M u l t i s t a g e C a n c e r M o d e l . (Version: 1.7; Date: 0 5 / 1 6 / 2 0 0 8 ) 13 I n p u t D a t a F i l e : C : \ C a n c \ 1 8 m s c 2 I P e r c l u n g a d e n c a r c . ( d ) 14 G n u p l o t P l o t t i n g File: C : \ C a n c \ 1 8 m s c 2 I P e r c l u n g a d e n c a r c . p l t 15 T h u A p r 01 1 2 : 5 8 : 5 5 2 0 1 0 16 17 18 0 19 20 21 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 22 23 P[response] = b a c k g r o u n d + (1-background)*[1-EXP( 24 -betai*dose^1-beta2*dose^2)] 25 26 The parameter betas are restricted to be positive 27 28 29 Dependent variable = Mean 30 Independent variable = Dose 31 32 Total number of observations = 4 33 Total n u m b e r of records w i t h m i s s i n g va l u e s = 0 34 Total number of parameters in model = 3 35 Total number of specified p arameters = 0 36 Degree of polynomial = 2 37 38 39 Maximum number of iterations = 250 40 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 41 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1e-008 42 43 44 45 Default Initial Parameter Values 46 Background = 0.0889033 47 Beta(1) = 0 This document is a draftfor review purposes only and does not constitute Agency policy. F-136 DRAFT--DO NOT CITE OR QUOTE 1 Beta(2) = 4.12413e-005 2 3 4 Asymptotic Correlation Matrix of Parameter Estimates 5 6 ( *** The model parameter(s) -Beta(1) 7 have been estimated at a boundary point, or have been specified by the user, 8 and do not appear in the correlation matrix ) 9 10 Background Beta(2) 11 12 B a c k g r o u n d 1 -0.45 13 14 Beta(2) -0.45 1 15 16 17 18 P a r a m e t e r E s t i m a t e s 19 20 95.0% Wald Confidence Interval 21 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 22 Background 0.0953987 * 23 Beta(1) 0* 24 Beta(2) 3.97322e-005 * 25 26 * - Indicates that this value is not calculated. 27 28 29 30 Analysis of Deviance Table 31 32 Model Log(likelihood) # Param's Deviance Test d.f. P-value 33 Full model -79.5959 4 34 Fitted model -81.4024 2 3.61287 2 0.1642 35 Reduced model -85.3351 1 11.4782 3 0.009402 36 37 AIC: 166.805 38 39 40 Goodness of Fit 41 Scaled 42 Dose Est. Prob. Expected Observed Size Residual 43 44 0.0000 0.0954 6.773 10.000 71 1.304 45 1.4000 0.0955 4.583 2.000 48 - 1 . 2 6 8 46 7.1000 0.0972 4.666 4.000 48 - 0 . 3 2 5 47 71.0000 0.2596 12.979 13.000 50 0.007 48 49 Chi^2 = 3.41 d.f. = 2 P-value = 0.1814 50 51 52 Benchmark Dose Computation 53 54 Specified effect 0.01 55 56 Risk Type Extra risk 57 58 Confidence level 0.95 59 60 BMD 15.9045 61 62 BMDL 2.1388 63 64 BMDU 26.2712 65 66 T a k e n t o g e t h e r , (2.1388 , 26.27 1 2 ) is a 90 two-sided confidence 67 interval for the BMD 68 69 Multistage Cancer Slope Factor 0.00467551 70 This document is a draftfor review purposes only and does not constitute Agency policy. F-137 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.18.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 2 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:58 04/01 2010 3 4 National T oxicology Program, 1982: Lung: Alveolar/Bronchiolar A denom a or Carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-138 DRAFT--DO NOT CITE OR QUOTE 1 F.2.19. National Toxicology Program, 1982: Liver: Hepatocellular Adenoma or Carcinoma 2 F.2.19.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Multistage Cancer, 1-Degree a 2 0.916 258.572 1.338E+00 8.620E-01 M ultistage Cancer, 2-D egree 2 0.916 258.572 1.338E+00 8.620E-01 final =0 M ultistage Cancer, 3-Degree 2 0.916 258.572 1.338E+00 8.620E-01 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.19.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 1982: Liver: H epatocellular A denom a or Carcinoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 1 9 m s c l I P e r c m i c e m l i v e r a d e n carc.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 1 9 m s c l l P e r c m i c e m l i v e r a d e n c a r c . p l t 14 T h u A p r _ 01 1 2 : 5 9 : 2 8 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 4 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.22264 46 Beta(1) = 0.0074005 47 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-139 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 Background Beta(1) 4 5 Background 1 -0.46 6 7 Beta(1) -0.46 1 8 9 10 11 P a r a m e t e r E s t i m a t e s 12 13 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 14 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 15 B a c k g r o u n d 0.219315 16 Beta(1) 0.00750879 17 18 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 19 20 21 22 Analysis of Deviance Table 23 24 Model Log(likelihood) # Param's Deviance Test d.f. P-value 25 Full model -127.199 4 26 Fitted model -127.286 2 0.174343 2 0.9165 27 Reduced model -135.589 1 16.7801 3 0.0007843 28 29 AIC: 258.572 30 31 32 Goodness of Fit 33 Scaled 34 Dose Est. Prob. Expected Observed Size Residual 35 36 0.0000 0.2193 16.010 15.000 73 -0.286 37 1.4000 0.2275 11.146 12.000 49 0.291 38 7.1000 0.2598 12.732 13.000 49 0.087 39 71.0000 0.5419 27.096 27.000 50 -0.027 40 41 Chi^2 = 0.17 d.f. = 2 P-value = 0.9164 42 43 44 Benchmark Dose Computation 45 46 Specified effect = 0.01 47 48 Risk Type = Extra ris 49 50 Confidence level = 0.95 51 52 BMD = 1.33848 53 54 BMDL = 0.861975 55 56 BMDU = 2.4671 57 58 T a k e n t o g e t h e r , (0.. 8 6 1 9 7 5 , 2 . 4 6 7 1 ) is a 90 two-sided confidence 59 interval for the BMD 60 61 M u l t i s t a g e C a n c e r S l o p e F a c t o r = 0.0116013 62 This document is a draftfor review purposes only and does not constitute Agency policy. F-140 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.19.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 11:59 04/01 2010 3 4 National T oxicology Program, 1982: Liver: H epatocellular A denom a or Carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-141 DRAFT--DO NOT CITE OR QUOTE 1 F.2.20. National Toxicology Program, 2006: Liver: Cholangiocarcinoma 2 F.2.20.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) M ultistage Cancer, 1-Degree M ultistage Cancer, 2-D egree Multistage Cancer, 3-Degree a 5 5 4 0.024 129.070 1.872E+00 1.404E+00 0.947 114.349 9.440E+00 5.290E+00 0.995 115.158 1.310E+01 4.468E+00 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.20.2. Outputfor Selected Model: Multistage Cancer, 3-Degree 6 National T oxicology Program, 2006: Liver: Cholangiocarcinom a 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 2 0 m s c 3 I P e r c l i v c h o - c a r c . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 2 0 m s c 3 I P e r c l i v c h o - c a r c . p l t 14 T h u A p r 01 1 3 : 0 0 : 0 3 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 - b e t a i * d o s e ^ 1 - b e t a 2*dose^2-beta3*dose^3) ] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 6 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 4 34 Total number of specified parameters = 0 35 Degree of p olynomial = 3 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0 46 Beta(1) = 0.000561481 47 Beta(2) = 1.74365e-005 48 Beta(3) = 1.40248e-006 This document is a draftfor review purposes only and does not constitute Agency policy. F-142 D RAFT-- DO N O T CITE OR QUOTE 1 2 Asymptotic Correlation Matrix of Parameter Estimates 3 4 ( *** The model parameter(s) -Background -Beta(1) 5 have been estimated at a boundary point, or have been specified by the user, 6 and do not appear in the correlation matrix ) 7 8 Beta(2) Beta(3) 9 10 Beta(2) 1 -0.99 11 12 Beta(3) -0.99 1 13 14 15 16 P a r a m e t e r E s t i m a t e s 17 18 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 19 V a r i a b l e Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 20 Background 0 21 Beta(1) 0 22 Beta(2) 4.35927e-005 23 Beta(3) 1.14186e-006 24 25 Indicates that this value is not calculated. 26 27 28 29 Analysis of Deviance Table 30 31 M o d e l Log(likelihood) # Param's Deviance Test d.f. P-value 32 Full model -55.408 6 33 Fitted model -55.5789 2 0.34181 4 0.987 34 Reduced model -96.9934 1 83.1708 5 <.0001 35 36 AIC: 115.158 37 38 39 Goodness of Fit 40 Scaled 41 Dose Est. Prob. Expected Observed Residual 42 43 0.0000 0. 0 0 0 0 0.000 0.000 49 0.000 44 2.1400 0. 0 0 0 2 0.010 0.000 48 - 0 . 1 0 1 45 7.1400 0. 0 0 2 6 0.121 0.000 46 - 0 .349 46 15.7000 0. 0 1 5 0 0.752 1.000 50 0.288 47 32.9000 0. 0 8 4 1 4.121 4.000 49 -0.062 48 71.4000 0. 4 7 1 6 24.994 25.000 53 0.002 49 50 Chi^2 = 0.22 d.f. = 4 value = 0.9945 51 52 53 B e n c h m a r k Dose C o m p u t a t i o n 54 55 S p e c i f i e d effect 0.01 56 57 Risk Type Extra risk 58 59 Confidence level 0.95 60 61 BMD 13.1014 62 63 BMDL 4.46755 64 65 BMDU 19.1783 66 67 T a k e n t o g e t h e r , (4.46755, 19.17 8 3 ) is a 90 two-sided confidence 68 interval for the BMD 69 70 Multistage Cancer Slope Factor 0.00223836 This document is a draftfor review purposes only and does not constitute Agency policy. F-143 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.20.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 3 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 12:00 04/01 2010 3 4 National T oxicology Program, 2006: Liver: Cholangiocarcinom a This document is a draftfor review purposes only and does not constitute Agency policy. F-144 DRAFT--DO NOT CITE OR QUOTE 1 F.2.21. National Toxicology Program, 2006: Liver: Hepatocellular adenoma 2 F.2.21.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value M ultistage Cancer, 1-Degree M ultistage Cancer, 2-D egree Multistage Cancer, 3-Degree a 5 5 5 0.131 0.857 0.999 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 82.310 4.393E+00 2.915E+00 73.656 1.475E+01 8.618E+00 71.216 2.379E+01 1.153E+01 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.21.2. Outputfor Selected Model: Multistage Cancer, 3-Degree 6 National T oxicology Program, 2006: Liver: H epatocellular adenoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 2 1 m s c 3 I P e r c l i v h e p a t ad.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 2 1 m s c 3 I P e r c l i v h e p a t a d . p l t 14 T h u A p r 01 1 3 : 0 0 : 3 6 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 - b e t a i * d o s e ^ 1 - b e t a 2*dose^2-beta3*dose^3) ] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 6 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 4 34 Total number of specified parameters = 0 35 Degree of p olynomial = 3 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0 46 Beta(1) = 0 47 Beta(2) = 0 48 Beta(3)= 7.77141e-007 This document is a draftfor review purposes only and does not constitute Agency policy. F-145 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Asymptotic Correlation Matrix of Parameter Estimates 4 5 ( *** The model parameter(s) -Background -Beta(1) -Beta(2) 6 have been estimated at a boundary point, or have been specified by the user, 7 and do not appear in the correlation matrix ) 8 9 Beta(3) 10 11 B e t a ( 3 ) 1 12 13 14 15 P a r a m e t e r E s t i m a t e s 16 17 95.0% W a l d C o n f i d e n c e I n t e r v a l 18 V a r i a b l e Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 19 B a c k g r o u n d 0 20 Beta(1) 0 21 Beta(2) 0 22 Beta(3) 7.46408e-007 23 24 * - Indicates that this value is not calculated. 25 26 27 28 Analysis of Deviance Table 29 30 Model Log(likelihood) # Param's Deviance Test d.f. P-value 31 Full model -34.4075 6 32 Fitted model -34.6078 1 0.40065 5 0.9953 33 R e d u c e d model -56.3333 1 43.8515 5 <.0001 34 35 AIC: 71.2156 36 37 38 Goodness of Fit 39 Scaled 40 Dose Est. Prob. Expected Observed Residual 41 42 0.0000 0. 0 0 0 0 0. 0 0 0 0. 0 0 0 49 0. 0 0 0 43 2.1400 0. 0 0 0 0 0. 0 0 0 0. 0 0 0 48 -0. 019 44 7.1400 0. 0 0 0 3 0. 012 0. 0 0 0 46 -0. 112 45 15.7000 0. 0 0 2 9 0. 144 0. 0 0 0 50 -0. 380 46 32.9000 0. 0 2 6 2 1. 2 8 5 1. 0 0 0 49 -0. 255 47 71.4000 0. 2 3 7 9 12. 609 13. 000 53 0. 1 2 6 48 49 Chi^2 = 0.24 d.f. = 5 P-value = 0.9986 50 51 52 Benchmark Dose Computation 53 54 Specified effect 0.01 55 56 Risk Type Extra risk 57 58 Confidence level 0.95 59 60 BMD 23.7904 61 62 BMDL 11.5343 63 64 BMDU 27.8755 65 66 T a k e n t o g e t h e r , (11.5343, 27.87 5 5 ) is a 90 two-sided confidence 67 interval for the BMD 68 69 Multistage Cancer Slope Factor 0.000866978 70 This document is a draftfor review purposes only and does not constitute Agency policy. F-146 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.21.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 3 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 12:00 04/01 2010 3 4 National T oxicology Program, 2006: Liver: H epatocellular adenoma This document is a draftfor review purposes only and does not constitute Agency policy. F-147 DRAFT--DO NOT CITE OR QUOTE 1 F.2.22. National Toxicology Program, 2006: Oral mucosa: squamous cell carcinoma 2 F.2.22.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 4 0.386 M ultistage Cancer, 2-D egree 4 0.587 M ultistage Cancer, 3-Degree 4 0.587 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 125.484 4.751E+00 2.956E+00 123.245 1.635E+01 3.845E+00 123.245 1.635E+01 3.844E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.22.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 2006: Oral mucosa: squamous cell carcinoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 2 2 m s c l I P e r c o r a l carc.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 2 2 m s c l l P e r c o r a l c a r c . p l t 14 T h u A p r _ 01 1 3 : 0 1 : 1 1 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 6 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.00607545 46 Beta(1) = 0.00265195 47 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-148 DRAFT--DO NOT CITE OR QUOTE 1 Asymptotic Correlation Matrix of Parameter Estimates 2 3 Background Beta(1) 4 5 Background 1 -0.6 6 7 Beta(1) -0.6 1 8 9 10 11 P a r a m e t e r E s t i m a t e s 12 13 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 14 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 15 Background 0.0171416 * 16 Beta(1) 0.00211536 * 17 18 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 19 20 21 22 Analysis of Deviance Table 23 24 Model Log(likelihood) # Param's Deviance Test d.f. P-value 25 Full model -57.5353 6 26 Fitted model -60.7418 2 6.41293 4 0.1704 27 Reduced model -67.7782 1 20.4858 5 0.001013 28 29 AIC: 125.484 30 31 32 Goodness of Fit 33 Scaled 34 Dose Est. Prob. Expected Observed Residual 35 36 0.0000 0.0171 0.840 1.000 49 0.176 37 2.1400 0.0216 1.036 2.000 48 0.958 38 7.1400 0.0319 1.466 1.000 46 -0.391 39 15.7000 0.0492 2.462 0.000 50 -1.609 40 32.9000 0.0832 4.078 4.000 49 - 0 .040 41 71.4000 0.1549 8.211 10.000 53 0.679 42 43 Chi^2 = 4.15 d.f. = 4 P-value = 0.3855 44 45 46 Benchmark Dose Computation 47 48 Specified effect 0.01 49 50 Risk Type Extra risk 51 52 Confidence level 0.95 53 54 BMD 4.75111 55 56 BMDL 2.9556 57 58 BMDU 9.19454 59 60 T a k e n t o g e t h e r , (2.9556 , 9.19454) is a 90 two-sided confidence 61 i n t e r v a l for t he B M D 62 63 M u l t i s t a g e C a n c e r Slope Factor 0.0033834 64 This document is a draftfor review purposes only and does not constitute Agency policy. F-149 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.22.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 12:01 04/01 2010 3 4 National T oxicology Program, 2006: Oral mucosa: squamous cell carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-150 DRAFT--DO NOT CITE OR QUOTE 1 F.2.23. National Toxicology Program, 2006: Pancreas: adenoma or carcinoma 2 F.2.23.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 5 0.796 M ultistage Cancer, 2-D egree 5 0.977 M ultistage Cancer, 3-Degree 5 0.997 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 28.316 2.120E+01 9.335E+00 26.230 3.270E+01 1.389E+01 25.427 4.057E+01 1.755E+01 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.23.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 National T oxicology Program, 2006: Pancreas: adenoma or carcinoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 2 3 m s c l I P e r c p a n e ad carc.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 2 3 m s c l l P e r c p a n c a d c a r c . p l t 14 T h u A p r _ 01_ 1 3 : 0 1 : 4 3 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 6 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0 46 Beta(1) = 0.000817541 47 48 49 Asymptotic Correlation Matrix of Parameter Estimates This document is a draftfor review purposes only and does not constitute Agency policy. F-151 DRAFT--DO NOT CITE OR QUOTE 1 2 ( *** The model parameter(s) -Background 3 have been estimated at a boundary point, or have been specified by the user, 4 and do not appear in the correlation matrix ) 5 6 Beta(1) 7 8 Beta(1) 1 9 10 11 12 P a r a m e t e r E s t i m a t e s 13 14 95.0% W a l d C o n f i d e n c e I n t e r v a l 15 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 16 B a c k g r o u n d 0 17 B eta(1) 0 . 0 0 0 4 7 4 0 0 4 18 19 * - I n d i c a t e s t h a t t h i s v a l u e is not c a l c u l a t e d . 20 21 22 23 A n alysis of Deviance Table 24 25 Model Log(likelihood) # Param's Deviance Test d.f. P-value 26 Full model -11.4096 6 27 Fitted model -13.1581 1 3.49702 5 0.6238 28 Reduced model -16.7086 1 10.598 5 0.05996 29 30 AIC: 28.3163 31 32 33 G o o d n e s s of Fit 34 Scaled 35 Dose Est. Prob. Expected Observed Size Residu 36 37 0 .0000 0. 0 0 0 0 0. 0 0 0 0. 0 0 0 48 0. 0 0 0 38 2 .1400 0. 0 0 1 0 0. 0 4 9 0. 0 0 0 48 -0 221 39 7 .1400 0. 0 0 3 4 0. 1 5 5 0. 0 0 0 46 -0 395 40 15 .7000 0. 0 0 7 4 0. 3 7 1 0. 0 0 0 50 -0 611 41 32 .9000 0. 0 1 5 5 0. 7 4 3 0. 0 0 0 48 -0 869 42 71 .4000 0. 0 3 3 3 1. 697 3. 0 0 0 51 1. 017 43 44 i^ 2 = 2.37 d.f. = 5 P -value = 0.7964 45 46 47 Benchmark Dose Computation 48 49 Specified effect = 0.01 50 51 R i s k T y p e = Extra risk 52 53 C o n f i d e n c e l e vel = 0.95 54 55 BMD = 21.2031 56 57 BMDL = 9.33481 58 59 BMDU = 65.4351 60 61 T a k e n t o g e t h e r , (9.33481 , 65.4351) is a 90 two-sided confidence 62 interval for the BMD 63 64 Multistage Cancer Slope Factor = 0.00107126 65 This document is a draftfor review purposes only and does not constitute Agency policy. F-152 DRAFT--DO NOT CITE OR QUOTE 1 F.2.23.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level Fraction Affected 2 12:01 04/01 2010 3 4 National T oxicology Program, 2006: Pancreas: adenoma or carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-153 DRAFT--DO NOT CITE OR QUOTE 1 F.2.24. National Toxicology Program, 2006: Lung: Cystic keratinizing epithelioma 2 F.2.24.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value M ultistage Cancer, 1-Degree Multistage Cancer, 2-Degree a 5 5 0.192 0.771 M ultistage Cancer, 3-Degree 5 0.961 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 60.806 6.922E+00 4.187E+00 54.363 1.858E+01 1.069E+01 51.847 2.778E+01 1.556E+01 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.24.2. Outputfor Selected Model: Multistage Cancer, 2-Degree 6 National T oxicology Program, 2006: Lung: Cystic keratinizing epitheliom a 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 2 4 _ m s c 2 _ 1 P e r c _ l u n g _ e p i t h . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 2 4 m s c 2 I P e r c l u n g e p i t h . p l t 14 T h u A p r _ 01 1 3 : 0 2 : 1 9 2010 15 16 17 0 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1-beta2*dose^2)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 6 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 3 34 Total number of specified parameters = 0 35 Degree of p olynomial = 2 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0 46 Beta(1) = 0 47 Beta(2) = 3.77591e-005 48 This document is a draftfor review purposes only and does not constitute Agency policy. F-154 DRAFT--DO NOT CITE OR QUOTE 1 2 Asymptotic Correlation Matrix of Parameter Estimates 3 4 ( *** The model parameter(s) -Background -Beta(1) 5 have been estimated at a boundary point, or have been specified by the user, 6 and do not appear in the correlation matrix ) 7 8 Beta(2) 9 10 Beta(2) 1 11 12 13 14 P a r a m e t e r E s t i m a t e s 15 16 95.0% W a l d C o n f i d e n c e I n t e r v a l 17 V a r i a b l e Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 18 B a c k g r o u n d 0 19 Beta(1) 0 20 Beta(2) 2.91011e-005 21 22 Indicates that this value is not calculated. 23 24 25 26 Analysis of Deviance Table 27 28 Model Log(likelihood) # Param's Deviance Test d.f. P-value 29 Full model -23.958 6 30 Fitted model -26.1815 1 4.44693 5 0.487 31 R e d u c e d m o d e l -40.2069 1 32.4976 5 <.0001 32 33 AIC: 54.363 34 35 36 Goodness of Fit 37 Scaled 38 Dose Est. Prob. Expected Observed Size Residual 39 40 0.0000 0.0000 0. 0 0 0 0. 0 0 0 49 0. 0 0 0 41 2.1400 0.0001 0. 0 0 6 0. 0 0 0 48 -0. 080 42 7.1400 0.0015 0. 068 0. 0 0 0 46 -0. 261 43 15.7000 0.0071 0. 3 5 0 0. 0 0 0 49 -0. 594 44 32.9000 0.0310 1. 5 1 9 0. 0 0 0 49 -1. 252 45 71.4000 0.1379 7. 1 7 0 9. 0 0 0 52 0. 7 3 6 46 47 Chi^2 = 2.54 d.f. = 5 P-value = 0.7708 48 49 50 Benchmark Dose Computation 51 52 Specified effect 0.01 53 54 Risk Type Extra risk 55 56 Confidence level 0.95 57 58 BMD 18.5839 59 60 BMDL 10.6878 61 62 BMDU 25.1324 63 64 T a k e n t o g e t h e r , (10.6878, 25.13 2 4 ) is a 90 two-sided confidence 65 interval for the BMD 66 67 Multistage Cancer Slope Factor 0.000935646 68 69 This document is a draftfor review purposes only and does not constitute Agency policy. F-155 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.24.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 2 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 12:02 04/01 2010 3 4 National T oxicology Program, 2006: Lung: Cystic keratinizing epitheliom a This document is a draftfor review purposes only and does not constitute Agency policy. F-156 DRAFT--DO NOT CITE OR QUOTE 1 F.2.25. Toth et al., 1979: Liver: Tumors 2 F.2.25.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value Multistage Cancer, 1-Degree a 1 0.254 M ultistage Cancer, 2-D egree 1 0.254 AIC BMD BMDL (ng/kg-d) (ng/kg-d) 155.946 2.689E+00 1.522E+00 155.946 2.689E+00 1.522E+00 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.25.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 Toth et al., 1979: Liver: Tumors 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ C a n c \ 2 5 m s c l I P e r c ad r co r 1yr.(d) 13 G n u p l o t P l o t t i n g F i l e : C : \ C a n c \ 2 5 m s c l l P e r c a d r c o r l y r . p l t 14 T h u A p r 01_ 1 3 : 1 0 : 2 5 2010 15 16 17 T a b l e 1 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -beta1*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = Mean 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 3 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.240176 46 Beta(1) = 0.00374745 47 48 49 Asymptotic Correlation Matrix of Parameter Estimates 50 51 Background Beta(1) This document is a draftfor review purposes only and does not constitute Agency policy. F-157 DRAFT--DO NOT CITE OR QUOTE 1 2 Background 1 -0.53 3 4 Beta(1) -0.53 1 5 6 7 8 Parameter Estimates 9 10 95.0% W a l d C o n f i d e n c e I n t e r v a l 11 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 12 B a c k g r o u n d 0.2418 13 Beta(1) 0.00373791 14 15 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 16 17 18 19 A n a l y s i s of D e v i a n c e T a b l e 20 21 Model Log(likelihood) # Param's Deviance Test d.f. P-value 22 Full model -75.3127 3 23 Fitted model -75.9728 2 1.3201 1 0.2506 24 Reduced model -79.4897 1 8.35401 2 0.01534 25 26 AIC: 155.946 27 28 29 Goodness of Fit 30 Scaled 31 Dose Est. Prob. Expected Observed Size Residual 32 33 0.0000 0.2418 9.188 7.000 38 - 0 .829 34 1.0000 0.2446 10.764 13.000 44 0 . 7 8 4 35 100.0000 0.4783 21.044 21.000 44 - 0 . 0 1 3 36 37 Chi^2 = 1.30 d.f. = 1 P-value = 0.2537 38 39 40 Benchmark Dose Computation 41 42 Specified effect 0.01 43 44 Risk Type Extra risk 45 46 Confidence level 0.95 47 48 BMD 2.68876 49 50 BMDL 1.52183 51 52 BMDU 7.54263 53 54 T a k e n t o g e t h e r , (1.52183, 7.54263) is a 90 two-sided confidence 55 inter v a l for the BMD 56 57 Multistage Cancer Slope Factor 0.00657103 58 This document is a draftfor review purposes only and does not constitute Agency policy. F-158 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.25.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 12:10 04/01 2010 3 4 Toth et al., 1979: Liver: Tumors This document is a draftfor review purposes only and does not constitute Agency policy. F-159 DRAFT--DO NOT CITE OR QUOTE 1 F.2.26. Della Porta et al., 1987: Table 4, B6C3 mice, male, hepatocellular carcinoma 2 F.2.26.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) M ultistage Cancer, 1-Degree Multistage Cancer, 2-Degree a 1 1 0.073 164.110 9.255E+00 6.946E+00 0.899 160.823 7.359E+01 9.825E+00 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.26.2. Outputfor Selected Model: Multistage Cancer, 2-Degree 6 D ella Porta et al., 1987: Table 4, B6C 3 m ice, m ale, hepatocellular carcinoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ 9 4 _ D P o r t a _ 1 9 8 7 _ M a l e _ H e p _ C a r c _ M u l t i C a n c 2 _ 1 . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 9 4 _ D P o r t a _ 1 9 8 7 _ M a l e _ H e p _ C a r c _ M u l t i C a n c 2 _ 1 . p l t 14 Fri A p r 02 1 3 : 5 8 : 0 2 2010 15 16 17 T a b l e 4, B 6 C 3 m i c e , M a l e , H e p a t o c e l l u l a r c a r c i n o m a 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1-beta2*dose^2)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = DichEff 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 3 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 3 34 Total number of specified parameters = 0 35 Degree of p olynomial = 2 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.110507 46 Beta(1) = 0 47 Beta(2) = 1.88069e-006 48 49 50 Asymptotic Correlation Matrix of Parameter Estimates 51 This document is a draftfor review purposes only and does not constitute Agency policy. F-160 DRAFT--DO NOT CITE OR QUOTE 1 ( *** The model parameter(s) -Beta(1) 2 have been estimated at a boundary point, or have been specified by the user, 3 and do not appear in the correlation matrix ) 4 5 Background Beta(2) 6 7 Background 1 -0.62 8 9 Beta(2) -0.62 1 10 11 12 13 P a r a m e t e r E s t i m a t e s 14 15 9 5 . 0 % W a l d C o n f i d e n c e I n t e r v a l 16 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 17 B a c k g r o u n d 0.114031 18 B e t a(1) 0 19 Beta(2) 1.8559e-006 20 21 * - In d i c a t e s that this v a l u e is not calculated. 22 23 24 25 Analysis of Deviance Table 26 27 Model Log(likelihood) # Param's Deviance Test d.f. P-value 28 Full model -78.4036 3 29 Fitted model -78.4116 2 0.0160146 1 0.8993 30 Reduced model -94.7394 1 32.6717 2 <.0001 31 32 AIC: 160.823 33 34 35 Goodness of Fit 36 Scaled 37 Dose Est. Prob. Expected Observed Size Residual 38 39 0.0000 0.1140 4.903 5.000 43 0.046 40 357.1429 0.3008 15.340 15.000 51 -0.104 41 714.2857 0.6563 32.815 33.000 50 0.055 42 43 Chi^2 = 0.02 d.f. = 1 P-value = 0.8994 44 45 46 Benchmark Dose Computation 47 48 Specified effect 0.01 49 50 Risk Type Extra risk 51 52 Confidence level 0.95 53 54 BMD 73.5891 55 56 BMDL 9.82517 57 58 BMDU 88.9247 59 60 T a k e n t o g e t h e r , (9.82517, 88.9247) is a 90 two-sided confidence 61 i n t e r v a l for t he B M D 62 63 M u l t i s t a g e C a n c e r Slope Factor 0.00101779 64 This document is a draftfor review purposes only and does not constitute Agency policy. F-161 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.26.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 2 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 12:58 04/02 2010 3 4 D ella Porta et al., 1987: Table 4, B6C 3 m ice, m ale, hepatocellular carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-162 DRAFT--DO NOT CITE OR QUOTE 1 F.2.27. Della Porta et al., 1987: Table 4, B6C3 mice, female, hepatocellular adenoma 2 F.2.27.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of X2pFreedom Value Multistage Cancer, 1-Degree a Multistage Cancer, 2-Degree 1 0 0.468 NA AIC BMD BMDL (ng/kg-d) (ng/kg-d) 99.355 3.695E+01 2.245E+01 100.803 1.345E+02 2.353E+01 Notes aBest-fitting model, BMDS output presented in this appendix 3 4 5 F .2.27.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 D ella Porta et al., 1987: Table 4, B6C 3 m ice, fem ale, hepatocellular adenoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ 9 5 _ D P o r t a _ 1 9 8 7 _ F e m a l e _ H e p _ A d e n _ M u l t i C a n c 1 _ 1 . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 9 5 _ D P o r t a _ 1 9 8 7 _ F e m a l e _ H e p _ A d e n _ M u l t i C a n c 1 _ 1 . p l t 14 Fri A p r 02 1 3 : 5 8 : 3 2 2010 15 16 17 T a b l e 4, B 6 C 3 m i c e , F e m a l e , H e p a t o c e l l u l a r a d e n o m a 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -betai*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = DichEff 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 3 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.0244051 46 Beta(1) = 0.000306055 47 48 49 Asymptotic Correlation Matrix of Parameter Estimates 50 51 Background Beta(1) This document is a draftfor review purposes only and does not constitute Agency policy. F-163 DRAFT--DO NOT CITE OR QUOTE 1 2 Background 1 -0.72 3 4 Beta(1) -0.72 1 5 6 7 8 Parameter Estimates 9 10 95.0% W a l d C o n f i d e n c e I n t e r v a l 11 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 12 Background 0.0369416 13 Beta(1) 0.000272012 14 15 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 16 17 18 19 A n a l y s i s of D e v i a n c e T a b l e 20 21 Model Log(likelihood) # Param's Deviance Test d.f. P-value 22 Full model -47.4015 3 23 Fitted model -47.6775 2 0.552146 1 0.4574 24 Reduced model -51.6367 1 8.47042 2 0.01448 25 26 AIC: 99.3551 27 28 29 Goodness of Fit 30 Scaled 31 Dose Est. Prob. Expected Observed Size Residual 32 33 0.0000 0.0369 1.810 2.000 49 0.144 34 357.1429 0.1261 5.296 4.000 42 - 0 . 6 0 2 35 714.2857 0.2070 9.936 11.000 48 0 . 379 36 37 Chi^2 = 0.53 d.f. = 1 P-value = 0.4677 38 39 40 Benchmark Dose Computation 41 42 Specified effect 0.01 43 44 Risk Type Extra risk 45 46 Confidence level 0.95 47 48 BMD 36.9482 49 50 BMDL 22.4477 51 52 BMDU 86.1826 53 54 T a k e n t o g e t h e r , (22.4477, 86.1826) is a 90 two-sided confidence 55 inter v a l for the BMD 56 57 Multistage Cancer Slope Factor 0.000445481 58 This document is a draftfor review purposes only and does not constitute Agency policy. F-164 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.27.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 12:58 04/02 2010 3 4 D ella Porta et al., 1987: Table 4, B6C 3 m ice, fem ale, hepatocellular adenoma This document is a draftfor review purposes only and does not constitute Agency policy. F-165 DRAFT--DO NOT CITE OR QUOTE 1 F.2.28. Della Porta et al., 1987: Table 4, B6C3 mice, female, hepatocellular carcinoma 2 F.2.28.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model Degrees of x2pFreedom Value AIC BMD BMDL (ng/kg-d) (ng/kg-d) Multistage Cancer, 1-Degree a M ultistage Cancer, 2-D egree 1 1 0.010 116.588 2.425E+01 1.605E+01 0.010 116.588 2.425E+01 1.605E+01 final =0 Notes a Best-fitting m odel, BM D S output presented in this appendix 3 4 5 F .2.28.2. Outputfor Selected Model: Multistage Cancer, 1-Degree 6 D ella Porta et al., 1987: Table 4, B6C 3 m ice, fem ale, hepatocellular carcinoma 7 8 9 10 11 M u l t i s t a g e C a n c e r M o d e l . ( V e r s i o n : 1.7; D a t e : 0 5 / 1 6 / 2 0 0 8 ) 12 I n p u t D a t a File: C : \ 1 \ 9 6 _ D P o r t a _ 1 9 8 7 _ F e m a l e _ H e p _ C a r c _ M u l t i C a n c 1 _ 1 . ( d ) 13 G n u p l o t P l o t t i n g F i l e : C : \ 1 \ 9 6 _ D P o r t a _ 1 9 8 7 _ F e m a l e _ H e p _ C a r c _ M u l t i C a n c 1 _ 1 . p l t 14 Fri A p r 02 1 3 : 5 9 : 0 1 2010 15 16 17 T a b l e 4, B 6 C 3 m i c e , F e m a l e , H e p a t o c e l l u l a r c a r c i n o m a 18 19 20 The f o r m of the p r o b a b i l i t y f u n c t i o n is: 21 22 P[response] = background + (1-background)*[1-EXP( 23 -beta1*dose^1)] 24 25 The parameter betas are restricted to be positive 26 27 28 Dependent variable = DichEff 29 Independent variable = Dose 30 31 T o t a l n u m b e r of o b s e r v a t i o n s = 3 32 Total number of records with missing values = 0 33 Total n u m b e r of p a r a m e t e r s in m odel = 2 34 Total number of specified parameters = 0 35 Degree of p olynomial = 1 36 37 38 Maximum number of iterations = 250 39 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 40 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 41 42 43 44 Default Initial Parameter Values 45 Background = 0.0903848 46 Beta(1) = 0.000261828 47 48 49 Asymptotic Correlation Matrix of Parameter Estimates 50 51 Background Beta(1) This document is a draftfor review purposes only and does not constitute Agency policy. F-166 D RAFT-- DO N O T CITE OR QUOTE 1 2 Background 1 -0.8 3 4 Beta(1) -0.8 1 5 6 7 8 Parameter Estimates 9 10 95.0% W a l d C o n f i d e n c e I n t e r v a l 11 V a r i a b l e Estimate Std. Err Lower Conf. Limit Upper Conf. Limit 12 Background 0.0300271 13 Beta(1) 0.000414523 14 15 * - I n d i c a t e s t h a t t h i s v a l u e is n o t c a l c u l a t e d . 16 17 18 19 A n a l y s i s of D e v i a n c e T a b l e 20 21 Model Log(likelihood)1 # Param's Deviance Test d.f. P-value 22 Full model -53.1726 3 23 Fitted model -56.2941 2 6.24292 1 0.01247 24 Reduced model -60.7146 1 15.084 2 0.0005303 25 26 AIC: 116.588 27 28 29 Goodness of Fit 30 Scaled 31 Dose Est. Prob. Expected Observed Size Residual 32 33 0.0000 0.0300 1.471 1.000 49 - 0 .395 34 357.1429 0.1635 6.867 12.000 42 2 . 142 35 714.2857 0.2786 13.373 9.000 48 - 1 . 4 0 8 36 37 Chi^2 = 6.72 d.f. = 1 P-value = 0.0095 38 39 40 Benchmark Dose Computation 41 42 Specified effect 0.01 43 44 Risk Type Extra risk 45 46 Confidence level 0.95 47 48 BMD 24.2455 49 50 BMDL 16.0512 51 52 BMDU 49.7176 53 54 T a k e n t o g e t h e r , (16.0512, 49.7176) is a 90 two-sided confidence 55 inter v a l for the BMD 56 57 Multistage Cancer Slope Factor 0.000623007 58 This document is a draftfor review purposes only and does not constitute Agency policy. F-167 DRAFT--DO NOT CITE OR QUOTE Fraction Affected 1 F.2.28.3. F ig u r e f o r S e le c te d M o d e l: M u ltis ta g e C a n cer, 1 -D e g re e Multistage Cancer Model with 0.95 Confidence Level 2 12:59 04/02 2010 3 4 D ella Porta et al., 1987: Table 4, B6C 3 m ice, fem ale, hepatocellular carcinoma This document is a draftfor review purposes only and does not constitute Agency policy. F-168 DRAFT--DO NOT CITE OR QUOTE F.3. R E FE R E N C E S 1 D ella Porta G; Dragani TA; S ozzi D; S ozzi G. (1978) C arcinogenic effects o f infantile and long-term 2 ,3 ,7 ,8 2 tetrachlorodibenzo-p-dioxin treatment in the m ouse. Tumori 73: 99-107. 3 Goodman, DG; Sauer, RM . (1992) H epatotoxicity and carcinogenicity in fem ale Sprague-D aw ley rats treated w ith 4 2,3,7,8-tetrachlorordibenzo-p-dioxin (TCDD): a Pathology W orking Group reevaluation. R egul T oxicol Pharmacol 5 15:245-252. 6 K ociba, RJ; K ey es, D G ; B eyer, JE; et al. (1 9 7 8 ) R esu lts o f a tw o -y ea r chronic to x icity and on co g en icity study o f 7 2,3,7,8-tetrachlorodibenzo-p-dioxin in rats. T oxicol A ppl Pharm acol 4 6 (2 ):2 7 9 -3 0 3 . 8 NTP (National T oxicology Program). (1982) Bioassay o f 2,3,7,8-tetrachlorodibenzo-p-dioxin for possible 9 carcinogenicity (gavage study). Tech. Rept. Ser. N o. 201. U .S. Department o f H ealth and Human Services, Public 1 0 Health Service, R esearch Triangle Park, NC. 11 NTP (National Toxicology Program). (2006) NTP technical report on the toxicology and carcinogenesis studies o f 1 2 2,3,7,8-tetrachlorodibenzo-p-dioxin (T C D D ) (C A S N o. 1746-01-6) in fem ale Harlan Sprague-D aw ley rats (Gavage 13 Studies). Natl T oxicol ProgramTech Rep 521. Public Health Service, N ational Institute o f Health, U .S. Department 1 4 o f H ealth and Human Services, R esearch Triangle Park, NC. 1 5 Toth, KJ; Sugar, S; S o m fa i-R elle S; et al. (1 9 7 8 ) C arcinogenic b io a ssa y o f the herbicide 2,4,5-trich lorph en oxy 1 6 ethanol (TCPE) w ith Sw iss mice. Prog B iochem Pharmacol 14:82-93. This document is a draftfor review purposes only and does not constitute Agency policy. F-l 69 DRAFT--DO NOT CITE OR QUOTE [This page intentionally left blank.] DRAFT DO N O T CITE OR QUOTE May 2010 External R eview Draft APPENDIX G Endpoints Excluded From Reference Dose Derivation Based on Toxicological Relevance NOTICE THIS D O C U M E N T IS A N E X T ER N A L REVIEW DRA FT. It has not been form ally released by the U .S. Environmental Protection A gency and should not at this stage be construed to represent A gency policy. It is being circulated for com m ent on its technical accuracy and policy im plications. National Center for Environmental A ssessm ent O ffice o f Research and D evelopm ent U .S. Environmental Protection A gency Cincinnati, OH 1 A P P E N D IX G . E N D P O IN T S E X C L U D E D F R O M R E F E R E N C E D O SE D E R IV A T IO N 2 B A SED O N T O X IC O L O G IC A L R E L E V A N C E 3 4 5 T he N atio n al A cadem y o f S ciences (N A S) co m m ittee com m ented on th e lo w dose m odel 6 p red ictio n s and the n eed to discu ss th e b io lo g ical significance o f th e n o n can cer h ealth effects 7 m odeled in th e 2003 R eassessm ent. In selecting p o in t o f departure (P O D ) candidates from the 8 anim al b io assay s fo r d eriv atio n o f th e referen ce d o se (R fD ), U .S. E n v iro n m en tal P ro tectio n 9 A gency (E P A ) had to co n sid er th e to xicological relev an ce o f th e iden tified en dpoint(s) from any 10 g iv en study. O ften en d p o in ts/effects m ay b e sensitive, b u t lack g en eral to x ico lo g ical 11 s ig n ific a n c e d u e to n o t b e in g c le a rly a d v e rs e (d e fin e d in th e In te g ra te d R is k In fo rm a tio n S y ste m 12 (IR IS ) glo ssary as a bio ch em ical change, functional im pairm ent, or p ath o lo g ic lesio n th at affects 13 th e p e rfo rm a n c e o f th e w h o le o rg a n ism , o r re d u c e s a n o rg a n is m 's ab ility to re sp o n d to an 14 additional environm ental challenge), b ein g an adaptive response, or n o t b ein g clearly lin k ed to 15 d o w n stre am fu n c tio n a l o r p a th o lo g ical alteratio n s. It is stan d ard E P A R fD d eriv a tio n p o licy n o t 16 to b ase a reference valu e on endpoints th at are n ot adverse or not obvious precursors to an 17 ad v erse effect. F o r select studies, a ratio n ale fo r lack o f to x ico lo g ical relev an ce o f p articu lar 18 en d p o in ts rep o rted is listed here. T h ese en d p o in ts w ere n o t co n sid ered fo r d eriv atio n o f th e R fD . 19 K itch in and W o o d s (1979) ad m in istered fem ale S prague-D aw ley rats a single gav ag e 20 dose o f 2 ,3 ,7 ,8 -tetrach lo ro d ib en zo -p -d io x in (T C D D ) and m easu red cy to ch ro m e P 4 5 0 lev els and 21 b en zo (a)p y ren e h ydroxylase (B P H ) activity as a m ark er o f h ep atic m icrosom al cytochrom e 22 P 4 4 8 -m ed iated en zy m e activity. T hey fo u n d a statistically sig n ifican t in crease in B P H at doses 23 >2 ng/kg and a significant increase in cytochrom e P 450 levels at doses > 600 ng/kg. A ryl 24 hyd ro carb o n h ydrolase and E R O D w ere b o th significantly increased 3 m onths after exposure; 25 h o w ev er th e elev atio n did n o t m ain tain statistical significance at 6 m onths. N o o th er in d icato rs 26 o f h ep atic effects w ere analyzed. C Y P in d u ctio n alone is n o t co n sid ered a significant 27 toxico lo g ically adverse effect given th at C Y P s are in duced as a m eans o f h epatic p rocessing o f 28 x en o b io tic agents. A d d itio n ally , th e ro le o f C Y P in d u ctio n in h ep ato to x icity and carcin o g en icity 29 o f T C D D is u n know n, and C Y P in d u ctio n is n o t con sid ered a relev an t P O D w ith o u t obvious 30 pathological significance. 31 In m u ltip le stu d ies b y H a sso u n et al. (1 9 9 8 , 2 0 0 0 , 2 0 0 2 , 2 0 0 3 ), v a rio u s in d ic a to rs o f 32 ox id ativ e stress w ere m easu red in h ep atic and b rain tissu e o f fem ale B 6C 3F 1 m ice and S prague- This document is a draftfo r review purposes only and does not constitute Agency policy. G-1 DRAFT--DO NOT CITE OR QUOTE 1 D a w le y rats fo llo w in g 13 or 3 0 w e e k s o f T C D D g a v a g e d o sin g (5 d ays a w ee k ). B iom ark ers for 2 o x id a tiv e stress in clu d ed p rod u ction su p eroxid e an ion, lip id p eroxid ation , and D N A sin gle-stran d 3 breaks. T h e authors report a statistically sig n ifica n t effec t on several o x id a tiv e stress m arkers as 4 a resu lt o f T C D D exp osu re, th e lo w e st d o se p rod u cin g an effec t b ein g 0 .3 2 n g /k g -d a y (H assou n 5 et al., 1 9 9 8 ). In th is stu d y, all o x id a tiv e stress m ark ers w er e sig n ifica n tly effec te d , b u t n o other 6 in d ica to rs o f brain p a th o lo g y w er e a sse sse d . T h us, it is im p ra ctica b le to lin k th e m arkers o f 7 o x id a tiv e stress to a to x ic o lo g ic a l o u tc o m e in th e brain, and th is stud y and its en d p o in ts are n ot 8 co n sid ered relevan t P O D can d idates. 9 B u r le so n et al. (1 9 9 6 ) a n a ly ze d th e e ffe c t o f a T C D D o n viral h o st re sista n ce fo llo w in g a 10 sin g le g a v a g e d o se o f T C D D b y m ea su rin g m o rta lity m ed ia ted b y in flu e n z a v iru s c h a lle n g e in 11 B 6 C 3 F 1 fe m a le m ic e . T h e stu d y au th ors fo u n d th at T C D D at > 1 0 n g /k g -d a y in cr ea se d 12 in flu en za -in d u ced m ortality. T h e exp erim en tal d esig n ca lls for a 30% m ortality in untreated 13 a n im a ls (15% w a s a ch ie v ed ); m ortality, itself, is n o t a d irect resu lt o f T C D D ex p o su re. N o n e o f 14 th e other im m u n o lo g ica lly -relev a n t m easu res w ere a ffected b y T C D D treatm en t in th is study, 15 and n o other effec ts w ere reported. T h e interpretation o f th ese resu lts w ith resp ect to h um ans is 16 p ro b lem a tic. F u rth erm o re, th e fin d in g s w e r e n o t rep ro d u ced b y N o h a r a et al. (2 0 0 2 ) u sin g th e 17 sam e exp erim en tal d esig n (see S ectio n 2 .4 .2 ). T herefore, th is en d p oin t is n ot con sid ered relevant 18 as a P O D candidate. 19 T o exam in e the central n ervou s system resp on se to T C D D , K u ch iiw a et al (2 0 0 2 ) 2 0 an a ly zed th e effec t o f in u tero and lactation al T C D D ex p o su re on th e seroton ergic sy stem in the 21 b rain stem o f m a le d d Y m ice. F em a le m ice w ere ad m in istered T C D D b y oral g a v a g e o n ce a 2 2 w e e k fo r 8 w e e k s prior to p reg n a n cy and, u sin g an im m u n o cy to ch em ica l d etectio n m eth o d , th e 23 raphe n u clei in th e b rain stem o f m a le offsp rin g w a s m on itored for sero to g erg ic neurons. T C D D 2 4 at 0 .7 n g /k g -d a y cau sed a 2 5 -5 0 % red u ction in th e im m u n o sta in in g o f serotonin , h o w ev er there 25 w e r e n o d iffe r e n c e s in extern al m o rp h o lo g y , birth or p o stn atal b o d y w e ig h ts b e tw e e n 26 T C D D -ex p o sed and control offsp rin g. T h e authors su g g est that th ese fin d in g s m ay in d icate that 2 7 T C D D acts as a n eu roteratogen b y m ed ia tin g lo n g -term alteration s in n eu ron al seroton in 28 sy n th esis and seroton ergic fu n ction . H o w ev er, n o other relevan t n eu rotoxicity en d p oin ts w ere 29 exam in ed or reported. T hus, redu ced seroton in is n ot an ad verse en d p oin t o f to x ico lo g ica l 30 sig n ifica n ce in and o f itself, and th is study is d eem ed u n su itab le as a P O D candidate. This document is a draftfo r review purposes only and does not constitute Agency policy. G-2 DRAFT--DO NOT CITE OR QUOTE 1 M a lly and C hipm an (2 0 0 2 ) evalu ated the effec t o f T C D D on gap ju n ction s, 2 h y p o th esizin g that as a n o n g en o to x ic carcin ogen , T C D D m ay in d u ce tum or form ation b y 3 d isturbin g tissu e h o m eo sta sis. F em a le F 3 4 4 rats w ere d o sed w ith T C D D b y oral g a v a g e for 4 eith er 3 co n se cu tiv e d ays or 2 d ays a w e e k for 2 8 days. G ap ju n ctio n co n n ex in (C x ) p laqu e 5 ex p ressio n and h ep a to cy te p roliferation w a s m easu red . T h e stud y authors report a d ecrea se in 6 C x 3 2 p laq u e n u m b er and area in th e liv e r o f rats e x p o se d to 0 .7 n g /k g -d a y and h igher, h o w e v er 7 th ey did n o t fin d an a sso cia ted in crea se in h ep a to cy te p roliferation . N o clin ica l sig n s o f to x ic ity 8 w ere ob served , and h isto lo g ica l ex am in ation o f th e liv er rev ea led n o ab n orm alities. In the 9 a b sen ce o f ad dition al in d icators o f h ep a to to x icity , a d ecrease in C x 3 2 p laq u e form ation is n ot 10 clearly lin k ed to T C D D -m ed ia ted h ep a to to x icity or h ep atocarcin ogen icty, nor is it con sid ered an 11 a d v erse e ffe c t. T h is en d p o in t is n o t c o n sid er ed a to x ic o lo g ic a lly re le v a n t P O D . 12 V a n d en H e u v e l et al. (1 9 9 4 ) a n a ly ze d ch a n g es in h ep a tic m R N A fo llo w in g a sin g le 13 a d m in istratio n o f T C D D to fe m a le S p r a g u e -D a w ley rats b y oral g a v a g e. F ou r d a y s after 14 treatm ent, an im als w ere sa crificed and liv ers w ere ex cised . U sin g reverse tran scriptase- 15 p o ly m era se ch ain reaction (R T -P C R ) on h ep atic R N A , th ey com p ared le v e ls o f " d io x in 16 r e sp o n siv e " m R N A 's (C Y P 1 A 1 , U D P -g lu c u r o n o sy ltr a n sfe r a s e I, p la sm in o g e n a ctiv a to r in h ib ito r 17 2, and transform ing grow th factor a ) at variou s d o ses o f T C D D and at control (b a selin e) lev els. 18 T h ey determ in ed that C Y P 1A 1 elicited th e m o st sen sitiv e resp on se to T C D D , w ith a statistically 19 sig n ifica n t in crea se (3 -fo ld ) in m R N A from rat liv ers e x p o se d to 1 n g /k g -d a y T C D D . In d u ction 2 0 o f C Y P 1 A 1 ex p ressio n is n ot co n sid ered an ad verse effect, as th e role o f C Y P 1 A 1 in 21 T C D D -m ed ia ted ca rcin o g en icity is u n settled . T h erefore, in th e a b sen ce o f other in d icators o f 22 h ep atoxicity, in creases in liv er C Y P 1A 1 can n ot b e con sid ered to x ic o lo g ic a lly relevan t for a P O D 23 can d idate. 2 4 D e v ito et al. (1 9 9 4 ) a sse sse d th e a ctiv ity o f C Y P 1 A 1 and C Y P 1 A 2 , th e a m o u n t o f 25 p h osp h orylation o f p h osp h otyrosyl p rotein s (p p 32, p p 34, and p p 38), and th e le v e ls o f estrogen 2 6 recep tor in the liver, uterus, lu n g and sk in tissu e o f fem a le B 6 C 3 F 1 m ice ad m in istered T C D D for 27 5 d ays a w e e k for 13 w eek s. T h e authors h y p o th esized that th ese m easu rem en ts m ay b e 28 sen sitiv e b iom arkers for ex p o su re to T C D D . B o d y w e ig h ts w ere a lso recorded w eek ly . 29 In d u ction o f C Y 1 A 1 and C Y P 1 A 2 , as w e ll as in creased p h osp h orylated form s o f p p 32, p p34, 30 and p p 38 w ere sen sitiv e in d icators o f T C D D exp osu re, w ith statistically sig n ifica n t ch an ges seen 31 at 1.07 n g /k g -d a y . E R O D a ctiv ity in th e lin g , skin, and liv e r w a s a lso o b serv ed w ith sig n ifica n t This docum ent is a draftfo r review purposes only a n d does not constitute Agency policy. G -3 D R A F T -- D O N O T C ITE O R Q U O T E 1 in crea ses at th is d ose. H o w ev er, th e authors d id n ot fin d a ch a n g e in rat b o d y or term in al organ 2 w eig h ts, n or did th ey n o te an y p a th o lo g y in th e an im als at th is d o se le v el. T h e role o f C Y P s and 3 p h o sp h orylated p p 3 2 , p p 3 4 , and p p 38 in T C D D -m ed ia ted to x ic ity is u n k n o w n , and ch a n g es in 4 th e a ctivity or fu n ctio n o f th ese p rotein s are n ot co n sid ered ad verse. T h erefore, th ese en d p oin ts 5 are n ot con sid ered su itable as P O D s. 6 B e c a u se T C D D had b een d etected in th e soil o f con tam in ated lo ca tio n s, d eterm in in g the 7 b io a v a ila b ility o f T C D D from in g ested soil m ay b e im portant to th e calcu la tio n o f safe exp o su re 8 le v e ls . L u c ie r et al. (1 9 8 6 ) fe d ad u lt fe m a le S p r a g u e -D a w le y rats T C D D co n ta m in a ted so il or 9 g a v e them T C D D in corn oil at variou s d o ses and com pared the effects o f T C D D on b io ch em ica l 10 param eters from liv er tissu e. T h ey fou n d that eq u iv a len t d o ses o f T C D D in corn oil and soil 11 p ro d u ced sim ila r in c r e a se s in h ep a tic aryl h y d ro ca rb o n h y d r o x y la se a ctiv ity (A H H ) and U D P 12 glu cu ron yltran sferase activity. T h ey determ in ed that A H H w a s statistically in d u ced 1 .8 -fo ld at 13 15 n g /k g in corn o il and 4 0 n g /k g in so il. C y to ch ro m e P 4 5 0 w a s sig n ifica n tly in crea sed at h ig h er 14 d o ses. N o clin ica l sig n s o f acu te to x ic ity or ch a n g es in b o d y w e ig h t w ere ob served . T he 15 asso cia tio n b etw een A H H activity and T C D D -m ed ia ted h ep a to to x icity is u n k n ow n and n o 16 a d v erse en d p o in ts w ere m easu red . T h us, th is en d p o in t is n ot su itab le as a P O D can d idate. 17 S u g ita -K o n ish i e t al. (2 0 0 3 ) in v e stig a te d th e c h a n g e in h o st re sista n c e o f m ic e o ffsp r in g 18 lactation ally ex p o sed to T C D D . P regnan t C 5 7 B L /6 N C ji m ice w ere ad m in istered T C D D v ia 19 d rin k in g w a ter from parturition to w e a n in g o f th e o ffsp rin g (1 7 d ays). O n e grou p o f o ffsp rin g 2 0 w a s th en in fe c te d w ith Listeria m onocytogenes and b lo o d and sp lee n sa m p les w e r e c o lle c te d 21 v a rio u s tim e p o in ts p o st in fectio n . U n in fected , T C D D ex p o se d o ffsp rin g w ere w e ig h e d and their 2 2 sp leen s and th y m u ses rem o v ed for assay o f cellu lar con ten t and p rotein ex p ression . T C D D 23 ex p o su re cau sed a sta tistica lly -sig n ifica n t d ecrease in relative sp leen w eig h t and a statistically- 2 4 sig n ifica n t in crea se in th y m ic C D 4 + c e lls in th e h ig h -d o se grou p (1 1 .3 n g /k g -d a y ). O ffsp rin g 25 in fe c te d w ith Listeria fo llo w in g T C D D ex p o su re ex h ib ited a sta tistica lly sig n ific a n t in cr ea se in 2 6 serum tu m or n ecro sis factor alpha (T N F -a ) 2 d ays after in fectio n in b oth se x e s in th e lo w - 2 7 (1 .1 4 n g /k g -d a y ) and h ig h -d o se grou p s. T h e authors co n clu d e that ex p o su re to T C D D d isrupted 2 8 th e h o st resista n ce o f th e o ffsp rin g at th e lo w e s t d o se tested , d esp ite th e p rim ary im m u n e 29 param eters b ein g u n affected . W ith ou t an o b v io u s association b etw een T C D D and im m u n e 30 fu n ction , h ow ever, this en d p oin t is n ot su itab le for id en tification o f a L O A E L . T hus, the 31 L O A E L fo r th is stu d y is 11.3 n g /k g -d a y , and th e N O A E L is 1.1 4 n g /k g -d a y . This docum ent is a draftfo r review purposes only a n d does not constitute Agency policy. G -4 D R A F T -- D O N O T C ITE O R Q U O T E 1 S e w a ll et al. (1 9 9 3 ) in v e stig a te d a ltera tio n s in th e ep id erm a l g ro w th fa cto r recep to r 2 (E G F R ) p ath w ay in a tw o -sta g e in itiation p rom otion m o d el o f T C D D h ep atic cancer. E G F R 3 sig n a lin g h as b een im p lica ted in th e altered cell gro w th in d u ctio n b y tu m or p rom oters. F em a le 4 S p ra g u e-D a w ley rats w ere ad m in istered T C D D b iw e e k ly b y oral g a v a g e for 3 0 w e e k s fo llo w in g 5 in itiation b y a sin g le d o se o f d ieth yln itrosam in e (D E N ). A group a lso receiv ed T C D D w ith ou t 6 prior D E N in itiation . L iv ers w ere h arvested and fix e d from sa crificed an im als and sectio n s 7 tested for E G F R b in d in g, au top h osp h orylation , im m u n o lo ca liza tio n , and h ep atic cell 8 p roliferation. T h e authors report a sig n ifica n t d o se-d ep en d en t d ecrease in p lasm a m em brane 9 E G F R m axim u m b in d in g cap acity in T C D D -e x p o se d rats b eg in n in g at 3 .5 n g/k g -d a y . H o w ev er, 10 at th is sam e d o se, th e authors n o te a statistically sig n ifica n t d ecrease in ce ll p roliferation (as 11 m ea su red b y D N A re p lic a tio n la b e lin g ), w ith in c r e a se s in p ro lifera tio n o n ly o ccu rrin g at h ig h er 12 d o ses (1 2 5 n g/k g-d ay). N o other in dicators o f h ep atic to x icity or tu m origen icity w ere a ssessed . 13 T h e ro le o f E G F R in T C D D -m e d ia te d h ep a to to x icity and h ep a to ca rcin o g en icity is u n k n o w n , and 14 as such, this en d p oin t can n ot b e u n eq u iv o ca lly lin k ed to T C D D -in d u ced h ep atic effects nor 15 la b eled as ad verse. T h us, it is n ot su itab le as a P O D can d idate. 16 17 G .1. R E FE R E N C E S 18 Burleson, GR; Lebrec, H; Y ang, YG; et al. (1996) E ffect o f 2,3,7,8-tetrachlorodibenzo-p-dioxin (TC D D ) on 19 influenza virus host resistance in m ice. Fund Appl T oxicol 29:40-47. 2 0 D evito, MJ; M a, X; Babish, JG; et al. (1994) D ose-resp onse relationships in m ice fo llo w in g subchronic exposure to 2 1 2.3.7.8- tetrachlorodibenzo-p-dioxin: CYP1A1, C YP1A2, estrogen receptor, and protein tyrosine phosphoylation. 2 2 Toxicol Appl Pharmacol 124:82-90. 2 3 H assoun, EA; W ilt, SC; D eV ito, MJ; et al. (1998) Induction o f oxidative stress in brain tissues o f m ice after 2 4 subchronic exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. T oxicol Sci 42:23-27. 2 5 H assoun, EA; Li, F; Abushaban, A; et al. (2000) The relative abilities o f TC D D and its congeners to induce 2 6 oxidative stress in the hepatic and brain tissues o f rats after subchronic exposure. T oxicology 1 4 5:103-113. 2 7 Hassoun, EA; W ang, H; Abushaban, A. (2002) Induction o f oxidative stress follow ing chronic exposure to TCDD, 2 8 2.3.4.7.8- pentachlorodibenzofuran, and 2 ,3 ',4 ,4 ',5-pentachlorobiphenyl. J T oxicol Environ H ealth A 6 5 :8 2 5 -8 4 2 . 2 9 Hassoun, EA; Al-Ghafri, M; Abushaban, A. (2003) The role o f antioxidant enzym es in TC DD-induced oxidative 3 0 stress in various brain regions o f rats after subchronic exposure. Free Rad B io l M edicine 3 5 (9 ):1 0 2 8 -1 0 3 6 . 31 Kitchin, KT; W oods, JS. (1979) 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TC D D) effects on hepatic m icrosom al 3 2 cytochrome P-448-m ediated enzym e activities. Toxicol Appl Pharmacol 47:537-546. This document is a draftfo r review purposes only and does not constitute Agency policy. G-5 DRAFT--DO NOT CITE OR QUOTE 1 K uchiiw a, S; C heng, S-B; N agatom o, I; et al. (2 0 0 2 ) In utero and lactational exp osu re to 2,3,7,8-tetrachlorodibenso2 /-dioxin decreases serotonin-immunoreactive neurons in raphe nuclei o f male m ouse offspring. N eurosci Lett 3 317:73-76. 4 Lucier, GW; R um baugh, RC; M cC o y , Z; et al. (1 9 8 6 ) In gestion o f so il contam inated w ith 2,3,7,8-tetrachloro5 dibenzo-p-dioxin (T C D D ) alters hepatic enzym e activities in rats. Fund A ppl T oxicol 6 :3 6 4 -3 7 1 . 6 M ally, A; Chipman, JK. (2002) N on-genotoxic carcinogens: early effects on gap junctions, cell proliferation and 7 apoptosis in the rat. T o x ic o lo g y 1 8 0 :2 3 3 -2 4 8 . 8 N ohara, K; Izum i, H; Tamura, S; et al. (2 0 0 2 ) E ffec t o f lo w -d o se 2,3,7,8-tetrach lorod ib en zo-p -d ioxin (T C D D ) on 9 influenza A virus-induced mortality in m ice. T oxicology 170:131-138. 1 0 Sew all, CH; Lucier, GW; Tritscher, AM ; et al. (1993) TC D D -m ediated changes in hepatic epiderm al grow th factor 11 receptor may be a critical event in the hepatocarcinogenic action o f TCDD. Carcinogenesis 14:1885-1893. 1 2 Sugita-K onishi, Y; K obayashi, K; N aito, H; et al. (2003) E ffect o f lactational exposure to 2 ,3 ,7 ,8 13 tetrachlorodibenzo-p-dioxin o n the susceptibility to Listeria infection. B io sc i B io tech n o l B io ch em 6 7 (1 ):8 9 -9 3 . 1 4 U .S. EPA. (2003) Exposure and human health reassessm ent o f 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and 15 related com pounds [N A S review draft]. V olum es 1 -3 . N ational Center for Environm ental A ssessm ent, W ashington, 1 6 D C ; E P A /6 0 0 /P -0 0 /0 0 1 Cb. A vailab le at: h ttp ://w w w .ep a .g o v /n cea w w w 1 /p d fs/d io x in /n a s-rev iew /. 1 7 V anden H euvel, JP; Clark, GC; Tritscher, A; et al. (1 9 9 4 ) A ccu m u lation o f polych lorinated d ib en zo-p -d ioxin s and 1 8 dibenzofurans in liver o f control laboratory rats. Fundam A ppl T ox ico l 2 3 :46 5 -4 6 9 . 19 This document is a draftfo r review purposes only and does not constitute Agency policy. G-6 DRAFT--DO NOT CITE OR QUOTE [T h is p a g e in ten tio n a lly le ft b lan k .] DRAFT D O N O T C ITE O R Q U O T E M ay 2010 E xternal R ev ie w D raft APPENDIX H Cancer Precursor Benchmark Dose Modeling N O T IC E T H IS D O C U M E N T IS A N E X T E R N A L R E V IE W D R A F T . It has n ot b een form ally released b y the U .S . E n viron m en tal P rotection A g en cy and sh ou ld n ot at th is stage b e con stru ed to represent A g en cy p o licy . It is b ein g circulated for com m en t on its tech n ical accuracy and p o licy im p lication s. N ation al C enter for E n viron m en tal A ssessm en t O ffice o f R esearch and D evelop m en t U .S . E n viron m en tal P rotection A g en cy C in cin n ati, O H C O N T E N T S-- A PPE N D IX H: C ancer Precursor B enchm ark D ose M odeling A P P E N D I X H . C A N C E R P R E C U R S O R B E N C H M A R K D O S E M O D E L I N G ..........................H - 1 H .1 . B M D S I N P U T T A B L E S ................................................................................................................................................ H - 1 H . 1 .1 . H a s s o u n e t a l. ( 2 0 0 0 ) ....................................................................................................................................... H - 1 H . 1 .2 . K i t c h i n a n d W o o d s ( 1 9 7 9 ) ........................................................................................................................... H - 1 H . 1 .3 . N a t i o n a l T o x i c o l o g y P r o g r a m ( 2 0 0 6 ) , 3 1 W e e k E x p o s u r e ...............................................H - 2 H . 1 .4 . N a t i o n a l T o x i c o l o g y P r o g r a m ( 2 0 0 6 ) , 5 3 W e e k E x p o s u r e ...............................................H - 2 H . 1 .5 . V a n d e n H e u v e l e t a l. ( 1 9 9 4 ) .......................................................................................................................H - 2 H .2 . A L T E R N A T E D O S E : W H O L E B L O O D B M D S R E S U L T S .......................................................... H - 3 H . 2 .1 . H a s s o u n e t a l., 2 0 0 0 : C y t o c h r o m e C R e d u c t a s e ........................................................................ H - 3 H . 2 . 1 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ........................................................H - 3 H . 2 . 1 . 2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 5 ) ...................................................... H - 3 H . 2 . 1 . 3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 5 ) ........................................................H - 6 H . 2 .2 . H a s s o u n e t a l., 2 0 0 0 : D N A S i n g l e - S t r a n d B r e a k s ....................................................................H - 7 H . 2 . 2 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ........................................................H - 7 H . 2 . 2 . 2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ...................................................... H - 7 H . 2 . 2 . 3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) .....................................................H - 1 0 H . 2 . 2 . 4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n t e d : P o w e r , U n r e s t r i c t e d .............. H - 1 1 H . 2 . 2 . 5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n t e d : P o w e r , U n r e s t r i c t e d ..............H - 1 3 H . 2 .3 . H a s s o u n e t a l., 2 0 0 0 : T B A R S ............................................................................................................... H - 1 4 H . 2 . 3 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .....................................................H - 1 4 H . 2 . 3 . 2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l .................................................................................... H - 1 4 H . 2 . 3 . 3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ......................................................................................H - 1 7 H . 2 .4 . K i t c h i n a n d W o o d s , 1 9 7 9 : B a p H y d r o x y l a s e A c t i v i t y ...................................................... H - 1 8 H . 2 . 4 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .....................................................H - 1 8 H . 2 . 4 . 2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 5 ) ................................................... H - 1 8 H . 2 . 4 . 3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 5 ) .....................................................H - 2 1 H . 2 .5 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : L i v e r E R O D 5 3 W e e k s .................................... H - 2 2 H . 2 . 5 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .....................................................H - 2 2 H . 2 . 5 . 2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l .................................................................................... H - 2 2 H . 2 . 5 . 3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l ......................................................................................H - 2 5 H . 2 .6 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : L u n g E r o d 5 3 W e e k s .........................................H - 2 6 H . 2 . 6 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .....................................................H - 2 6 H . 2 . 6 . 2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ................................................... H - 2 6 H . 2 . 6 . 3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) .....................................................H - 2 9 H . 2 . 6 . 4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n t e d : P o w e r , U n r e s t r i c t e d .................H - 3 0 H . 2 . 6 . 5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n t e d : P o w e r , U n r e s t r i c t e d .................H - 3 2 H . 2 .7 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : L a b e l i n g I n d e x 3 1 W e e k s .............................. H - 3 3 H . 2 . 7 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .....................................................H - 3 3 H . 2 . 7 . 2 . O u t p u t f o r S e l e c t e d M o d e l : P o l y n o m i a l , 5 - d e g r e e .......................................... H - 3 3 H . 2 . 7 . 3 . F i g u r e f o r S e l e c t e d M o d e l : P o l y n o m i a l , 5 - d e g r e e ............................................H - 3 6 H . 2 .8 . V a n d e n H e u v e l e t a l., 1 9 9 4 : H e p a t i c C Y P 1 A 1 M r n a E x p r e s s i o n ............................ H - 3 7 H . 2 . 8 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .....................................................H - 3 7 H . 2 . 8 . 2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l .................................................................................. H - 3 7 This document is a draftfor review purposes only and does not constitute Agency policy. H-ii DRAFT--DO NOT CITE OR QUOTE C O N T E N T S (continued) H . 2 . 8 . 3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l .................................................................................... H - 4 0 H .3 . A D M I N I S T E R E D D O S E B M D S R E S U L T S ........................................................................................... H - 4 1 H . 3 .1 . H a s s o u n e t a l., 2 0 0 0 : C y t o c h r o m e C R e d u c t a s e ..................................................................... H - 4 1 H . 3 . 1 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .....................................................H - 4 1 H . 3 . 1 . 2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) .................................................. H - 4 1 H . 3 . 1 . 3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ................................................... H - 4 4 H . 3 . 1 . 4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n t e d : P o w e r , U n r e s t r i c t e d .............. H - 4 4 H . 3 . 1 . 5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n t e d : P o w e r , U n r e s t r i c t e d ..............H - 4 7 H . 3 .2 . H a s s o u n e t a l., 2 0 0 0 : D N A S i n g l e - S t r a n d B r e a k s .................................................................. H - 4 8 H . 3 . 2 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s ......................................................H - 4 8 H . 3 . 2 . 2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ................................................................................... H - 4 8 H . 3 . 2 . 3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l .................................................................................... H - 5 1 H . 3 . 2 . 4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n t e d : H i l l , U n r e s t r i c t e d ......................H - 5 1 H . 3 . 2 . 5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n t e d : H i l l , U n r e s t r i c t e d ..................... H - 5 4 H . 3 .3 . H a s s o u n e t a l., 2 0 0 0 : T B A R S ............................................................................................................... H - 5 5 H . 3 . 3 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .....................................................H - 5 5 H . 3 . 3 . 2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) .................................................. H - 5 5 H . 3 . 3 . 3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ................................................... H - 5 8 H . 3 . 3 . 4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n t e d : P o w e r , U n r e s t r i c t e d .............. H - 5 8 H . 3 . 3 . 5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n t e d : P o w e r , U n r e s t r i c t e d ..............H - 6 1 H . 3 .4 . K i t c h i n a n d W o o d s , 1 9 7 9 : B a p H y d r o x y l a s e A c t i v i t y ...................................................... H - 6 2 H . 3 . 4 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .....................................................H - 6 2 H . 3 . 4 . 2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 5 ) .................................................. H - 6 2 H . 3 . 4 . 3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 5 ) ................................................... H - 6 5 H . 3 .5 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : L i v e r E R O D 5 3 W e e k s .................................... H - 6 6 H . 3 . 5 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .....................................................H - 6 6 H . 3 . 5 . 2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ................................................................................... H - 6 6 H . 3 . 5 . 3 . F i g u r e f o r S e l e c t e d M o d e l : H i l l .................................................................................... H - 6 9 H . 3 .6 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : L u n g E r o d 5 3 W e e k s .........................................H - 7 0 H . 3 . 6 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .....................................................H - 7 0 H . 3 . 6 . 2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) .................................................. H - 7 0 H . 3 . 6 . 3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 4 ) ................................................... H - 7 3 H . 3 . 6 . 4 . O u t p u t f o r A d d i t i o n a l M o d e l P r e s e n t e d : P o w e r , U n r e s t r i c t e d .............. H - 7 3 H . 3 . 6 . 5 . F i g u r e f o r A d d i t i o n a l M o d e l P r e s e n t e d : P o w e r , U n r e s t r i c t e d ..............H - 7 6 H . 3 .7 . N a t i o n a l T o x i c o l o g y P r o g r a m , 2 0 0 6 : L a b e l i n g I n d e x 3 1 W e e k s .............................. H - 7 7 H . 3 . 7 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .....................................................H - 7 7 H . 3 . 7 . 2 . O u t p u t f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 2 ) .................................................. H - 7 7 H . 3 . 7 . 3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 2 ) ................................................... H - 8 0 H . 3 .8 . V a n d e n H e u v e l e t a l., 1 9 9 4 : H e p a t i c C Y P 1 A 1 M r n a E x p r e s s i o n ............................ H - 8 1 H . 3 . 8 . 1 . S u m m a r y T a b l e o f B M D S M o d e l i n g R e s u l t s .....................................................H - 8 1 H . 3 . 8 . 2 . O u t p u t f o r S e l e c t e d M o d e l : H i l l ................................................................................... H - 8 1 H . 3 . 8 . 3 . F i g u r e f o r S e l e c t e d M o d e l : E x p o n e n t i a l ( M 5 ) ................................................... H - 8 4 This document is a draftfor review purposes only and does not constitute Agency policy. H-iii DRAFT--DO NOT CITE OR QUOTE 1 A P PE N D IX H . C A N C E R P R E C U R SO R B E N C H M A R K D O SE M O D E L IN G 2 3 4 H .1 . B M D S IN P U T T A B L E S 5 H .1 .1 . H a sso u n et al. (2 0 0 0 ) Endpoint Cytochrome C reductase d DNA single-strand breaks f TBARs e Administered Dose (ng/kg-day) 0 3 10 22 46 100 Internal Dose (ng/kg blood) a 0 1.94 4.61 8.15 14.01 25.34 n=6 0.15 0.07 n=6 0.18 0.05 b n=6 0.19 0.06 n=6 n=6 n= 6 0.27 0.06 c 0.39 0.06 c 0.44 0.11 c 7.41 1.54 10.78 1.25 b'c 13.6 1.69 c 15.3 1.71 c 20.4 2.25 c 23.5 1.37 c 1.47 0.29 1.55 0.54 b 2.15 0.36 c 2.28 0.25 c 2.62 0.52 c 2.29 0.49 c aFrom the E m on d P B P K m od el d escribed in 3.3. bLO EL for selected endpoint. cStatistically sign ifican t as com pared to control (p < 0.05). dV alu es are the m ean SD. D ata obtained from Table 1 in H assou n et al. 2000. eV alu es are the m ean SD. D ata obtained from T able 2 in H assou n et al. 20 0 0 . V a lu e s are the m ean SD. D ata obtained from T able 3 in H assou n et al. 20 0 0 . 6 7 8 H .1 .2 . K itc h in a n d W o o d s (1 9 7 9 ) Administered Dose (ng/kg-day) 0 0.6 2 4 20 60 Internal Dose (ng/kg blood) a Endpoint BaP hydroxylase activity f (continued on next line) 0 n=9 4.9 0.37 0.06 n=4 0.20 n=4 0.38 n=4 1.61 n=4 4.9 0.59 b 6.7 0.70c,d 7.2 0.90 d 8.3 0.13 e Administered Dose (ng/kg-day) 4.15 n=4 14 2.5 e 200 600 2000 5000 20,000 Internal Dose (ng/kg blood) a Endpoint BaP hydroxylase activity f (continued) 11.59 n=4 59 3.4 e 30.26 n=4 96 23 e 90.90 n=4 155 8.2 e 218.02 n=4 182 13e 863.18 n=4 189 13e aF rom the E m ond P B P K m od el describ ed in 3.3. bN O E L for selected endpoint. cL O E L for selected endpoint. S ta tistic a lly significant as com pared to control (p < 0.05). S ta tistica lly significant as com pared to control (p < 0.001). V a lu es are the m ean SE. Data obtained from Table 3 in K itchin and W oods 1979. 9 This document is a draftfor review purposes only and does not constitute Agency policy. H-1 DRAFT--DO NOT CITE OR QUOTE 1 H .1 .3 . N a tio n a l T o x ico lo g y P ro g r a m (2 0 0 6 ), 31 W e e k E x p o su re Endpoint 0 0 n=9 2.14 2.33 n = 10 Administered Dose (ng/kg-day) 7.14 15.7 Internal Dose (ng/kg blood) a 5.32 n = 10 9.21 n = 10 32.9 15.66 n = 10 71.4 28.13 n = 10 Labeling Index ,week 31 c 0.33 0.0.06 0.85 0.21 b 0.96 0.23 b 0.79 0.15 b 1.33 0.36 b 3.85 0.97 b aFrom the E m on d P B P K m od el d escribed in 3.3. S ta tistic a lly significant as com pared to control (p < 0.05). cV alu es are the m ean SE. D ata obtained from Table 11 in N T P 2006. 2 3 4 H .1 .4 . N a tio n a l T o x ico lo g y P ro g r a m (2 0 0 6 ), 53 W e e k E x p o su re Administered Dose (ng/kg-day) 0 2.14 7.14 15.7 32.9 71.4 Internal Dose (ng/kg blood) a Endpoint 0.00 n=8 2.46 n=8 5.53 n=8 9.54 n=8 16.18 n=8 29.04 n=8 Liver EROD, week 53 c Lung EROD, week 53 c 30.22 1.59 569.38 24.62 b 1280.00 95.30 b 1551.16 112.36 b 1726.81 107.58 b 1871.47 109.14 b 3.01 0.56 27.15 1.87 b 42.85 3.94 b 36.57 4.59 b 43.75 6.56 b 43.71 2.24 b aFrom the E m on d P B P K m od el d escribed in 3.3. S ta tistic a lly significant as com pared to control (p < 0.01). cV alu es are the m ean SE. D ata obtained from Table 12 in N T P 2006. 5 6 H .1 .5 . V a n d e n H e u v e l et al. (1 9 9 4 ) Administered Dose (ng/kg-day) Endpoint 0 0.1 1 10 100 Internal Dose (ng/kg blood) a 1,000 10,000 0.00 0.01 0.11 0.88 6.45 48.32 434.50 Hepatic CYP1A1 mRNA Expression c n = 13 5.4 1.0 n=5 7.2 2.5 n = 12 14.8 4.3 b n=7 12.8 1.7 b n=7 n = 11 n=5 536 121b 18000 4590 36700 9900 bb aF rom the E m ond P B P K m od el describ ed in 3.3. Statistically sign ifican t as com pared to control (p < 0.05). cV alu es are the m ean SE. D ata obtained from Table 2 in van d en H e u v e l 1994. 7 This document is a draftfor review purposes only and does not constitute Agency policy. H-2 DRAFT--DO NOT CITE OR QUOTE 1 H .2 . A L T E R N A T E D O S E : W H O L E B L O O D B M D S R E S U L T S 2 H .2 .1 . H a sso u n et al., 2 0 0 0 : C y to c h r o m e C R e d u c ta se 3 H .2 .1 .1 . Summary Table o f BMDS Modeling Results______________ Model a Degrees of Freedom x 2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes exponential (M 2) 4 0.016 -143.333 9.274E+00 7.737E+00 exponential (M 3) 4 0 .0 1 6 -1 4 3 .3 3 3 9 .2 7 4 E + 0 0 7 .7 3 7 E + 0 0 p o w er hit b oun d (d = 1) exponential (M 4) exponential (M5) b H ill 3 0.339 -150.139 3.364E+00 2.170E+00 2 0.788 -151.027 5.913E+00 3.102E+00 2 0.743 -150.910 6.208E+00 3.190E+00 linear polynom ial, 5degree power 4 0.170 -149.086 5.613E+00 4.429E+00 4 0.170 -149.086 5.613E+00 4.429E+00 4 0 .1 7 0 -1 4 9 .0 8 6 5 .6 1 3 E + 0 0 4 .4 2 9 E + 0 0 p o w er b ound hit (p ow er = 1) a C onstant variance m odel selected (p = 0 .3 8 7 1 ) b Best-fitting m odel, BM D S output presented in this appendix 4 5 6 H .2 .1 .2 . Outputfor Selected Model: Exponential (M5) 7 H a sso u n et al., 2 0 0 0 : C y to ch ro m e C red u ctase 8 9 10 11 E x p o n e n t i a l M o d e l . ( V e r s i o n : 1 . 6 1 ; D a t e : 7 / 2 4 / 2 0 0 9 ) 12 I n p u t D a t a File: C : \ 5 \ B l o o d \ 1 7 _ H a s _ 2 0 0 0 _ C y t C L i v _ E x p C V _ 1 . ( d ) 13 G n u p l o t P l o t t i n g Fil e : 14 Fri A p r 30 1 4 : 1 4 : 3 4 2 0 1 0 15 16 17 T B A R s , l i v e r o n l y ( T a b l e 2) 18 19 20 The form of the response function by Model: 21 M o d e l 2: Y dose] = a * exp{sign * b * dose} 22 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 23 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 24 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 25 26 Note: Y[dose] is the m e dian response for exposure = dose; 27 sign = +1 for increasing trend in data; 28 sign = -1 for d e c r e a s i n g trend. 29 30 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 31 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 32 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 33 34 35 D e p endent v a r i a b l e = Mea n This document is a draftfor review purposes only and does not constitute Agency policy. H-3 DRAFT--DO NOT CITE OR QUOTE 1 Independent variable = Dose 2 Data are assumed to be distributed: normally 3 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 4 r h o is s e t t o 0. 5 A constant v a r i a n c e model is fit. 6 7 Total number of dose groups = 6 8 Total number of records with missing values = 0 9 Maximum number of iterations = 250 10 R e l a t i v e F u n c t i o n C o n v e r g e n c e has b e e n set to: 1 e - 0 0 8 11 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 12 13 M L E s o l u t i o n p r o v i d e d : E x a c t 14 15 16 I n i t i a l P a r a m e t e r V a l u e s 17 18 Variable Model 5 19 20 lnalpha -5.48625 21 rho(S) 0 22 a 0.1387 23 b 0.0225296 24 c 6.40231 25 d 1 26 2 7 (S) = S p e c i f i e d 28 29 30 31 P a r a m e t e r E s t i m a t e s 32 33 Variable Model 5 34 35 lnalpha -5.47298 36 rho 0 37 a 0.156024 38 b 0.0891513 39 c 2.85355 40 d 2.14235 41 42 43 Table of Stats From Input Data 44 45 Dose N Obs Mean Obs Std Dev 46 47 06 0.146 0.06614 48 1.938 6 0.177 0.05389 49 4.614 6 0.191 0.05634 50 8.147 6 0.271 0.05634 51 14.01 6 0.388 0.06369 52 25.34 6 0.444 0.1102 53 54 55 E s t i m a t e d Val u e s of Interest 56 57 Dose Est Mean Est Std Scaled Residual 58 59 0 0.156 0. 0 6 4 8 -0.3789 60 1.938 0.1627 0. 0 6 4 8 0.5416 61 4.614 0.1961 0. 0 6 4 8 -0.1919 62 8.147 0.2705 0. 0 6 4 8 0.01769 63 14.01 0.3874 0. 0 6 4 8 0.02224 64 25.34 0.4443 0. 0 6 4 8 -0.0107 65 66 67 68 Other models for which likelihoods are calculated: 69 70 Model A1: Yij = Mu(i) + e ( i j 71 V ar{e(ij)} = Sigma^2 This document is a draftfor review purposes only and does not constitute Agency policy. H-4 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A2: Yij = Mu(i) + e(ij) 3 V a r { e i ;ij)} = S i g m a ( i ) ^ 2 4 5 Model A3: Yij = Mu(i) + e(ij) 6 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 7 8 M o d e l R: Yij = M u + e(i) 9 V a r { e i :ij)} = S i g m a ^ 2 10 11 12 L i k e l i h o o d s of I n t e r e s t 13 14 Model Log(likelihood) DF AIC 15 16 A1 80.75258 7 -147.5052 17 A2 83.37355 12 - 1 4 2 . 7 4 7 1 18 A3 80.75258 7 -147.5052 19 R 55.82002 2 -107.64 20 5 80.51364 5 -151.0273 21 22 23 Additive constant for all log-likelihoods = -33.08. This constant added to the 24 above values gives the log-likelihood including the term that does not 25 depend on the model parameters. 26 27 28 Explanation of Tests 29 30 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 31 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 32 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 33 34 T e s t 7a: D o e s M o d e l 5 f i t t h e d a t a ? (A3 v s 5) 35 36 37 Tests of Interest 38 39 Test -2*log(Likelihood Ratio) D. F. p-value 40 41 Test 1 55.11 10 < 0.0001 42 Test 2 5.242 5 0.3871 43 Test 3 5.242 5 0.3871 44 Test 7a 0.4779 2 0.7875 45 46 47 The p - v a l u e for T e s t 1 is less t h a n .05. T h e r e a p p e a r s to be a 48 difference between response and/or variances among the dose 49 levels, it seems appropriate to model the data. 50 51 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 52 variance model appears to be appropriate here. 53 54 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 55 v a r i a n c e appears to be app r o p r i a t e here. 56 57 T h e p - v a l u e f o r T e s t 7a is g r e a t e r t h a n .1. M o d e l 5 s e e m s 58 to a dequately describe the data. 59 60 61 B e n c h m a r k Dose Computations: 62 63 Sp e c i f i e d Effect = 1.000000 64 65 Risk Type = Estimated standard deviations from control 66 67 Confidence Level = 0.950000 68 69 BMD = 5.91298 70 71 BMDL = 3.10234 This document is a draftfor review purposes only and does not constitute Agency policy. H-5 DRAFT--DO NOT CITE OR QUOTE Mean R esponse 1 H.2.1.3. F ig u r e f o r S e le c te d M o d e l: E x p o n e n tia l (M 5) Exponential Model 5 with 0.95 Confidence Level 2 14:14 04/30 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. H-6 DRAFT--DO NOT CITE OR QUOTE 1 H.2.2. Hassoun et al., 2000: DNA Single-Strand Breaks 2 H.2.2.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model a Degrees of Freedom x 2pValue AIC BMD (ng/kg) BMDL (ng/kg) Notes exponential (M 2) 4 <0.0001 111.134 6.551E+00 5.472E+00 exponential (M 3) exponential (M4) b exponential (M 5) 4 < 0 .0 0 0 1 1 1 1.134 6 .5 5 1 E + 0 0 5 .4 7 2 E + 0 0 p o w er hit b oun d (d = 1) 3 0.231 78.588 1.207E+00 9.165E-01 3 0.231 7 8 .5 8 8 1 .207E + 00 9 .165E -01 p o w er hit b oun d (d = 1) H ill 3 0.230 7 8 .5 9 0 1 .097E + 00 7.9 6 6 E -0 1 n lo w er b ound hit (n = 1) linear 4 <.0001 97.616 3.552E+00 2.890E+00 polynom ial, 5degree 4 <.0001 97.616 3.552E+00 2.890E+00 power 4 < .0001 9 7 .6 1 6 3 .5 5 2 E + 0 0 2 .8 9 0 E + 0 0 p o w er b ound hit (p ow er = 1) power, unrestricted c 3 0.132 79.893 4.522E-01 2.027E-01 unrestricted (power = 0.576) a C onstant variance m odel selected (p = 0 .7 5 2 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 H .2 .2 .2 . O utputfo r Selected M odel: E xponential (M4) 6 H a sso u n et al., 2 0 0 0 : D N A sin g le-stra n d b reak s 7 8 Exponential Model. (Version: 1.61; Date: 7/24/2009) 9 Input Data File: C:\5\Blood\18_Has_2000_SSB_ExpCV_1.(d) 10 G n u p l o t P l o t t i n g File: 11 F r i A p r 30 1 4 : 1 5 : 1 6 2 0 1 0 12 13 14 D N A s i n g l e - s t r a n d b r e a k s , l i v e r o n l y ( T a b l e 3) 15 16 17 T h e f o r m of t h e r e s p o n s e f u n c t i o n b y M o d e l : 18 M o d e l 2: Y dose] = a * exp{sign * b * dose} 19 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 20 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 21 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 22 23 Note: Y[dose] is the m e d i a n response for exposure = dose; 24 sign = +1 for increasing trend in data; 25 sign = -1 for d e c r e a s i n g trend. 26 2 7 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 2 8 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 2 9 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 30 31 This document is a draftfor review purposes only and does not constitute Agency policy. H-7 DRAFT--DO NOT CITE OR QUOTE 1 Dependent variable = Mean 2 Independent variable = Dose 3 Data are assumed to be distributed: normally 4 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 5 r h o is s e t t o 0. 6 A constant v a r i a n c e model is fit. 7 8 Total number of dose groups = 6 9 Total number of records with missing values = 0 10 M a x i m u m n u m b e r of i t e r a t i o n s = 250 11 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 12 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1 e - 0 0 8 13 14 M L E s o l u t i o n p r o v i d e d : E x a c t 15 16 17 I n i t i a l P a r a m e t e r V a l u e s 18 19 Variable Model 4 20 21 lnalpha 0.841244 22 rho(S) 0 23 a 7.0395 24 b 0.103521 25 c 3.50522 26 d 1 27 2 8 (S) = S p e c i f i e d 29 30 31 32 Parameter Estimates 33 34 Variable Model 4 35 36 lnalpha 0.960789 37 rho 0 38 a 7.7528 39 b 0.075429 40 c 3.39665 41 d 1 42 43 44 Table of Stats From Input Data 45 46 Dose N Obs Mean Obs Std Dev 47 48 06 7.41 1.543 49 1.938 6 10.78 1.249 50 4.614 6 13.6 1.69 51 8.147 6 15.3 1.715 52 14.01 6 20.4 2.254 53 25.34 6 23.5 1.372 54 55 56 Estimated Values of Interest 57 58 Dose Est Mean Est Std Scaled Residual 59 60 0 7.753 1.617 -0.5194 61 1.938 10.28 1.617 0.7575 62 4.614 13.21 1.617 0.5853 63 8.147 16.28 1.617 -1.49 64 14.01 19.87 1.617 0.7958 65 25.34 23.59 1.617 -0.1293 66 67 68 69 Other models for which likelihoods are calculated: 70 71 Model A1: Yij = M u ( i ) + e (ij) This document is a draftfor review purposes only and does not constitute Agency policy. H-8 DRAFT--DO NOT CITE OR QUOTE 1 Var{e(ij)} = Sigma^2 2 3 Model A2: Yij Mu(i) + e(ij) 4 Var{e(ij)} Sigma(i)^2 5 6 Model A3: Yij Mu(i) + e(ij) 7 Var{e(ij)} e x p ( l a l p h a + log(mean(i)) * rho) 8 9 M o d e l R: Yij M u + e(i) 10 Var{e(ij)} Sigma^2 11 12 13 L i k e l i h o o d s o f I n t e r e s t 14 15 Model Log(likelihood) DF AIC 16 17 A1 -33.14239 7 80.28478 18 A2 -31.81197 12 87.62394 19 A3 -33.14239 7 80.28478 20 R -80.44209 2 164.8842 21 4 -35.29421 4 78.58842 22 23 24 Additive constant for all log-likelihoods = -33.08. This constant added to the 25 above values gives the log-likelihood including the term that does not 26 depend on the model parameters. 27 28 Explanation of Tests 29 30 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 31 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 32 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 33 34 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 35 36 37 Tests of Interest 38 39 Test -2*log(Likelihood Ratio) D. F. p-value 40 41 Test 1 97.26 10 < 0.0001 42 Test 2 2.661 5 0.7521 43 Test 3 2.661 5 0.7521 44 Test 6a 4.304 3 0.2305 45 46 47 The p - v a l u e for Test 1 is less t h a n .05. T h e r e a p p e a r s to be a 48 difference between response and/or variances among the dose 49 levels, it seems appropriate to model the data. 50 51 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 52 variance model appears to be appropriate here. 53 54 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 55 v a r i a n c e appears to be app r o p r i a t e here. 56 57 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 58 to a dequately describe the data. 59 60 61 B e n c h m a r k Dose Computations: 62 63 Sp e c i f i e d Effect = 1.000000 64 65 Risk Type = Estimated standard deviations from control 66 67 Confidence Level = 0.950000 68 69 BMD = 1.20684 70 71 BMDL = 0.916526 This document is a draftfor review purposes only and does not constitute Agency policy. H-9 DRAFT--DO NOT CITE OR QUOTE Mean R esponse 1 H.2.2.3. F ig u r e f o r S e le c te d M o d e l: E x p o n e n tia l (M 4) Exponential Model 4 with 0.95 Confidence Level 2 14:15 04/30 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. H-10 DRAFT--DO NOT CITE OR QUOTE 1 H.2.2.4. O u tp u tf o r A d d itio n a l M o d e l P r e s e n te d : P o w e r, U n re stric te d 2 Hassoun et al., 2000: DNA single-strand breaks 3 4 5 Power Model. (Version: 2.15; Date: 04/07/2008) 6 Input Data File: C:\5\Blood\18_Has_2000_SSB_PwrCV_U_1.(d) 7 Gnuplot Plotting File: C:\5\Blood\18_Has_2000_SSB_PwrCV_U_1.plt 8 Fri Apr 30 14:15:20 2010 9 10 11 D N A s i n g l e - s t r a n d b r e a k s , l i v e r o n l y ( T a b l e 3) 12 13 14 T h e f o r m of t h e r e s p o n s e f u n c t i o n is: 15 16 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e Ap o w e r 17 18 19 D e p e n d e n t v a r i a b l e = M e a n 20 Independent variable = Dose 21 rho is set to 0 22 The power is not restricted 23 A constant v a r i a n c e model is fit 24 25 Total number of dose groups = 6 26 Total number of records with missing values = 0 27 Maximum number of iterations = 250 28 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 29 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 30 31 32 33 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 34 alpha = 2.7831 35 rho = 0S p e c i f i e d 36 control = 7.41 37 slope = 2.16848 38 power = 0.620048 39 40 41 A s y m p t o t i c C o r r e l a t i o n M a t r i x of P a r ameter Estimates 42 43 ( *** The model parameter(s) -rho 44 have been estimated at a boundary point, or have been specified by the user, 45 and do not appear in the correlation matrix ) 46 47 alpha control slope power 48 49 alpha 1 2.5e-009 -4.6e-009 5.7e-009 50 51 control 2.5e-009 1 -0.79 0.66 52 53 slope -4.6e-009 -0.79 1 -0.97 54 55 power 5.7e-009 0.66 -0.97 1 56 57 58 59 Parameter Estimates 60 61 95.0% W a l d Confi d e n c e Interval 62 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 63 alpha 2.71022 0.638804 1.45818 3.96225 64 control 7.26415 0.644159 6.00163 8.52668 65 slope 2.60017 0.530762 1.55989 3.64044 66 power 0.575946 0.0589669 0.460373 0.691519 67 68 69 70 Table of Data and Estimated Values of Interest This document is a draftfor review purposes only and does not constitute Agency policy. H-11 DRAFT--DO NOT CITE OR QUOTE 1 2 Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled R 3 4 5 06 7.41 7.26 1.54 1.65 0.217 6 1.938 6 10.8 11.1 1.25 1.65 -0.432 7 4.614 6 13.6 13.5 1.69 1.65 0.094 8 8.147 6 15.3 16 1.71 1.65 -0.993 9 14.01 6 20.4 19.2 2.25 1.65 1.85 10 25.34 6 23.5 24 1.37 1.65 - 0 . 7 3 5 11 12 13 14 M o d e l D e s c r i p t i o n s for l i k e l i h o o d s c a l c u l a t e d 15 16 17 M o d e l A1: Yij = Mu(i) + e(ij) 18 V a r { e i ;i j ) } = S i g m a ^ 2 19 20 Model A2: Yij = Mu(i) + e(ij) 21 V a r { e i :i j ) } = S i g m a ( i ) ^2 22 23 Model A3: Yij = Mu(i) + e(ij) 2 4 V a r { e i :i j ) } = S i g m a ^ 2 25 Model A3 uses any fixed variance parameters that 26 were specified by the user 27 2 8 M o d e l R: Yi = M u + e(i) 29 Var{e(i)} = Sigma^2 30 31 32 Likelihoods of Interest 33 34 Model Log(likelihood) # Param's AIC 35 A1 -33.142389 7 80.284779 36 A2 -31.811970 12 8 7 . 6 2 3 9 4 0 37 A3 -33.142389 7 80.284779 38 fitted -35.946504 4 79.893008 39 R -80.442086 2 164.884172 40 41 42 Explanation of Tests 43 4 4 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 45 (A2 vs. R) 4 6 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 4 7 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 4 8 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 49 (Note: When rho=0 the results of Test 3 and Test 2 will be the same.) 50 51 T e s t s of I n t e r e s t 52 53 Test -2*log(Likelihood Ratio) Test df p-value 54 55 Test 1 97.2602 10 <.0001 56 Test 2 2.66084 5 0.7521 57 Test 3 2.66084 5 0.7521 58 Test 4 5.60823 3 0.1323 59 60 The p - v a l u e for Test 1 is less t h a n .05. T h e r e a p p e a r s to be a 61 d i f f e r e n c e b e t w e e n response and/or va r i a n c e s among the dose levels 62 It seems appropriate to model the data 63 6 4 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 65 model appears to be appropriate here 66 67 6 8 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 69 to be appropriate here 70 71 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s This document is a draftfor review purposes only and does not constitute Agency policy. H-12 DRAFT--DO NOT CITE OR QUOTE 1 to adequately describe the data 2 3 4 Benchmark Dose Computation 5 6 Specified effect = 1 7 8 Risk Type = Estimated standard deviations from the control mean 9 10 C o n f i d e n c e l e v e l = 0.95 11 12 B M D = 0 . 4 5 2 2 2 1 13 14 15 B M D L = 0 . 2 0 2 6 8 8 16 17 18 H .2 .2 .5 . F ig u re f o r A d d itio n a l M o d el P resented: Pow er, U nrestricted Power Model with 0.95 Confidence Level Mean Response 19 14:15 04/30 2010 20 dose This document is a draftfor review purposes only and does not constitute Agency policy. H-13 DRAFT--DO NOT CITE OR QUOTE 1 H.2.3. Hassoun et al., 2000: T B A R S 2 H.2.3.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model a Degrees of Freedom x 2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes exponential (M 2) 4 0.001 -8.517 1.736E+01 1.223E+01 exponential (M 3) 4 0.001 -8 .5 1 7 1.736E +01 1.223E +01 p o w er hit bound (d = 1) exponential (M 4) 3 0.188 -19.755 2.189E+00 1.151E+00 exponential (M 5) 2 0.240 -19.681 3.470E+00 1.525E+00 Hill b 2 0.272 -19.935 3.292E+00 1.737E+00 linear polynom ial, 5degree power 4 0.002 -9.793 1.444E+01 9.622E +00 4 0.002 -9.793 1.444E+01 9.622E +00 4 0.002 -9.793 1.444E +01 9 .6 2 2 E + 0 0 p o w er b oun d hit (p ow er = 1) a C onstant variance m odel selected (p = 0 .3 3 4 8 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 H .2 .3 .2 . O utputfo r Selected M odel: H ill 6 H a sso u n et al., 2 0 0 0 : T B A R S 7 8 9 Hill Model. (Version: 2.14; Date: 06/26/2008) 10 I n p u t D a t a File: C : \ 5 \ B l o o d \ 1 9 _ H a s _ 2 0 0 0 _ T B A R s L i v _ H i l l C V _ 1 . ( d ) 11 G n u p l o t P l o t t i n g F i l e : C : \ 5 \ B l o o d \ 1 9 _ H a s _ 2 0 0 0 _ T B A R s L i v _ H i l l C V _ 1 . p l t 12 Fri A p r 30 1 4 : 1 6 : 0 2 2 0 1 0 13 14 15 T B A R s , l i v e r o n l y ( T a b l e 2) 16 17 18 T h e f o r m of t h e r e s p o n s e f u n c t i o n is: 19 2 0 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 21 22 23 D e p endent v a r i a b l e = Mea n 24 Independent variable = Dose 25 rho is set to 0 26 Power parameter restricted to be greater than 1 27 A constant v ariance model is fit 28 29 Total number of dose groups = 6 30 Total number of records with missing values = 0 31 M a x i m u m n u m b e r of i t e r a t i o n s = 250 32 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 33 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1e-0 0 8 34 35 36 This document is a draftfor review purposes only and does not constitute Agency policy. H-14 DRAFT--DO NOT CITE OR QUOTE 1 Default Initial Parameter Values 2 alpha 0.178788 3 rho 0 Specified 4 intercept 1.469 5 1.15 6 1.2785 7 5.08547 8 9 10 A s y m p t o t i c C o r r e l a t i o n M a t r i x of P a r a m e t e r E s t i m a t e s 11 12 ( *** T h e m o d e l p a r a m e t e r ( s ) - r h o 13 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , o r h a v e b e e n s p e c i f i e d b y t h e u s e r , 14 a n d do n o t a p p e a r in t h e c o r r e l a t i o n m a t r i x ) 15 16 alpha intercept v n k 17 18 a l p h a 1 2.8e-008 -4.4e-008 4.9e-008 -1.5e-008 19 20 intercept 2.8e-008 1 -0.82 0.48 0.52 21 22 v -4.4e-008 -0.82 1 -0.61 -0.22 23 24 n 4.9e-008 0.48 -0.61 1 0.29 25 26 k -1.5e-008 0.52 -0.22 0.29 1 27 28 29 30 Parameter Estimates 31 32 95.0% Wald Confidence Interval 33 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 34 alpha 0.16017 0.0377523 0.0861764 0.234163 35 intercept 1.46138 0.152797 1.1619 1.76086 36 v 0.963033 0.20228 0.566571 1.3595 37 n 3.44642 2.43468 -1.32547 8.21832 38 k 3.63417 1.02019 1.63464 5.6337 39 40 41 42 Table of Data and Estimated Values of Interest 43 44 Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled R 45 46 47 06 1.47 1.46 0.291 0. 4 0.0466 48 1.938 6 1.55 1.56 0.539 0. 4 -0.0696 49 4.614 6 2.15 2.13 0.363 0. 4 0.12 Lf) O 50 8.147 6 2.28 2.37 0.247 0. 4 51 14.01 6 2.62 2.42 0.517 0. 4 1.25 52 25.34 6 2.29 2.42 0.487 0. 4 -0.803 53 54 55 56 Model Descriptions for likelihoods calculated 57 58 59 Model A1: Yij = Mu(i) + e(ij) 6 0 V a r { e i ;i j ) } = S i g m a ^ 2 61 62 Model A2: Yij = Mu(i) + e(ij) 63 V a r { e i :i j ) } = S i g m a ( i ) ^2 64 65 Model A3: Yij = Mu(i) + e(ij) 6 6 V a r { e i :i j ) } = S i g m a ^ 2 67 Model A3 uses any fixed variance parameters that 68 were specified by the user 69 7 0 M o d e l R: Yi = M u + e(i) 71 Var{e(i)} = Sigma^2 This document is a draftfor review purposes only and does not constitute Agency policy. H-15 DRAFT--DO NOT CITE OR QUOTE 1 2 3 Likelihoods of Interest 4 5 Model Log(likelihood) # Param's AIC 6 A1 16.269770 7 -18.539539 7 A2 19. 1 2 7 8 2 7 12 - 1 4 . 2 5 5 6 5 4 8 A3 16.269770 7 -18.539539 9 fitted 14.967391 5 -19.934782 10 R 2 . 4 4 2 9 4 0 2 - 0 . 8 8 5 8 8 0 11 12 13 E x p l a n a t i o n o f T e s t s 14 15 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 16 (A2 vs. R) 17 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 18 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 19 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 20 (Note When rho=0 the results of Test 3 and Test 2 will be the same 21 22 Tests of Interest 23 24 Test -2*log(Likelihood Ratio) Test df p-vali 25 26 Test 1 33.3698 10 0.000236 27 Test 2 5.71611 5 0.3348 28 Test 3 5.71611 5 0.3348 29 Test 4 2.60476 2 0.2719 30 31 The p - v a l u e for T e s t 1 is less t h a n .05. T h e r e a p p e a r s to be a 32 difference between response and/or variances among the dose levels 33 It s e ems a p p r o p r i a t e to m o d e l the d a t a 34 35 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 36 model appears to be appropriate here 37 38 39 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 40 to be appropriate here 41 4 2 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 43 to a d e q u a t e l y d e s c r i b e the data 44 45 46 Benchmark Dose Computation 47 48 Specified effect = 1 49 50 Risk Type = Estimated standard deviations from the control mean 51 52 Confidence level = 0.95 53 54 BMD = 3.29185 55 56 BMDL = 1.73738 57 This document is a draftfor review purposes only and does not constitute Agency policy. H-16 DRAFT--DO NOT CITE OR QUOTE Mean R esponse 1 H.2.3.3. F ig u r e f o r S e le c te d M o d e l: H ill Hill Model with 0.95 Confidence Level 2 15:22 04/30 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. H-17 DRAFT--DO NOT CITE OR QUOTE 1 H.2.4. Kitchin and Woods, 1979: B a p Hydroxylase Activity 2 H.2.4.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts _______ Model a Degrees of Freedom x2pValue AIC BMD (ng/kg) BMDL (ng/kg) Notes exponential (M 2) 9 <0.0001 452.100 2.960E+02 1.446E+02 exponential (M 3) 9 < 0 .0 0 0 1 4 5 2 .1 0 0 2 .9 6 0 E + 0 2 1 .446E + 02 p o w er hit b ound (d = 1) exponential (M 4) 8 0.002 232.110 3.182E-01 2.373E-01 exponential (M5) b 7 0.015 227.004 9.321E-01 4.900E-01 Hill 8 <.0001 479.250 5.340E+00 4.528E+00 linear polynom ial, 8degree power 9 <.0001 291.380 4.552E-01 3.303E-01 6 <.0001 468.198 1.012E+03 7.899E-01 9 < .0001 2 9 1 .3 8 0 4.5 5 2 E -0 1 3 .303E -01 p o w er b oun d hit (p ow er = 1) a N on-constant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 H .2 .4 .2 . O utputfo r Selected M odel: E xponential (M5) 6 K itch in and W o o d s, 1979: B aP H y d ro x y la se A ctiv ity 7 8 9 Exponential Model. (Version: 1.61; Date: 7/24/2009) 10 I n p u t D a t a File: C : \ 5 \ B l o o d \ 2 7 _ K i t c h i n _ 1 9 7 9 _ H y d r o l a s e _ E x p _ 1 . ( d ) 11 G n u p l o t P l o t t i n g Fil e : 12 Fri A p r 30 14 :1 7 : 2 8 2 0 1 0 13 14 15 K i t c h i n 1 9 7 9 , T b l 3 , B a P h y d r o l a s e a c t i v i t y 16 17 18 T h e f o r m of t h e r e s p o n s e f u n c t i o n b y M o d e l : 19 M o d e l 2: Y dose] = a * exp{sign * b * dose} 20 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 21 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 22 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 23 24 Note: Y[dose] is the m e dian response for exposure = dose; 25 sign = +1 for increasing trend in data; 26 sign = -1 for d e c r e a s i n g trend. 27 2 8 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 2 9 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 30 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 31 32 33 D e p e n d e n t v a r i a b l e = M e a n 34 Independent variable = Dose 35 Data are a s sumed to be distributed: n o r m a l l y 36 Variance Model: exp(lnalpha +rho *ln(Y[dose])) This document is a draftfor review purposes only and does not constitute Agency policy. H-18 DRAFT--DO NOT CITE OR QUOTE 1 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) rho) 2 3 T otal n u m b e r of d o s e g r o u p s = 11 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 MLE solution provided: Exact 10 11 12 I n i t i a l P a r a m e t e r V a l u e s 13 14 Variable Model 5 15 16 lnalpha -3.27793 17 rho 1.92227 18 a 4 . 6 5 5 19 b 0 . 0 0 4 1 2 0 6 20 c 42.6316 21 d 1 22 23 24 25 Parameter Estimates 26 27 Variable Model 5 28 29 lnalpha -2.64071 30 rho 1.94046 31 a 5 . 4 6 2 4 8 32 b 0.0382278 33 c 3 0 . 9 2 0 8 34 d 1.42906 35 36 37 Table of Stats From Input Data 38 39 Dose N Obs Mean Obs Std 40 41 09 4.9 1.11 42 0.0645 4 4.9 1.18 43 0.2023 4 6.7 1.4 44 0.3839 4 7.2 1.8 45 1.613 4 8.3 0.26 46 4.146 4 14 5 47 11.59 4 59 6.8 48 30.26 4 96 46 49 90.9 4 155 16.4 50 218 4 182 26 51 863.2 4 189 26 52 53 54 Estimated Values of Interest 55 56 Dose Est Mean Est Std Scaled Residual 57 58 0 5.462 1.387 -1.217 59 0.0645 5.493 1.394 -0.8507 60 0.2023 5.619 1.425 1.516 61 0.3839 5.854 1.483 1.815 62 1.613 8.483 2.126 -0.1723 63 4.146 16.8 4.125 -1.358 64 11.59 49.32 11.73 1.65 65 30.26 121.2 28.06 -1.796 66 90.9 168.5 38.62 -0.6975 67 218 168.9 38.72 0.6765 68 863.2 168.9 38.72 1.038 69 70 71 This document is a draftfor review purposes only and does not constitute Agency policy. H-19 DRAFT--DO NOT CITE OR QUOTE 1 Other models for which likelihoods are calculated: 2 3 Model A1: Yij = Mu(i) + e(ij) 4 V a r { e i ;ij)} = S i g m a ^ 2 5 6 Model A2: Yij = Mu(i) + e(ij) 7 V a r { e i :ij)} = S i g m a ( i ) ^ 2 8 9 Model A3: Yij = Mu(i) + e(ij) 10 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 11 12 M o d e l R: Yij = M u + e(i) 13 V a r { e i ij)} = S i g m a ^ 2 14 15 16 L i k e l i h o o d s of I n t e r e s t 17 18 Model Log(likelihood) DF AIC 19 20 A1 -158.1306 12 3 4 0 . 2 6 1 3 21 A2 -84.80028 22 2 1 3 . 6 0 0 6 22 A3 -98.82189 13 223.6438 23 R -234.6252 2 473.2504 24 5 -107.5022 6 227.0044 25 26 27 Additive constant for all log-likelihoods = -45.03. This constant added to the 28 above values gives the log-likelihood including the term that does not 29 depend on the model parameters. 30 31 32 Explanation of Tests 33 34 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 35 T e s t 2: A r e H o m o g e n e o u s ? (A2 vs. A1) 36 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 37 38 T e s t 7a: D o e s M o d e l 5 f i t t h e d a t a ? (A3 v s 5) 39 40 41 Tests of Interest 42 43 Test -2*log(Likelihood Ratio) D. F. p-value 44 45 Test 1 299.6 20 < 0.0001 46 Test 2 146.7 10 < 0.0001 47 Test 3 28.04 9 0.0009381 48 Test 7a 17.36 7 0.01521 49 50 51 The p - v a l u e for T e s t 1 is less t h a n .05. T h e r e a p p e a r s to be a 52 difference between response and/or variances among the dose 53 levels, it s e ems a p p r o p r i a t e to m o d e l the data. 54 55 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 56 variance model appears to be appropriate. 57 58 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t to 59 consider a different variance model. 60 61 T h e p - v a l u e f o r T e s t 7a is l e s s t h a n .1. M o d e l 5 m a y n o t a d e q u a t e l y 62 describe the data; you may want to consider another model. 63 64 65 Benchmark Dose Computations: 66 67 Specified Effect = 1.000000 68 69 Risk Type = Estimated standard deviations from control 70 71 C o n f i d e n c e Level = 0.950000 This document is a draftfor review purposes only and does not constitute Agency policy. H-20 DRAFT--DO NOT CITE OR QUOTE 1 2 BMD 0.9321 3 4 BMDL = 0.490004 5 6 7 H .2 .4 .3 . F igure fo r Selected M odel: E xponential (M5) Exponential Model 5 with 0.95 Confidence Level Mean R esponse 8 14:17 04/30 2010 9 dose This document is a draftfor review purposes only and does not constitute Agency policy. H-21 DRAFT--DO NOT CITE OR QUOTE 1 H.2.5. National Toxicology Program, 2006: Liver E R O D 53 W e e k s 2 H.2.5.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts _____________ Model a Degrees of Freedom x 2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes exponential (M 2) 4 <0.0001 648.094 2.011E+01 1.464E+01 exponential (M 3) 4 < 0 .0 0 0 1 6 4 8 .0 9 4 2.0 1 1 E + 0 1 1.464E +01 p o w er hit b ound (d = 1) exponential (M 4) 3 0.015 521.251 1.430E-02 9.808E-03 exponential (M 5) Hill b 2 2 0.354 514.812 7.656E-02 3.202E-02 0.760 513.286 1.853E-01 9.351E-02 linear polynom ial, 5degree power 4 <.0001 639.841 1.034E+01 6.557E-03 1 <.0001 14.000 error error 4 < .0 0 0 1 5 9 2 .8 8 9 2 .2 5 4 E -0 2 1.527E -02 p o w er b oun d hit (p ow er = 1) a N on-con stant variance m odel selected (p = < .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 H .2 .5 .2 . O utputfo r Selected M odel: H ill 6 N ation al T o x ico lo g y Program , 2006: L iver E R O D 53 W eek s 7 8 9 Hill Model. (Version: 2.14; Date: 06/26/2008) 10 I n p u t D a t a File: C : \ 5 \ B l o o d \ 4 6 _ N T P _ 2 0 0 6 _ E R O D l i v 5 3 _ H i l l _ 1 . ( d ) 11 G n u p l o t P l o t t i n g F i l e : C : \ 5 \ B l o o d \ 4 6 _ N T P _ 2 0 0 6 _ E R O D l i v 5 3 _ H i l l _ 1 . p l t 12 S u n M a y 02 1 5 : 3 4 : 2 1 2 0 1 0 13 14 15 0 16 17 18 T h e f o r m of t h e r e s p o n s e f u n c t i o n is: 19 2 0 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 21 22 23 D e p endent v a r i a b l e = Mea n 24 Independent variable = Dose 25 Power parameter restricted to be greater than 1 26 The v ariance is to be mode l e d as Var(i) = exp(lalpha + rho * ln(mean(i))) 27 28 Total number of dose groups = 6 29 Total number of records with missing values = 0 30 M a x i m u m number of iterations = 250 31 R e l a t i v e F u n c t i o n C o n v e r g e n c e has b e e n set to: 1e-0 0 8 32 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 33 34 35 36 Default Initial Parameter Values This document is a draftfor review purposes only and does not constitute Agency policy. H-22 DRAFT--DO NOT CITE OR QUOTE 1 lalpha 11.0197 2 rho 0 3 intercept 30.215 4 1841.26 5 7.0105 6 6.95814 7 8 9 Asymptotic Correlation Matrix of Parameter Estimates 10 11 lalpha rho intercept v n k 12 CO CO r CO oo CO r o CO <-- 1 o CO <-- 1 o 13 l a l p h a 1 -0.97 0.065 -0.025 0.046 14 15 rho -0.97 1 0.17 -0.093 0.025 -0.048 16 17 i n t e r c e p t 0.17 1 -0.022 0.011 0.00084 18 19 v 0.065 -0.093 -0.022 1 0.87 20 21 n -0.025 0.025 0.011 1 -0.83 22 23 k 0.046 -0.048 0.00084 0.87 1 24 25 26 27 Parameter Estimates 28 29 95.0% Wald Confidence Interval 30 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 31 lalpha -4.47504 0.923978 -6.286 -2.66407 32 rho 2.12799 0.137849 1.85781 2.39817 33 i n t e r c e p t 30.2685 1.41935 27.4866 33.0504 34 v 1813.88 100.554 1616.8 2010.96 35 n 2.02516 0.29717 1.44272 2.6076 36 k 3.78554 0.349266 3.101 4.47009 37 38 39 40 Table of Data and Estimated Values of Interest 41 42 Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 43 44 45 08 30.2 30.3 4.5 4.02 -0.0377 46 2.458 8 569 564 69.6 90.3 0.17 47 5.533 8 1.28e+003 1.27e+003 270 214 0.137 48 9.543 8 1.55e+003 1.6e+003 318 274 -0.529 49 16.18 8 1.73e+003 1.75e+003 304 302 -0.248 50 29.04 8 1.87e+003 1.82e+003 309 313 0.507 51 52 53 54 Model Descriptions for likelihoods calculated 55 56 57 Model A1: Yij = Mu(i) + e(ij ) 58 Var{ei;ij)} = Sigma^2 59 60 Model A2: Yij = Mu(i) + e(ij ) 61 Var{ei;ij)} = Sigma(i) ^2 62 63 Model A3: Yij = Mu(i) + e(ij ) 64 Var{ei;ij)} = e x p ( l a l p h a + rho*ln(Mu(i))) 65 Model A3 uses any fixed variance parameters that 66 were specified by the user 67 6 8 M o d e l R: Yi = Mu + e(i ) 69 Var{e(i)} = Sigma^2 70 71 This document is a draftfor review purposes only and does not constitute Agency policy. H-23 DRAFT--DO NOT CITE OR QUOTE 1 Likelihoods of Interest 2 3 Model Log(likelihood) # Param's AIC 4 A1 -285.269096 7 584.538193 5 A2 -249.237836 12 5 2 2 . 4 7 5 6 7 1 6 A3 -250.368300 8 516.736600 7 fitted -250.643212 6 513.286424 8 R -338.451300 2 680.902600 9 10 11 E x p l a n a t i o n o f T e s t s 12 13 T e s t 1: D o r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 14 ( A 2 vs. R) 15 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? ( A 1 vs A2) 16 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 17 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 18 (Note W h e n r h o = 0 the r e s u l t s of T e s t 3 a n d T e s t 2 w i l l be t h e s a m e 19 20 Tests of Interest 21 22 Test -2*log(Likelihood Ratio) Test df p-value 23 24 Test 1 178.427 10 <.0001 25 Test 2 72.0625 5 <.0001 26 Test 3 2.26093 4 0.6879 27 Test 4 0.549824 2 0.7596 28 29 The p - v a l u e for Test 1 is less t h a n .05. T h e r e a p p e a r s to be a 30 difference between response and/or variances among the dose levels 31 It s e ems a p p r o p r i a t e to m o d e l the d a t a 32 33 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 34 model appears to be appropriate 35 36 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 37 to be appropriate here 38 39 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 40 to adequately describe the data 41 42 43 B e n c h m a r k Dose C o m p u t a t i o n 44 45 Specified effect = 1 46 47 Risk Type = Estimated standard deviations from the control mean 48 49 Confidence level = 0.95 50 51 BMD = 0.185269 52 53 BMDL = 0.0935065 54 This document is a draftfor review purposes only and does not constitute Agency policy. H-24 DRAFT--DO NOT CITE OR QUOTE Mean R esponse 1 H.2.5.3. F ig u r e f o r S e le c te d M o d e l: H ill Hill Model with 0.95 Confidence Level 2 15:34 05/02 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. H-25 DRAFT--DO NOT CITE OR QUOTE 1 H.2.6. National Toxicology Program, 2006: L u n g Erod 53 W e e k s 2 H.2.6.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts ____________ Model a Degrees of Freedom x 2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes exponential (M 2) 4 <0.0001 314.332 3.281E+01 2.047E+01 exponential (M 3) exponential (M4) b exponential (M 5) 4 < 0 .0 0 0 1 55 5 .0 6 1 5 .2 1 0 E + 0 0 8 .194E -01 p o w er hit b oun d (d = 1) 3 0.302 255.955 9.586E-02 5.907E-02 2 0.276 256.882 1.044E+00 6.588E-02 Hill 2 0.275 256.882 1.903E+00 3.469E-01 linear 4 <.0001 313.237 2.662E+01 1.251E+01 polynom ial, 5degree 5 <.0001 330.180 error 2 .7 1 8 E + 0 1 power 4 < .0001 3 1 3 .2 3 7 2 .6 6 2 E + 0 1 1.251E +01 p o w er b ound hit (p ow er = 1) power, unrestricted c 3 0.032 261.083 1.875E-07 1.875E-07 unrestricted (power = 0.18) a N on-con stant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 H .2 .6 .2 . O utputfo r Selected M odel: E xponential (M4) 6 N ation al T o x ico lo g y Program , 2006: L u ng E R O D 53 W eek s 7 8 9 Exponential Model. (Version: 1.61; Date: 7/24/2009) 10 I n p u t D a t a File: C : \ 5 \ B l o o d \ 5 2 _ N T P _ 2 0 0 6 _ L u n g E R O D 5 3 _ E x p _ 1 . ( d ) 11 G n u p l o t P l o t t i n g Fil e : 12 Fri A p r 30 1 4 : 2 0 : 2 7 2 0 1 0 13 14 15 T b l 12, W e e k 53, L u n g M i c r o s o m e s E R O D 16 17 18 T h e f o r m of t h e r e s p o n s e f u n c t i o n b y M o d e l : 19 M o d e l 2: Y dose] = a * exp{sign * b * dose} 20 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 21 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 22 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 23 24 Note: Y[dose] is the m e dian response for exposure = dose; 25 sign = +1 for increasing trend in data; 26 sign = -1 for d e c r e a s i n g trend. 27 2 8 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 2 9 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 30 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 31 This document is a draftfor review purposes only and does not constitute Agency policy. H-26 DRAFT--DO NOT CITE OR QUOTE 1 2 Dependent variable = Mean 3 Independent variable = Dose 4 Data are assumed to be distributed: normally 5 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 6 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) rho) 7 8 Total number of dose groups = 6 9 Total number of records with missing values = 0 10 M a x i m u m n u m b e r of i t e r a t i o n s = 250 11 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 12 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1 e - 0 0 8 13 14 M L E s o l u t i o n p r o v i d e d : E x a c t 15 16 17 I n i t i a l P a r a m e t e r V a l u e s 18 19 Variable Model 4 20 21 lnalpha -0.80064 22 rho 1.47683 23 a 2.86045 24 b 0.134268 25 c 16.0581 26 d 1 27 28 29 30 Parameter Estimates 31 32 Variable Model 4 33 34 lnalpha -1.14455 35 rho 1.63458 36 a 3.06102 37 b 0.371249 38 c 14.1551 39 d 1 40 41 42 Table of Stats From Input Data 43 44 Dose N Obs Mean Obs Std Dev 45 46 08 3.011 1.584 47 2.458 8 27.15 5.269 48 5.533 8 42.85 11.15 49 9.543 8 36.57 12.99 50 16.18 8 43.75 18.55 51 29.04 8 43.71 6.322 52 53 54 Estimated Values of Interest 55 56 Dose Est Mean Est Std Scaled Residual 57 58 0 3. 061 1.408 -0.1005 59 2.458 27 .16 8.383 -0.003073 60 5.533 38 .17 11.07 1.196 61 9.543 42 .16 12.01 -1.318 62 16.18 43 .23 12.26 0.1191 63 29.04 43 .33 12.28 0.08864 64 65 66 67 Other models for which likelihoods are calculated: 68 69 Model A1: Yij = M u ( i ) + e (ij) 70 Var{e( ij)} = Sigma^2 71 This document is a draftfor review purposes only and does not constitute Agency policy. H-27 DRAFT--DO NOT CITE OR QUOTE 1 Model A2: Yij Mu(i) + e(ij) 2 Var{e(ij)} Sigma(i)^2 3 4 Model A3: Yij Mu(i) + e(ij) 5 Var{e(ij)} exp(lalpha + log(mean(i) rho) 6 7 M o d e l R: Yij M u + e(i) 8 Var{e(ij)} Sigma^2 9 10 11 L i k e l i h o o d s o f I n t e r e s t 12 13 Model Log(likelihood) DF AIC 14 15 A1 -135.2677 7 284.5353 16 A2 -115.6885 12 25 5 . 3 7 7 1 17 A3 -121.1517 8 258.3034 18 R -162.0902 2 328.1805 19 4 -122.9773 5 255.9546 20 21 22 Additive constant for all log-likelihoods = -44.11. This constant added to the 23 above values gives the log - l i k e l i h o o d i n c luding the ter m that does not 24 depend on the model parameters. 25 26 27 Explanation of Tests 28 2 9 T e s t 1 D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 30 T e s t 2 A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 31 T e s t 3 A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 32 33 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 34 35 36 Tests of Interest 37 38 Test -2*log(Likelihood Ratio) D. F. p-value 39 40 Test 1 92.8 10 < 0.0001 41 Test 2 39.16 5 < 0.0001 42 Test 3 10.93 4 0.0274 43 Test 6a 3.651 3 0.3017 44 45 46 The p - v a l u e for T e s t 1 is less t h a n .05. T h e r e a p p e a r s to be a 47 difference between response and/or variances among the dose 48 levels, it seems appropriate to model the data. 49 50 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 51 v a r i a n c e m o d e l a p p e a r s to be a p p r o p r i a t e . 52 53 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t to 54 consider a different variance model. 55 56 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 57 to a dequately describe the data. 58 59 60 Benchmark Dose Computations: 61 62 Specified Effect = 1.000000 63 64 Risk Type = Estimated standard deviations from control 65 66 Confidence Level = 0.950000 67 68 BMD = 0.09586 69 70 BMDL = 0.0590734 This document is a draftfor review purposes only and does not constitute Agency policy. H-28 DRAFT--DO NOT CITE OR QUOTE Mean R esponse 1 H.2.6.3. F ig u r e f o r S e le c te d M o d e l: E x p o n e n tia l (M 4) Exponential Model 4 with 0.95 Confidence Level 2 14:20 04/30 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. H-29 DRAFT--DO NOT CITE OR QUOTE 1 H.2.6.4. O u tp u tf o r A d d itio n a l M o d e l P r e s e n te d : P o w e r, U n r e s tr ic te d 2 National Toxicology Program, 2006: Lung EROD 53 Weeks 3 4 5 Power Model. (Version: 2.15; Date: 04/07/2008) 6 Input Data File: C:\5\Blood\52_NTP_2006_LungEROD53_Pwr_U_1.(d) 7 Gnuplot Plotting File: C:\5\Blood\52_NTP_2006_LungEROD53_Pwr_U_1.plt 8 Fri_Apr 30 14:20:33 2010 9 10 11 T b l 12, W e e k 53, L u n g M i c r o s o m e s E R O D 12 13 14 T h e f o r m of t h e r e s p o n s e f u n c t i o n is: 15 16 Y [ d o s e ] = c o n t r o l + s l o p e * d o s e Ap o w e r 17 18 19 D e p e n d e n t v a r i a b l e = M e a n 20 Independent variable = Dose 21 The p o w e r is not restr i c t e d 22 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) * rho) 23 24 Total number of dose groups = 6 25 Total number of records with missing values = 0 26 Maximum number of iterations = 250 27 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 28 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 29 30 31 32 Default Initial Parameter Values 33 lalpha = 4.76968 34 rho = 0 35 control = 3.011 36 slope = 23.2411 37 power = 0.187468 38 39 40 Asymptotic Correlation Matrix of Parameter Estimates 41 42 lalpha rho control slope power 43 44 lalpha 1 -0.96 -0.49 0.1 -0.045 45 46 rho -0.96 1 0.45 -0.13 0.05 47 48 control -0.49 0.45 1 -0.14 0.048 49 50 slope 0.1 -0.13 -0.14 1 -0.94 51 52 power -0.045 0.05 0.048 -0.94 1 53 54 55 56 Parameter Estimates 57 58 95.0% Wald Confidence Interval 59 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limi 60 lalpha -1.02668 0.818488 -2.63088 0.577531 61 rho 1.63033 0.24056 1.15884 2.10182 62 control 3.01543 0.519355 1.99751 4.03335 63 slope 23.8167 3.70401 16.5569 31.0764 64 power 0.179731 0.0639681 0.054356 0.305106 65 66 67 68 Table of Data and Estimated Values of Interest 69 70 Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res. This document is a draftfor review purposes only and does not constitute Agency policy. H-30 DRAFT--DO NOT CITE OR QUOTE 1 2 3 08 3.01 3.02 1.58 1.47 -0.00851 4 2.458 8 27.1 31 5.27 9.84 -1.11 5 5.533 8 42.8 35.4 11.2 11 1.92 6 9.543 8 36.6 38.7 13 11.8 -0.52 7 16.18 8 43.7 42.3 18.5 12.7 0.323 8 29.04 8 43.7 46.6 6.32 13.7 -0.605 9 10 11 12 M o d e l D e s c r i p t i o n s for l i k e l i h o o d s c a l c u l a t e d 13 14 15 M o d e l A1: Yij = Mu(i) + e(ij ) 16 V a r { e ( ij)} = S i g m a ^ 2 17 18 M o d e l A2: Yij = Mu(i) + e(ij ) 19 V a r { e ( ij)} = Si g m a ( i ) ^2 20 21 Model A3: Yij = Mu(i) + e(ij ) 22 Var{e( ij)} = exp(lalpha + rho*ln(Mu(i))) 23 Model A3 uses any fixed v a r i a n c e p a r a m e t e r s that 24 were specified by the user 25 2 6 M o d e l R: Yi = Mu + e(i ) 2 7 V a r { e (i)} = S i g m a ^ 2 28 29 30 Likelihoods of Interest 31 32 Model Log(likelihood) # Param's AIC 33 A1 -135.267662 7 284.535325 34 A2 -115.688533 12 2 5 5 . 3 7 7 0 6 7 35 A3 -121.151707 8 258.303413 36 fitted -125.541690 5 261.083380 37 R -162.090242 2 328.180484 38 39 40 Explanation of Tests 41 4 2 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 43 (A2 vs. R) 4 4 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 45 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 4 6 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 47 (Note: When rho=0 the results of Test 3 and Test 2 will be the same.) 48 49 Tests of Interest 50 51 Test -2*log(Likelihood Ratio) Test df p-value 52 53 Test 1 92.8034 10 <.0001 54 Test 2 39.1583 5 <.0001 55 Test 3 10.9263 4 0.0274 56 Test 4 8.77997 3 0.03236 57 58 The p - v a l u e for Test 1 is less t h a n .05. T h e r e a p p e a r s to be a 59 difference between response and/or variances among the dose levels 60 It seems appropriate to model the data 61 6 2 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 63 model appears to be a p p r o priate 64 65 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t t o c o n s i d e r a 66 different variance model 67 6 8 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 69 model 70 71 This document is a draftfor review purposes only and does not constitute Agency policy. H-31 DRAFT--DO NOT CITE OR QUOTE 1 Benchmark Dose Computation 2 3 Specified effect = 1 4 5 Risk Type = Estimated standard deviations from the control mean 6 7 Confidence level = 0.95 8 9 BMD = 1.8745e-007 10 11 12 B M D L = 1 . 8 7 4 5 e - 0 0 7 13 14 15 H . 2 .6 . 5 . F ig u r e f o r A d d itio n a l M o d e l P resen ted : P ow er, U n restricted Power Model with 0.95 Confidence Level Mean Response 16 14:20 04/30 2010 17 dose This document is a draftfor review purposes only and does not constitute Agency policy. H-32 DRAFT--DO NOT CITE OR QUOTE 1 H .2 .7 . N a tio n a l T o x ico lo g y P ro g r a m , 2 0 0 6 : L a b elin g In d e x 31 W ee k s 2 H . 2 . 7 . 1 . S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts ______________________ Model a Degrees of Freedom x 2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes exponential (M 2) 4 0.000 46.547 8.660E+00 6.926E+00 exponential (M 3) 4 0.000 4 6 .5 4 7 8 .6 6 0 E + 0 0 6 .9 2 6 E + 0 0 p o w er hit boun d (d = 1) exponential (M 4) 3 <0.0001 50.958 3.151E+00 1.865E+00 exponential (M 5) 3 < 0 .0 0 0 1 50 .9 5 8 3 .1 5 1 E + 0 0 1.864E + 00 p o w er hit boun d (d = 1) H ill 3 <.0001 50.963 3.145E+00 error n lo w er b ound hit (n = 1) linear polynomial, 5degree b power 4 0.000 48.958 3.151E+00 1.865E+00 3 0.000 46.230 7.607E+00 3.125E+00 4 0.000 4 8 .9 5 8 3 .1 5 1 E + 0 0 1.865E + 00 p o w er bound hit (p ow er = 1) a N on-con stant variance m odel selected (p = < .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 H .2 .7 .2 . O utputfo r Selected M odel: Polynom ial, 5-degree 6 N ation al T o x ico lo g y Program , 2006: L ab elin g In dex 31 W eek s 7 8 9 Polynomial Model. (Version: 2.13; Date: 04/08/2008) 10 I n p u t D a t a File: C : \ 5 \ B l o o d \ 3 8 _ N T P _ 2 0 0 6 _ H e p I n d e x _ P o l y 5 _ 1 . ( d ) 11 G n u p l o t P l o t t i n g F i l e : C : \ 5 \ B l o o d \ 3 8 _ N T P _ 2 0 0 6 _ H e p I n d e x _ P o l y 5 _ 1 . p l t 12 F r i _ A p r 30 1 4 : 2 1 : 1 6 2 0 1 0 13 14 15 T b l 11, 3 1 w k , H e p C e l l P r o l i f e r a t i o n L a b e l i n g I n d e x 16 17 18 T h e f o r m of t h e r e s p o n s e f u n c t i o n is: 19 2 0 Y [ d o s e ] = b e t a _ 0 + b e t a _ 1 * d o s e + b e t a _ 2 * d o s e ^ 2 + ... 21 22 23 D e p endent v a r i a b l e = Mea n 24 Independent variable = Dose 25 The polynomial coefficients are restricted to be positive 26 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) * rho) 27 28 Total number of dose groups = 6 29 Total number of records with missing values = 0 30 M a x i m u m number of iterations = 250 31 R e l a t i v e F u n c t i o n C o n v e r g e n c e has b e e n set to: 1e-0 0 8 32 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 33 34 35 36 Default Initial Parameter Values This document is a draftfor review purposes only and does not constitute Agency policy. H-33 DRAFT--DO NOT CITE OR QUOTE 1 lalpha = 0.708431 2 rho = 0 3 beta 0 = 0.327 4 beta 1 = 0 5 beta 2 = 0 6 beta 3 = 0 7 beta 4 = 0 8 beta 5 = 0 9 10 11 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 12 13 ( *** The model parameter(s) -beta 2 -beta 3 -beta 4 14 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , or h a v e b e e n s p e c i f i e d b y t h e user, 15 a n d d o n o t a p p e a r i n t h e c o r r e l a t i o n m a t r i x ) 16 17 lalpha rho beta 0 beta 1 beta 5 18 19 l a l p h a 1 -0.086 0.012 -0.032 0.043 20 21 rho -0.086 1 -0.0027 -0.011 0.076 22 23 beta 0 0.012 -0.0027 1 -0.6 0.23 24 CO Lf) o CO Lf) o 25 beta 1 -0.032 -0.011 -0.6 1 26 27 beta 5 0.043 0.076 0.23 1 28 29 30 31 P a r a m e t e r E s t i m a t e s 32 33 95.0% W a l d C o n f i d e n c e I n t e r v a l 34 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 35 lalpha -0.501559 0.185039 -0.864229 -0.138889 36 rho 1.90452 0.272948 1.36955 2.43948 37 beta 0 0.500197 0.102837 0.298641 0.701753 38 beta 1 0.0525247 0.0192967 0.0147038 0.0903456 39 beta 2 8.00068e-025 NA 40 beta 3 0 NA 41 beta 4 0 NA 42 beta 5 1.08658e-007 6.10451e-008 -1.09879e-008 2.28305e-007 43 44 NA - Indicates that this parameter has hit a bound 45 implied b y some i nequality constraint and thus 46 has no standard error. 47 48 49 50 Table of Data and Estimated Values of Interest 51 52 Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 53 54 55 09 0.327 0.5 0.189 0.402 -1.29 56 2.331 10 0.852 0.623 0.651 0.496 1.46 57 5.315 10 0.956 0.78 0.737 0.614 0.907 58 9.207 10 0.792 0.991 0.462 0.772 -0.816 59 15.66 10 1.33 1.42 1.12 1.09 -0.266 60 28.13 10 3.85 3.89 3.08 2.84 -0.0523 61 62 63 64 Model Descriptions for likelihoods calculated 65 66 67 Model A1: Yij = M u ( i ) + e( 6 8 V a r { e i ;ij)} = S i g m a ^ 2 69 70 Model A2: Yij = M u ( i ) + e( 71 V a r { e i :ij)} = S i g m a ( i ) ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. H-34 DRAFT--DO NOT CITE OR QUOTE 1 2 Model A3: Yij = Mu(i) + e(ij) 3 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 4 Model A3 uses any fixed variance parameters that 5 were specified by the user 6 7 M o d e l R: Yi = M u + e(i) 8 Var{e(i)} = Sigma^2 9 10 11 L i k e l i h o o d s o f I n t e r e s t 12 13 Model Log(likelihood) # Param's AIC 14 A1 -47.234977 7 108.469953 15 A 2 - 8 . 6 7 9 2 5 6 12 4 1 . 3 5 8 5 1 2 16 A3 -8.980651 8 33.961301 17 fitted -18.115050 5 46.230101 18 R -63.448285 2 130.896571 19 20 21 E x p l a n a t i o n of Tests 22 23 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 2 4 (A2 vs. R) 25 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 2 6 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 2 7 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 28 (Note When rho=0 the results of Test 3 and Test 2 will be the same 29 30 Tests of Interest 31 32 Test -2*log(Likelihood Ratio) Test df 33 P -valu 34 Test 1 10 9.538 10 <. 0 0 0 1 35 Test 2 77 .1114 5 <. 0 0 0 1 36 Test 3 0. 6 0 2 7 9 4 0. 9 6 2 8 37 Test 4 18 .2688 3 0.0003871 38 39 p - v a l u e for T e s t 1 is less t h a n .05. T h e r e a p pea rs to be a 40 difference between response and/or variances among the dose levels 41 It seems app r o p r i a t e to model the data 42 43 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 44 model appears to be appropriate 45 4 6 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 47 to be appropriate here 48 4 9 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 50 model 51 52 53 B e n c h m a r k Dose C o m p u t a t i o n 54 55 Sp e c i f i e d effect = 1 56 57 Risk Type = Estimated standard deviations from the control mean 58 59 Confidence level = 0.95 60 61 BMD = 7.6073 62 63 64 BMDL = 3.12526 65 This document is a draftfor review purposes only and does not constitute Agency policy. H-35 DRAFT--DO NOT CITE OR QUOTE Mean R esponse 1 H.2.7.3. F ig u r e f o r S e le c te d M o d e l: P o ly n o m ia l, 5 -d e g re e Polynomial Model with 0.95 Confidence Level 2 14:21 04/30 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. H-36 DRAFT--DO NOT CITE OR QUOTE 1 H .2 .8 . V a n d en H e u v e l et a l., 1994: H e p a tic C Y P 1 A 1 M r n a E x p re ssio n 2 H . 2 . 8 . 1 . S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts ________________________ Model a Degrees of Freedom x 2p Value AIC BMD (ng/kg) BMDL (ng/kg) Notes exponential (M 2) 5 <0.0001 1147.626 1.769E+01 1.257E+01 exponential (M 3) 4 < 0 .0 0 0 1 1 1 49.626 1.769E +01 1.257E +01 p o w er hit bound (d = 1) exponential (M 4) 4 <0.0001 666.337 6.104E-02 2.871E-02 exponential (M 5) 3 <0.0001 635.591 1.252E+00 9.089E-01 Hillb 3 <.0001 664.418 2.429E-01 1.679E-01 linear polynom ial, 6degree power 5 <.0001 673.777 4.546E-02 2.487E-02 6 <.0001 1213.329 error 1.301E+03 4 <.0001 673.418 6.269E-02 3.196E-02 a N on-con stant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 H .2 .8 .2 . O utputfo r Selected M odel: H ill 6 V a n d en H eu v e l et al., 1994: H e p a tic C Y P 1 A 1 m R N A E x p ressio n 7 8 9 Hill Model. (Version: 2.14; Date: 06/26/2008) 10 I n p u t D a t a File: C : \ U s e p a \ B M D S 2 1 \ D a t a \ h i l _ V a n d e n _ m R N A _ S e t t i n g . ( d ) 11 G n u p l o t P l o t t i n g F i l e : C : \ U s e p a \ B M D S 2 1 \ D a t a \ h i l _ V a n d e n _ m R N A _ S e t t i n g . p l t 12 T u e M a y 18 0 5 : 2 4 : 4 8 2 0 1 0 13 14 15 B M D S M o d e l R u n 16 17 18 T h e f o r m of t h e r e s p o n s e f u n c t i o n is: 19 2 0 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 21 22 23 D e p endent v a r i a b l e = m R N A mean 24 Independent variable = blood conc 25 Power pa r a m e t e r is not restricted 26 The v ariance is to be mode l e d as Var(i) = exp(lalpha + rho * ln(mean(i))) 27 28 Total number of dose groups = 7 29 Total number of records with missing values = 0 30 M a x i m u m number of iterations = 250 31 R e l a t i v e F u n c t i o n C o n v e r g e n c e has b e e n set to: 1e-0 0 8 32 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 33 34 35 36 User Inputs Initial Parameter Values This document is a draftfor review purposes only and does not constitute Agency policy. H-37 DRAFT--DO NOT CITE OR QUOTE 1 lalpha = 1 2 rho = 1.9 3 intercept = 6 4 v= 36000 5 n= 1 6 k= 1000 7 8 9 Asymptotic Correlation Matrix of Parameter Estimates 10 11 lalpha rho intercept v n k 12 13 l a l p h a 1 -0.89 -0.43 0.27 0.68 -0.18 14 15 rho -0.89 1 0.31 -0.42 -0.72 0.22 16 17 i n t e r c e p t -0.43 0.31 1 -0.093 0.14 -0.04 18 19 v 0.27 -0.42 -0.093 1 0.075 0.7 20 21 n 0.68 -0.72 0.14 0.075 1 -0.52 22 23 k -0.18 0.22 -0.04 0.7 -0.52 1 24 25 26 27 Parameter Estimates 28 29 95.0% Wald Confidence Interval 30 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 31 lalpha -0.191631 0.711681 -1.5865 1.20324 32 rho 2.0275 0.132551 1.76771 2.28729 33 i n t e r c e p t 5.416 1.16292 3.13672 7.69529 34 v 41657.2 16561.5 9197.25 74117.2 35 n 1.29154 0.100513 1.09454 1.48854 36 k 97.8648 41.0376 17.4325 178.297 37 38 39 40 Table of Data and Estimated Values of Interest 41 42 Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res. 43 44 45 0 13 5.4 5.42 3.61 5.04 - 0.0115 46 ). 0 1 1 3 5 7.2 5.76 5.59 5.36 0.602 4 7 ).106 12 14.8 11.6 14.9 10.9 1.03 48 ). 8 828 7 12.8 100 4.5 97.2 -2.38 49 6.46 7 536 1.21e+003 320 1 .2 2 e + 0 0 3 -1.48 50 8.32 11 1. 8 e + 0 0 4 1.19e+004 1 .52e+004 1..2 4 e + 004 1.62 51 Lf) CO 5 3.67e+004 3.64e+004 2 .21e+004 3..8 2 e + 004 0.0199 52 53 54 55 Model Descr i p t i o n s for l i k e l ihoods calcu l a t e d 56 57 58 Model A1: Yij = Mu(i) + e(ij) 59 V ar{e (ij)} = Sigma^2 60 61 Model A2: Yij = Mu(i) + e(ij) 62 V ar{e (ij)} = Sigma(i)^2 63 64 Model A3: Yij = Mu(i) + e(ij) 65 V ar{e (ij)} = exp(l a l p h a + rho*ln(Mu(i))) 66 Model A3 uses any fixed variance parameters that 67 were specified by the user 68 6 9 M o d e l R: Yi M u + e(i) 70 Var{e(i)} Sigma^2 71 This document is a draftfor review purposes only and does not constitute Agency policy. H-38 DRAFT--DO NOT CITE OR QUOTE 1 2 Likelihoods of Interest 3 4 Model Log(likelihood) # Param's AIC 5 A1 -572.470944 8 1160.941889 6 A2 -290.799287 14 6 0 9 . 5 9 8 5 7 5 7 A3 -293.809342 9 605.618684 8 fitted -326.209186 6 664.418372 9 R -603.663396 2 1211.326792 10 11 12 E x p l a n a t i o n of T e s t s 13 14 T e s t 1: D o r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 15 ( A 2 vs. R) 16 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? ( A 1 vs A2) 17 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 18 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 19 (Note W h e n r h o = 0 the r e s u l t s of T e s t 3 a n d T e s t 2 w i l l be t h e s a m e 20 21 Tests of Interest 22 23 Test -2*log(Likelihood Ratio) Test df p-value 24 25 Test 1 625.728 12 <.0001 26 Test 2 563.343 6 <.0001 27 Test 3 6.02011 5 0.3043 28 Test 4 64.7997 3 <.0001 29 30 The p - v a l u e for Test 1 is less t h a n .05. T h e r e a p p e a r s to be a 31 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g the d o s e leve l s 32 It seems appropriate to model the data 33 34 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 35 model appears to be a p p r o priate 36 37 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 38 to be appropriate here 39 4 0 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 41 model 42 43 44 Benchmark Dose Computation 45 46 Specified effect = 24 47 48 Risk Type = Point risk 49 50 Confidence l e v e l = 0.95 51 52 BMD = 0.249203 53 54 BMDL = 0.167897 55 This document is a draftfor review purposes only and does not constitute Agency policy. H-39 DRAFT--DO NOT CITE OR QUOTE 1 H.2.8.3. F ig u r e f o r S e le c te d M o d e l: H ill Hill Model with 0.95 Confidence Level Mean R esponse 2 05:24 05/18 2010 3 dose This document is a draftfor review purposes only and does not constitute Agency policy. H-40 DRAFT--DO NOT CITE OR QUOTE 1 H .3 . A D M IN IS T E R E D D O S E B M D S R E S U L T S 2 H .3 .1 . H a sso u n et al., 2 0 0 0 : C y to c h r o m e C R e d u c ta se 3 H .3 .1 .1 . Sum m ary Table o f B M D S M odeling Results Model a Degrees of Freedom x 2p Value AIC BMD (ng/kg-d) BMDL (ng/kg-d) Notes exponential (M 2) 4 0.002 -139.075 3.939E+01 3.254E+01 exponential (M 3) 4 0 .0 0 2 -1 3 9 .0 7 5 3.9 3 9 E + 0 1 3.2 5 4 E + 0 1 p o w er hit bound (d = 1) exponential (M4) b 3 0.637 -151.807 9.085E+00 5.886E+00 exponential (M 5) 2 0.786 -151.023 1.420E+01 6.537E+00 Hill 2 0.741 -150.905 1.513E+01 6.277E+00 linear 4 0.032 -144.946 2.470E+01 1.933E+01 polynom ial, 5degree 4 0.032 -144.946 2.470E+01 1.933E+01 power 4 0 .0 3 2 -1 4 4 .9 4 6 2 .4 7 0 E + 0 1 1.933E +01 p o w er b oun d hit (p ow er = 1) power, unrestricted c 3 0.211 -148.989 6.573E +00 1.966E+00 unrestricted (power = 0.574) a C onstant variance m odel selected (p = 0 .3 8 7 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 4 5 6 H .3 .1 .2 . O utputfo r Selected M odel: E xponential (M4) 7 H a sso u n et al., 2 0 0 0 : C y to ch ro m e C red u ctase 8 9 10 E x p o n e n t i a l M o d e l . (Version: 1.61; Date: 7 / 2 4 / 2 0 0 9 ) 11 I n p u t D a t a F i l e : C : \ 5 \ 1 7 _ H a s _ 2 0 0 0 _ C y t C L i v _ E x p C V _ 1 . ( d ) 12 G n u p l o t P l o t t i n g File: 13 F r i A p r 30 2 1 : 1 5 : 2 0 2 0 1 0 14 15 16 T B A R s , l i v e r o n l y ( T a b l e 2) 17 18 19 T h e f o r m of t h e r e s p o n s e f u n c t i o n b y M o d e l : 20 M o d e l 2: Y dose] = a * exp{sign * b * dose} 21 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 22 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 23 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 24 25 Note: Y[dose] is the m e dian response for exposure = dose; 26 sign = +1 for increasing trend in data; 27 sign = -1 for d e c r e a s i n g trend. 28 2 9 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 30 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. This document is a draftfor review purposes only and does not constitute Agency policy. H-41 DRAFT--DO NOT CITE OR QUOTE 1 Model 4 is nested within Model 5. 2 3 4 Dependent variable = Mean 5 Independent variable = Dose 6 Data are assumed to be distributed: normally 7 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 8 r h o is s e t t o 0. 9 A constant v a r i a n c e model is fit. 10 11 T o t a l n u m b e r o f d o s e g r o u p s = 6 12 T o t a l n u m b e r of r e c o r d s w i t h m i s s i n g v a l u e s = 0 13 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 14 R e l a t i v e F u n c t i o n C o n v e r g e n c e has b e e n set to: 1 e - 0 0 8 15 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 16 17 M L E s o l u t i o n p r o v i d e d : E x a c t 18 19 20 Initial Parameter Values 21 22 Variable Model 4 23 24 lnalpha -5.48625 25 rho(S) 0 26 a 0.1387 27 b 0.027423 28 c 3.36121 29 d 1 30 31 (S) = S p e c i f i e d 32 33 34 35 P a r ameter Es t i m a t e s 36 37 Variable Model 4 38 39 lnalpha -5.43908 40 rho 0 41 a 0.141259 42 b 0.0235562 43 c 3.42165 44 d 1 45 46 47 Table of Stats From Input Data 48 49 Dose N Obs Mean Obs Std Dev 50 51 06 0. 1 4 6 0.06614 52 36 0. 177 0.05389 53 10 6 0. 1 9 1 0.05634 54 22 6 0. 2 7 1 0.05634 55 46 6 0. 3 8 8 0.06369 56 100 6 0. 444 0.1102 57 58 59 Estimated Values o f Interest 60 61 Dose Est Mean Est Std Scaled Residual 62 63 0 0.1413 0.06591 0 .1762 64 3 0.1646 0.06591 0 .4609 65 10 0.2131 0.06591 -0 .8196 66 22 0.2796 0.06591 -0 .3199 67 46 0.3676 0.06591 0 .7587 68 100 0.4509 0.06591 -0 .2564 69 70 71 This document is a draftfor review purposes only and does not constitute Agency policy. H-42 DRAFT--DO NOT CITE OR QUOTE 1 Other models for which likelihoods are calculated: 2 3 Model A1: Yij = Mu(i) + e(ij) 4 V a r { e i ;ij)} = S i g m a ^ 2 5 6 Model A2: Yij = Mu(i) + e(ij) 7 V a r { e i :ij)} = S i g m a ( i ) ^ 2 8 9 Model A3: Yij = Mu(i) + e(ij) 10 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 11 12 M o d e l R: Yij = M u + e(i) 13 V a r { e i ij)} = S i g m a ^ 2 14 15 16 L i k e l i h o o d s of I n t e r e s t 17 18 Model Log(likelihood) DF AIC 19 20 A1 80.75258 7 -147.5052 21 A2 83.37355 12 - 1 4 2 . 7 4 7 1 22 A3 80.75258 7 -147.5052 23 R 55.82002 2 -107.64 24 4 79.90337 4 -151.8067 25 26 27 Additive constant for all log-likelihoods = -33.08. This constant added to the 28 above values gives the log-likelihood including the term that does not 29 depend on the model parameters. 30 31 32 Explanation of Tests 33 34 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 35 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 36 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 37 38 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 39 40 41 Tests of Interest 42 43 Test -2*log(Likelihood Ratio) D. F. p-value 44 45 Test 1 55.11 10 < 0.0001 46 Test 2 5.242 5 0.3871 47 Test 3 5.242 5 0.3871 48 Test 6a 1.698 3 0.6373 49 50 51 The p - v a l u e for T e s t 1 is less t h a n .05. T h e r e a p p e a r s to be a 52 difference between response and/or variances among the dose 53 levels, it s e ems a p p r o p r i a t e to m o d e l the data. 54 55 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 56 variance model appears to be appropriate here. 57 58 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 59 variance appears to be appropriate here. 60 61 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 62 to adequately describe the data. 63 64 65 Benchmark Dose Computations: 66 67 Specified Effect = 1.000000 68 69 Risk Type = Estimated standard deviations from control 70 71 C o n f i d e n c e Level = 0.950000 This document is a draftfor review purposes only and does not constitute Agency policy. H-43 DRAFT--DO NOT CITE OR QUOTE 1 2 BMD 9.0851 3 4 BMDL = 5.88612 5 H .3 .1 .3 . F igure fo r Selected M odel: E xponential (M4) Exponential Model 4 with 0.95 Confidence Level 6 7 8 9 H .3 .1 .4 . O utputfo r A dditional M odel Presented: Power, Unrestricted 10 H a s s o u n e t a l., 2 0 0 0 : C y t o c h r o m e C r e d u c ta s e 11 12 13 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 14 I n p u t D a t a File: C : \ 5 \ 1 7 _ H a s _ 2 0 0 0 _ C y t C L i v _ P w r C V _ U _ 1 . ( d ) 15 G n u p l o t P l o t t i n g F i l e : C : \ 5 \ 1 7 _ H a s _ 2 0 0 0 _ C y t C L i v _ P w r C V _ U _ 1 . p l t 16 _ Fri A p r 30 2 1 7 1 5 : 2 6 2 0 1 0 17 18 19 T B A R s , l i v e r o n l y ( T a b l e 2) 20 21 22 The f o r m of the r e s p o n s e f u n c t i o n is: 23 24 Y[dose] = control + slope * d o s e Apower 25 26 27 Dependent variable = Mean 28 Independent variable = Dose 29 rho is set to 0 30 The power is not restricted 31 A c o n s t a n t v a r i a n c e m o d e l is fit This document is a draftfor review purposes only and does not constitute Agency policy. H -44 D R A F T -- D O N O T C ITE O R Q U O T E 1 2 Total number of dose groups = 6 3 Total number of records with missing values = 0 4 Maximum number of iterations = 250 5 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 6 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 7 8 9 10 D e f a u l t I n i t i a l P a r a m e t e r V a l u e s 11 alpha = 0.004972 12 rho = 0 Specified 13 control = 0.146 14 slope = 0.0109242 15 power = 0.717914 16 17 18 A s y m p t o t i c C o r r e l a t i o n M a t r i x of P a r a m e t e r E s t i m a t e s 19 20 ( *** The model parameter(s) -rho 21 have bee n es t i m a t e d at a b o u n d a r y point, or have bee n sp e c i f i e d b y the user, 22 and do not appear in the correlation matrix ) 23 24 alpha control slope power 25 26 alpha 1 -8.8e-010 -3.8e-009 4.5e-009 27 28 control -8.8e-010 1 -0.77 0.68 29 30 slope -3.8e-009 -0.77 1 -0.98 31 32 power 4.5e-009 0.68 -0.98 1 33 34 35 36 Parameter Estimates 37 38 95.0% Wald Confidence Interval 39 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 40 alpha 0.00469717 0.00110713 0.00252723 0.00686711 41 control 0.135495 0.0246289 0.0872229 0.183766 42 slope 0.0232652 0.013381 -0.00296103 0.0494915 43 power 0.573772 0.119032 0.340474 0.80707 44 45 46 47 Table of Data and Estimated Values of Interest 48 49 Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res. 50 51 52 06 0. 1 4 6 0.135 0.0661 0. 0 6 8 5 0.375 53 36 0. 177 0.179 0.0539 0. 0 6 8 5 -0.0784 54 10 6 0. 1 9 1 0.223 0.0563 0. 0 6 8 5 -1.13 55 22 6 0. 2 7 1 0.273 0.0563 0. 0 6 8 5 -0.056 56 46 6 0. 3 8 8 0.345 0.0637 0. 0 6 8 5 1.54 57 100 6 0. 444 0.462 0.11 0. 0 6 8 5 -0.653 58 59 60 61 M odel Descr i p t i o n s for l i k e l ihoods calcu l a t e d 62 63 64 Model A1: Yij = M u ( i ) + e( 65 V a r { e i ;ij)} = S i g m a ^ 2 66 67 Model A2: Yij = M u ( i ) + e( 6 8 V a r { e i :ij)} = S i g m a ( i ) ^ 2 69 70 Model A3: Yij = M u ( i ) + e( 71 V a r { e i :ij)} = S i g m a ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. H-45 DRAFT--DO NOT CITE OR QUOTE 1 Model A3 uses any fixed variance parameters that 2 were specified by the user 3 4 M o d e l R: Yi = M u + e(i) 5 Var{e(i)} = Sigma^2 6 7 8 Likelihoods of Interest 9 10 Model Log(likelihood) # Param's AIC 11 A1 80.752584 7 -147.505168 12 A2 8 3 . 3 7 3 5 4 7 12 - 1 4 2 . 7 4 7 0 9 4 13 A3 80.752584 7 -147.505168 14 fitted 78.494318 4 -148.988637 15 R 55.820023 2 -107.640047 16 17 18 E x p l a n a t i o n of T e s t s 19 2 0 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 21 (A2 vs. R) 2 2 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 23 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 2 4 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 25 (Note When rho=0 the results of Test 3 and Test 2 will be the same 26 27 Tests of Interest 28 29 Test -2*log(Likeli hood Ratio) Test df 30 P -val 31 Test 1 55.107 10 <. 0 0 0 1 32 Test 2 5 . 2 4 1 9 3 5 0. 3 8 7 1 33 Test 3 5 . 2 4 1 9 3 5 0. 3 8 7 1 34 Test 4 4 . 5 1 6 5 3 3 0. 2 1 0 8 35 36 p - v a l u e for T e s t 1 i s less t h a n .05. T h e r e a p pea rs t be a 37 difference between response and/or variances among the dose levels 38 It seems appropriate to model the data 39 4 0 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 41 model appears to be app r o p r i a t e here 42 43 4 4 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 45 to be appropriate here 46 4 7 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 48 to adequately describe the data 49 50 51 B e n c h m a r k Dose C o m p u t a t i o n 52 53 S p e c i f i e d e f f e c t = 1 54 55 Risk Type = Estimated standard deviations from the control mean 56 57 Confidence level = 0.95 58 59 BMD = 6.57302 60 61 62 BMDL = 1.96558 63 This document is a draftfor review purposes only and does not constitute Agency policy. H-46 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 H.3.1.5. F ig u r e f o r A d d itio n a l M o d e l P r e s e n te d : P o w e r, U n re stric te d Power Model with 0.95 Confidence Level 2 21:15 04/30 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. H-47 DRAFT--DO NOT CITE OR QUOTE 1 H.3.2. Hassoun et al., 2000: D N A Single-Strand Breaks 2 H.3.2.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model a Degrees of Freedom x2pValue AIC BMD (ng/kg-d) BMDL (ng/kg-d) Notes exponential (M 2) 4 <0.0001 120.828 3.006E+01 2.491E+01 exponential (M 3) 4 < 0 .0 0 0 1 120.828 3.0 0 6 E + 0 1 2 .4 9 1 E + 0 1 p ow er hit b ound (d = 1) exponential (M 4) exponential (M 5) Hill b 3 3 3 0.036 82.814 3.734E+00 2.783E+00 0.036 8 2 .8 1 4 3 .7 3 4 E + 0 0 2 .7 8 3 E + 0 0 p ow er hit b ound (d = 1) 0.068 8 1 .4 0 7 2 .8 9 0 E + 0 0 2 .0 0 7 E + 0 0 n lo w er b oun d hit (n = 1) linear 4 <.0001 111.165 1.807E+01 1.452E+01 polynom ial, 5degree 4 <.0001 111.165 1.807E+01 1.452E+01 power 4 < .0001 111.165 1.807E +01 1.452E +01 p ow er b oun d hit (p ow er = 1) H ill, unrestricted c 2 0.133 80.318 9.618E-01 2.114E-01 unrestricted (n = 0.613) a C onstant variance m odel selected (p = 0 .7 5 2 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 H .3 .2 .2 . O utputfo r Selected M odel: H ill 6 H a sso u n et al., 2 0 0 0 : D N A sin g le-stra n d b reak s 7 8 9 Hill Model. (Version: 2.14; Date: 06/26/2008) 10 I n p u t D a t a File: C : \ 5 \ 1 8 _ H a s _ 2 0 0 0 _ S S B _ H i l l C V _ 1 . ( d ) 11 G n u p l o t P l o t t i n g F i l e : C : \ 5 \ 1 8 _ H a s _ 2 0 0 0 _ S S B _ H i l l C V _ 1 . p l t 12 _ Fri A p r 30 2 1 : 1 6 : 2 8 2 0 1 0 13 14 15 D N A s i n g l e - s t r a n d b r e a k s , l i v e r o n l y ( T a b l e 3) 16 17 18 T h e f o r m of t h e r e s p o n s e f u n c t i o n is: 19 2 0 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 21 22 23 D e p endent v a r i a b l e = Mea n 24 Independent variable = Dose 25 rho is set to 0 26 Power parameter restricted to be greater than 1 27 A constant v ariance model is fit 28 29 Total number of dose groups = 6 30 Total number of records with missing values = 0 31 M a x i m u m n u m b e r of i t e r a t i o n s = 250 This document is a draftfor review purposes only and does not constitute Agency policy. H-48 DRAFT--DO NOT CITE OR QUOTE 1 Relative Function has been set to: 1e-008 2 Parameter has been set to: 1e-008 3 4 5 6 Default Initial Parameter Values 7 alpha 2.7831 8 rho 0 Specified 9 intercept 7.41 10 1 6 . 0 9 11 0 . 1 7 4 8 3 1 12 6 9 . 2 7 0 6 13 14 15 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 16 17 ( *** The model parameter(s) -rho -n 18 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , or h a v e b e e n s p e c i f i e d b y t h e user, 19 a n d do n o t a p p e a r in t h e c o r r e l a t i o n m a t r i x ) 20 21 alpha intercept v k 22 23 alpha 1 1.1e-007 1.9e-007 1.9e-007 24 25 intercept 1.1e-007 1 0.099 0.61 26 27 v 1.9e-007 0.099 1 0.79 28 29 k 1.9e-007 0.61 0.79 1 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 95.0% W a l d C o n f i d e n c e Interval 36 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Lim 37 alpha 2.82659 0.666233 1.5208 4.13238 38 intercept 8.16404 0.581043 7.02522 9.30286 39 v 20.1253 1.69013 16.8127 23.4379 40 n 1 NA 41 k 31.702 8.35815 15.3203 48.0836 42 43 N A - Indicates that this p a r a m e t e r has hit a bound 44 implied by some inequality constraint and thus 45 has no standard error. 46 47 48 49 Table of Data and Estimated Values of Interest 50 51 Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 52 53 54 06 7.41 8.16 1.54 1.68 -1.1 55 36 10.8 9.9 1.25 1.68 1.28 56 10 6 13.6 13 1.69 1.68 0.889 57 22 6 15.3 16.4 1.71 1.68 -1.62 58 46 6 20.4 20.1 2.25 1.68 0.469 59 100 6 23.5 23.4 1.37 1.68 0.0802 60 61 62 63 Model Descr i p t i o n s for l i k e l ihoods calcu l a t e d 64 65 66 Model A1: Yij = Mu(i) + e(ij) 6 7 V a r { e i ;i j ) } = S i g m a ^ 2 68 69 Model A2: Yij = Mu(i) + e(ij) 7 0 V a r { e i :i j ) } = S i g m a ( i ) ^ 2 71 This document is a draftfor review purposes only and does not constitute Agency policy. H-49 DRAFT--DO NOT CITE OR QUOTE 1 Model A3: Yij = Mu(i) + e(ij) 2 Var{e(ij)} = Sigma^2 3 Model A3 uses any fixed variance parameters that 4 were specified by the user 5 6 M o d e l R: Yi = M u + e(i) 7 Var{e(i)} = Sigma^2 8 9 10 L i k e l i h o o d s of I n t e r e s t 11 12 Model Log(likelihood) # Param's AIC 13 A1 -33.142389 7 80.284779 14 A2 -31.811970 12 8 7 . 6 2 3 9 4 0 15 A3 -33.142389 7 80.284779 16 fitted -36.703273 4 81.406545 17 R -80.442086 2 164.884172 18 19 20 Explanation of Tests 21 2 2 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 23 (A2 vs. R) 2 4 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 v s A2) 25 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 2 6 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? (A3 vs. f i t t e d ) 27 (Note: When rho=0 the results of Test 3 and Test 2 will be the same.) 28 29 Tests of Interest 30 31 Test -2*log(Likelihood Ratio) Test df p-value 32 33 Test 1 97.2602 10 <.0001 34 Test 2 2.66084 5 0.7521 35 Test 3 2.66084 5 0.7521 36 Test 4 7.12177 3 0.06812 37 38 The p - v a l u e for Test 1 is less t h a n .05. T h e r e a p p e a r s to be a 39 difference between response and/or variances among the dose levels 40 It seems appropriate to model the data 41 4 2 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 43 model appears to be app r o p r i a t e here 44 45 4 6 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 47 to be appropriate here 48 4 9 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 50 model 51 52 53 B e n c h m a r k Dose C o m p u t a t i o n 54 55 Sp e c i f i e d effect = 1 56 57 Risk Type = Estimated standard deviations from the control mean 58 59 Confidence level 0.95 60 61 BMD 2.88976 62 63 BMDL 2.00669 64 This document is a draftfor review purposes only and does not constitute Agency policy. H-50 DRAFT--DO NOT CITE OR QUOTE 1 H .3 .2 .3 . F igure fo r Selected M odel: H ill Hill Model with 0.95 Confidence Level 2 3 4 5 H .3 .2 .4 . O utputfo r A dditional M odel Presented: Hill, Unrestricted 6 H a sso u n et al., 2 0 0 0 : D N A sin g le-stra n d b reak s 7 8 9 Hill Model. (Version: 2.14; Date: 06/26/2008) 10 I n p u t D a t a File: C : \ 5 \ 1 8 _ H a s _ 2 0 0 0 _ S S B _ H i l l C V _ U _ 1 . ( d ) 11 G n u p l o t P l o t t i n g F i l e : C : \ 5 \ 1 8 _ H a s _ 2 0 0 0 _ S S B _ H i l l C V _ U _ 1 . p l t 12 _ Fri A p r 3 0 _ 2 1 : 1 6 : 3 0 2 0 1 0 13 14 15 D N A s i n g l e - s t r a n d b r e a k s , l i v e r o n l y ( T a b l e 3) 16 17 18 T h e f o r m of t h e r e s p o n s e f u n c t i o n is: 19 2 0 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 21 22 23 D e p endent v a r i a b l e = Mea n 24 Independent variable = Dose 25 rho is set to 0 26 Power pa r a m e t e r is not restricted 27 A constant v ariance model is fit 28 29 Total number of dose groups = 6 30 Total number of records with missing values = 0 31 M a x i m u m n u m b e r of i t e r a t i o n s = 250 This document is a draftfor review purposes only and does not constitute Agency policy. H-51 DRAFT--DO NOT CITE OR QUOTE 1 Relative Function has been set to: 1e-008 2 Parameter has been set to: 1e-008 3 4 5 6 Default Initial Parameter Values 7 alpha 2.7831 8 rho 0 Specified 9 intercept 7.41 10 1 6 . 0 9 11 0 . 1 7 4 8 3 1 12 6 9 . 2 7 0 6 13 14 15 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 16 17 ( j T h e m o d e l p a r a m e t e r ( s ) - r h o 18 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , or h a v e b e e n s p e c i f i e d b y t h e u s e r 19 a n d do n o t a p p e a r in t h e c o r r e l a t i o n m a t r i x ) 20 21 alpha intercept v n k 22 23 alpha 1 -2.2e-008 -4.6e-008 8. 4 e - 0 0 9 -4.3e -008 24 25 intercept -2.2e-008 1 -0.33 0.47 - 0.29 26 27 v -4.6e-008 -0.33 1 -0.95 1 28 29 n 8.4e-009 0.47 -0.95 1 - 0.96 30 31 k -4.3e-008 -0.29 1 -0.96 1 32 33 34 35 P a r ameter Estimates 36 37 95.0% Wald Confidence Interval 38 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 39 alpha 2.5942 0.611459 1.39576 3.79264 40 intercept 7.47627 0.665055 6.17278 8.77975 41 v 36.9014 25.5466 -13.1689 86.9718 42 n 0.612877 0.190055 0.240376 0.985377 43 k 148.104 303.532 -446.809 743.016 44 45 46 4 7 T a b l e o f D a t a a n d ]E s t i m a t e d V a l u e s o f I n t e r e s t 48 49 Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res. 50 51 52 06 7.41 7.48 1.54 1.61 - 0.101 53 36 10.8 10.6 1.25 1.61 0.313 54 10 6 13.6 13.4 1.69 1.61 0.286 55 22 6 15.3 16.2 1.71 1.61 -1.41 56 46 6 20.4 19.6 2.25 1.61 1.24 57 100 6 23.5 23.7 1.37 1.61 -0.33 58 59 60 61 M odel Descr i p t i o n s for l i k e l ihoods calcu l a t e d 62 63 64 Model A1: Yij = M u ( i ) + e( 65 V a r { e i ;ij)} = S i g m a ^ 2 66 67 Model A2: Yij = M u ( i ) + e( 6 8 V a r { e i :ij)} = S i g m a ( i ) ^ 2 69 70 Model A3: Yij = M u ( i ) + e( 71 V a r { e i :ij)} = S i g m a ^ 2 This document is a draftfor review purposes only and does not constitute Agency policy. H-52 DRAFT--DO NOT CITE OR QUOTE 1 Model A3 uses any fixed variance parameters that 2 were specified by the user 3 4 M o d e l R: Yi = M u + e(i) 5 Var{e(i)} = Sigma^2 6 7 8 Likelihoods of Interest 9 10 Model Log(likelihood) # Param's AIC 11 A1 -33.142389 7 80.284779 12 A2 -31.811970 12 8 7 . 6 2 3 9 4 0 13 A3 -33.142389 7 80.284779 14 fitted -35.159023 5 80.318046 15 R -80.442086 2 164.884172 16 17 18 E x p l a n a t i o n of T e s t s 19 2 0 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 21 (A2 vs. R) 2 2 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 23 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 2 4 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d 25 (Note When rho=0 the results of Test 3 and Test 2 will be the same 26 27 Tests of Interest 28 29 Test -2*log(Likelihood Ratio) Test df p-value 30 31 Test 1 97.2602 10 <.0001 32 Test 2 2.66084 5 0.7521 33 Test 3 2.66084 5 0.7521 34 Test 4 4.03327 2 0.1331 35 36 p - v a l u e for T e s t 1 is less t h a n .05. T h e r e a p p e a r s to be 37 difference between response and/or variances among the dose levels 38 It seems appropriate to model the data 39 4 0 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 41 model appears to be app r o p r i a t e here 42 43 4 4 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 45 to be appropriate here 46 4 7 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 48 to adequately describe the data 49 50 51 B e n c h m a r k Dose C o m p u t a t i o n 52 53 S p e c i f i e d e f f e c t = 1 54 55 Risk Type = Estimated standard deviations from the control mean 56 57 Confidence level = 0.95 58 59 BMD = 0.961789 60 61 BMDL = 0.211403 62 This document is a draftfor review purposes only and does not constitute Agency policy. H-53 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 H.3.2.5. F ig u r e f o r A d d itio n a l M o d e l P r e s e n te d : H ill, U n re stric te d Hill Model with 0.95 Confidence Level 2 21:16 04/30 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. H-54 DRAFT--DO NOT CITE OR QUOTE 1 H.3.3. Hassoun et al., 2000: T B A R S 2 H.3.3.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts Model a Degrees of Freedom x2pValue AIC BMD (ng/kg-d) BMDL (ng/kg-d) Notes exponential (M 2) 4 0.000 -6.143 7.977E+01 5.344E+01 exponential (M 3) 4 0.000 -6.143 7 .9 7 7 E + 0 1 5.3 4 4 E + 0 1 p ow er hit bound (d = 1) exponential (M4) b 3 0.340 -21.181 4.916E+00 2.300E+00 exponential (M 5) 2 0.240 -19.681 6.732E+00 2.470E+00 Hill 2 0.272 -19.932 6.261E+00 2.575E+00 linear 4 0.001 -7.019 6.904E+01 4.373E+01 polynom ial, 5degree 4 0.001 -7.019 6.904E+01 4.373E+01 power 4 0.001 -7 .0 1 9 6 .9 0 4 E + 0 1 4 .3 7 3 E + 0 1 p ow er b oun d hit (p ow er = 1) power, unrestricted c 3 0.023 -14.993 2.902E +00 6.150E -02 unrestricted (power = 0.263) a C onstant variance m odel selected (p = 0 .3 3 4 8 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 H .3 .3 .2 . O utputfo r Selected M odel: E xponential (M4) 6 H a sso u n et al., 2 0 0 0 : T B A R S 7 8 9 Exponential Model. (Version: 1.61; Date: 7/24/2009) 10 I n p u t D a t a File: C : \ 5 \ 1 9 _ H a s _ 2 0 0 0 _ T B A R s L i v _ E x p C V _ 1 . ( d ) 11 G n u p l o t P l o t t i n g Fil e : 12 Fri A p r 30 2 1 : 1 7 : 1 7 2 0 1 0 13 14 15 T B A R s , l i v e r o n l y ( T a b l e 2) 16 17 18 T h e f o r m of t h e r e s p o n s e f u n c t i o n b y M o d e l : 19 M o d e l 2: Y dose] = a * exp{sign * b * dose} 20 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 21 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 22 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 23 24 Note: Y[dose] is the m e dian response for exposure = dose; 25 sign = +1 for increasing trend in data; 26 sign = -1 for d e c r e a s i n g trend. 27 2 8 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 2 9 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 30 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 31 This document is a draftfor review purposes only and does not constitute Agency policy. H-55 DRAFT--DO NOT CITE OR QUOTE 1 2 Dependent variable = Mean 3 Independent variable = Dose 4 Data are assumed to be distributed: normally 5 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 6 r h o is s e t t o 0. 7 A constant v a r i a n c e model is fit. 8 9 Total number of dose groups = 6 10 T o t a l n u m b e r of r e c o r d s w i t h m i s s i n g v a l u e s = 0 11 M a x i m u m n u m b e r o f i t e r a t i o n s = 2 5 0 12 R e l a t i v e F u n c t i o n C o n v e r g e n c e has b e e n set to: 1 e - 0 0 8 13 P a r a m e t e r C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 14 15 M L E s o l u t i o n p r o v i d e d : E x a c t 16 17 18 I n i t i a l P a r a m e t e r V a l u e s 19 20 Variable Model 4 21 22 lnalpha -1.90388 23 rho(S) 0 24 a 1.39555 25 b 0.0194898 26 c 1.97051 27 d 1 28 2 9 (S) = S p e c i f i e d 30 31 32 33 P a r a m e t e r E s t i m a t e s 34 35 Variable Model 4 36 37 lnalpha -1.81059 38 rho 0 39 a 1.40436 40 b 0.0996859 41 c 1.74329 42 d 1 43 44 45 Table of Stats From Input Data 46 47 Dose N Obs Mean Obs Std Dev 48 49 06 1.469 0.2915 50 36 1.549 0.5389 51 10 6 2.15 0.3625 52 22 6 2.28 0.2474 53 46 6 2.619 0.5168 54 100 6 2.292 0.4874 55 56 57 Estimated Values of Interest 58 59 Dose Est Mean Est Std Scaled Residual 60 61 0 1.404 0.4044 0.3915 62 3 1.674 0.4044 -0.7582 63 10 2.063 0.4044 0.527 64 22 2.332 0.4044 -0.3134 65 46 2.438 0.4044 1.099 66 100 2.448 0.4044 -0.9458 67 68 69 70 Other models for which likelihoods are calculated: 71 This document is a draftfor review purposes only and does not constitute Agency policy. H-56 DRAFT--DO NOT CITE OR QUOTE 1 Model A1: Yij = Mu(i) + e(ij) 2 V a r { e i ;ij)} = S i g m a ^ 2 3 4 Model A2: Yij = Mu(i) + e(ij) 5 V a r { e i :ij)} = S i g m a ( i ) ^ 2 6 7 Model A3: Yij = Mu(i) + e(ij) 8 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 9 10 M o d e l R: Yij = M u + e(i) 11 V a r { e i ij)} = S i g m a ^ 2 12 13 14 L i k e l i h o o d s of I n t e r e s t 15 16 Model Log(likelihood) DF AIC 17 18 A1 16.26977 7 -18.53954 19 A2 19.12783 12 - 1 4 . 2 5 5 6 5 20 A3 16.26977 7 -18.53954 21 R 2.44294 2 -0.8858799 22 4 14.5907 4 -21.18141 23 24 25 Additive constant for all log-likelihoods = -33.08. This constant added to the 26 above values gives the log-likelihood including the term that does not 27 depend on the model parameters. 28 29 30 Explanation of Tests 31 32 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 33 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 34 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 35 36 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 37 38 39 Tests of Interest 40 41 Test -2*log(Likelihood Ratio) D. F. p-value 42 43 Test 1 33.37 10 0.000236 44 Test 2 5.716 5 0.3348 45 Test 3 5.716 5 0.3348 46 Test 6a 3.358 3 0.3396 47 48 49 The p - v a l u e for Test 1 is less t h a n .05. T h e r e a p p e a r s to be a 50 difference between response and/or variances among the dose 51 levels, it s e ems a p p r o p r i a t e to m o d e l the data. 52 53 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s 54 variance model appears to be appropriate here. 55 56 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 57 variance appears to be appropriate here. 58 59 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 60 to adequately describe the data. 61 62 63 B e n c h m a r k Dose Computations: 64 65 Specified Effect = 1.000000 66 67 Risk Type = Estimated standard deviations from control 68 69 Confidence Level = 0.950000 70 71 BMD = 4.91639 This document is a draftfor review purposes only and does not constitute Agency policy. H-57 DRAFT--DO NOT CITE OR QUOTE 1 2 BMDL 2.29952 3 H .3 .3 .3 . F igure fo r Selected M odel: E xponential (M4) Exponential Model 4 with 0.95 Confidence Level 4 5 6 7 H .3 .3 .4 . O utputfo r A dditional M odel Presented: Power, Unrestricted 8 H a sso u n et al., 2 0 0 0 : T B A R S 9 10 11 P o w e r M o d e l . ( V e r s i o n : 2 . 1 5 ; D a t e : 0 4 / 0 7 / 2 0 0 8 ) 12 I n p u t D a t a Fil e : C : \ 5 \ 1 9 H a s 2 0 0 0 T B A R s L i v P w r C V U 1. (d) 13 G n u p l o t P l o t t i n g F i l e : C : \ 5 \ 1 9 H a s 2 0 0 0 T B A R s L i v P w r C V U 1. 14 Fri A p r 30 21:17: 15 16 17 T B A R s , l i v e r o n l y ( T a b l e 2) 18 19 20 The f o r m of the r e s p o n s e f u n c t i o n is: 21 22 Y[dose] = control + slope * d o s e Apower 23 24 25 Dependent variable = Mean 26 Independent variable = Dose 27 rho is set to 0 28 The power is not restricted 29 A constant v ariance model is fit 30 31 T o t a l n u m b e r of d o s e g r o u p s = 6 This document is a draftfor review purposes only and does not constitute Agency policy. H-58 DRAFT--DO NOT CITE OR QUOTE 1 Total number of records with missing values = 0 2 Maximum number of iterations = 250 3 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 4 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 5 6 7 8 Default Initial Parameter Values 9 alpha = 0.178788 10 rho = 0S p e c i f i e d 11 control = 1.469 12 s l o p e = 0 . 0 7 5 6 5 3 8 13 power = 0.652114 14 15 16 A s y m p t o t i c C o r r e l a t i o n M a t r i x of P a r a m e t e r E s t i m a t e s 17 18 T h e m o d e l p a r a m e t e r ( s ) - r h o 19 h a v e b e e n e s t i m a t e d at a b o u n d a r y p o i n t , or h a v e b e e n s p e c i f i e d b y t h e user, 20 and do not appear in the correlation matrix ) 21 22 alpha control slope power 23 24 alpha 1 1.1e-008 -1.1e-009 -1.5e-008 25 26 control 1.1e-008 1 -0.75 0.47 27 28 slope -1.1e-009 -0.75 1 -0.91 29 30 power -1.5e-008 0.47 -0.91 1 31 32 33 34 Parameter Estimates 35 36 95.0% Wald Confidence Interval 37 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 38 alpha 0.194232 0.0457809 0 .104503 0.283961 39 control 1.42104 0.171077 1.08573 1.75634 40 slope 0.333105 0.166768 0.00624603 0.659963 41 power 0.262735 0.0983956 0. 0 6 9 8 8 3 6 0.455587 42 43 44 45 Table of Data and Estimated Values of Interest 46 47 Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res. 48 49 50 06 1.47 1.42 0.291 0.441 0.267 51 36 1.55 1.87 0.539 0.441 -1.76 52 10 6 2.15 2.03 0.363 0.441 0.661 53 22 6 2.28 2.17 0.247 0.441 0.603 54 46 6 2.62 2.33 0.517 0.441 1.6 55 100 6 2.29 2.54 0.487 0.441 -1.37 56 57 58 59 Model Descriptions for likelihoods calculated 60 61 62 Model A1: Yij = Mu(i) + e ( i j 63 V ar{e(ij)} = Sigma^2 64 65 Model A2: Yij = Mu(i) + e(ij| 66 Var{e(ij)} = Sigma(i)^2 67 68 Model A3: Yij = Mu(i) + e(ij) 69 Var{e(ij)} = Sigma^2 70 Model A3 uses any fixed variance parameters that 71 wer e sp e c i f i e d b y the user This document is a draftfor review purposes only and does not constitute Agency policy. H-59 DRAFT--DO NOT CITE OR QUOTE 1 2 M o d e l R: Yi = M u + e(i) 3 Var{e(i)} = Sigma^2 4 5 6 Likelihoods of Interest 7 8 Model Log(likelihood) # Param's AIC 9 A1 16.269770 7 -18.539539 10 A2 1 9 . 1 2 7 8 2 7 12 - 1 4 . 2 5 5 6 5 4 11 A3 16.269770 7 -18.539539 12 fitted 11.496634 4 -14.993268 13 R 2 . 4 4 2 9 4 0 2 - 0 . 8 8 5 8 8 0 14 15 16 E x p l a n a t i o n of T e s t s 17 18 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 19 (A2 vs. R) 2 0 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A1 vs A2) 21 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 2 2 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 23 (Note Whe n rho=0 the results of Test 3 and Test 2 will be the same 24 25 Tests of Interest 26 27 Test -2*log(Likelihood Ratio) Test df p-value 28 29 Test 1 33.3698 10 0.000236 30 Test 2 5.71611 5 0.3348 31 Test 3 5.71611 5 0.3348 32 Test 4 9.54627 3 0.02284 33 34 The p - v a l u e for Test 1 is less t h a n .05. T h e r e a p p e a r s to be a 35 d i f f e r e n c e b e t w e e n response and/or va r i a n c e s among the dose levels 36 It seems appropriate to model the data 37 38 T h e p - v a l u e f o r T e s t 2 is g r e a t e r t h a n .1. A h o m o g e n e o u s v a r i a n c e 39 model appears to be appropriate here 40 41 4 2 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 43 to be app r o p r i a t e here 44 45 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 46 model 47 48 49 Benchmark Dose Computation 50 51 S p e c i f i e d e f f e c t = 1 52 53 R i s k T y p e = Estimated standard deviations from the control mean 54 55 C o n f i d e n c e level = 0.95 56 57 BMD = 2.90232 58 59 60 BMDL = 0.0614971 61 This document is a draftfor review purposes only and does not constitute Agency policy. H-60 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 H.3.3.5. F ig u r e f o r A d d itio n a l M o d e l P r e s e n te d : P o w e r, U n re stric te d Power Model with 0.95 Confidence Level 2 21:17 04/30 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. H-61 DRAFT--DO NOT CITE OR QUOTE 1 H.3.4. Kitchin and Woods, 1979: B a p Hydroxylase Activity 2 H.3.4.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts _______ Model a Degrees of Freedom x2pValue AIC BMD (ng/kg-d) BMDL (ng/kg-d) Notes exponential (M 2) 9 <0.0001 452.693 7.939E+03 3.663E+03 exponential (M 3) 9 < 0 .0 0 0 1 45 2 .6 9 3 7 .9 3 9 E + 0 3 3.6 6 3 E + 0 3 p o w er hit b ound (d = 1) exponential (M 4) 8 0.015 226.600 5.458E+00 4.099E+00 exponential (M5) b 7 0.019 226.401 1.022E+01 4.807E+00 H ill 8 <.0001 504.527 error error n upper bound hit (n = 18) linear polynom ial, 8degree power 9 <.0001 299.732 8.276E+00 5.945E+00 3 <.0001 20.000 error error 9 < .0001 2 9 9 .7 3 2 8 .2 7 6 E + 0 0 5 .9 4 5 E + 0 0 p o w er b oun d hit (p ow er = 1) a N on-constant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 H .3 .4 .2 . O utputfo r Selected M odel: E xponential (M5) 6 K itch in and W o o d s, 1979: B aP H y d ro x y la se A ctiv ity 7 8 9 Exponential Model. (Version: 1.61; Date: 7/24/2009) 10 I n p u t D a t a File: C : \ 5 \ 2 7 _ K i t c h i n _ 1 9 7 9 _ H y d r o l a s e _ E x p _ 1 . ( d ) 11 G n u p l o t P l o t t i n g Fil e : 12 Fri A p r 30 2 1 : 1 8 : 0 4 2 0 1 0 13 14 15 K i t c h i n 1 9 7 9 , T b l 3 , B a P h y d r o l a s e a c t i v i t y 16 17 18 T h e f o r m of t h e r e s p o n s e f u n c t i o n b y M o d e l : 19 M o d e l 2: Y dose] = a * exp{sign * b * dose} 20 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 21 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 22 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 23 24 Note: Y[dose] is the m e dian response for exposure = dose; 25 sign = +1 for increasing trend in data; 26 sign = -1 for d e c r e a s i n g trend. 27 2 8 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 2 9 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 30 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 31 32 33 D e p e n d e n t v a r i a b l e = M e a n 34 Independent variable = Dose 35 Data are a s sumed to be distributed: n o r m a l l y 36 Variance Model: exp(lnalpha +rho *ln(Y[dose])) This document is a draftfor review purposes only and does not constitute Agency policy. H-62 DRAFT--DO NOT CITE OR QUOTE 1 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) rho) 2 3 T otal n u m b e r of d o s e g r o u p s = 11 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 MLE solution provided: Exact 10 11 12 I n i t i a l P a r a m e t e r V a l u e s 13 14 Variable Model 5 15 16 lnalpha -3.27793 17 rho 1.92227 18 a 4 . 6 5 5 19 b 0 . 0 0 0 1 7 7 4 3 2 20 c 42.6316 21 d 1 22 23 24 25 Parameter Estimates 26 27 Variable Model 5 28 29 lnalpha -2.64304 30 rho 1.93753 31 a 5 . 4 3 4 2 3 32 b 0.00191658 33 c 3 1 . 2 0 3 3 34 d 1.21503 35 36 37 Table of Stats From Input Data 38 39 Dose N Obs Mean Obs Std 40 41 09 4.9 1.11 42 0.6 4 4.9 1.18 43 24 6.7 1.4 44 44 7.2 1.8 45 20 4 8.3 0.26 46 60 4 14 5 47 200 4 59 6.8 48 600 4 96 46 49 2000 4 155 16.4 50 5000 4 182 26 51 2e+004 4 189 26 52 53 54 Estimated Values of Interest 55 56 Dose Est Mean Est Std Scaled Residual 57 58 0 5.434 1.375 -1.166 59 0.6 5.478 1.386 -0.8347 60 2 5.624 1.421 1.514 61 4 5.875 1.483 1.787 62 20 8.525 2.127 -0.2115 63 60 16.87 4.12 -1.394 64 200 49.41 11.67 1.643 65 600 119.4 27.43 -1.705 66 2000 168.6 38.31 -0.7091 67 5000 169.6 38.53 0.6454 68 2e+004 169.6 38.53 1.009 69 70 71 This document is a draftfor review purposes only and does not constitute Agency policy. H-63 DRAFT--DO NOT CITE OR QUOTE 1 Other models for which likelihoods are calculated: 2 3 Model A1: Yij = Mu(i) + e(ij) 4 V a r { e i ;ij)} = S i g m a ^ 2 5 6 Model A2: Yij = Mu(i) + e(ij) 7 V a r { e i :ij)} = S i g m a ( i ) ^ 2 8 9 Model A3: Yij = Mu(i) + e(ij) 10 V a r { e i :ij)} = e x p ( l a l p h a + l o g ( m e a n ( i ) ) * rho) 11 12 M o d e l R: Yij = M u + e(i) 13 V a r { e i ij)} = S i g m a ^ 2 14 15 16 L i k e l i h o o d s of I n t e r e s t 17 18 Model Log(likelihood) DF AIC 19 20 A1 -158.1306 12 3 4 0 . 2 6 1 3 21 A2 -84.80028 22 2 1 3 . 6 0 0 6 22 A3 -98.82189 13 223.6438 23 R -234.6252 2 473.2504 24 5 -107.2005 6 226.4011 25 26 27 Additive constant for all log-likelihoods = -45.03. This constant added to the 28 above values gives the log-likelihood including the term that does not 29 depend on the model parameters. 30 31 32 Explanation of Tests 33 34 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 35 T e s t 2: A r e H o m o g e n e o u s ? (A2 vs. A1) 36 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 37 38 T e s t 7a: D o e s M o d e l 5 f i t t h e d a t a ? (A3 v s 5) 39 40 41 Tests of Interest 42 43 Test -2*log(Likelihood Ratio) D. F. p-value 44 45 Test 1 299.6 20 < 0.0001 46 Test 2 146.7 10 < 0.0001 47 Test 3 28.04 9 0.0009381 48 Test 7a 16.76 7 0.01903 49 50 51 The p - v a l u e for T e s t 1 is less t h a n .05. T h e r e a p p e a r s to be a 52 difference between response and/or variances among the dose 53 levels, it s e ems a p p r o p r i a t e to m o d e l the data. 54 55 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 56 variance model appears to be appropriate. 57 58 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t to 59 consider a different variance model. 60 61 T h e p - v a l u e f o r T e s t 7a is l e s s t h a n .1. M o d e l 5 m a y n o t a d e q u a t e l y 62 describe the data; you may want to consider another model. 63 64 65 Benchmark Dose Computations: 66 67 Specified Effect = 1.000000 68 69 Risk Type = Estimated standard deviations from control 70 71 C o n f i d e n c e Level = 0.950000 This document is a draftfor review purposes only and does not constitute Agency policy. H-64 DRAFT--DO NOT CITE OR QUOTE 1 2 BMD 10.2235 3 4 BMDL = 4.80673 5 6 7 H .3 .4 .3 . F igure fo r Selected M odel: E xponential (M5) Exponential Model 5 with 0.95 Confidence Level Mean R esponse 8 21:18 04/30 2010 9 dose This document is a draftfor review purposes only and does not constitute Agency policy. H-65 DRAFT--DO NOT CITE OR QUOTE 1 H.3.5. National Toxicology Program, 2006: Liver E R O D 53 W e e k s 2 H.3.5.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts _____________ Model a Degrees of Freedom x 2p Value AIC BMD (ng/kg-d) BMDL (ng/kg-d) Notes exponential (M 2) 4 <0.0001 210.749 4.068E+01 2.856E+01 exponential (M 3) 4 < 0 .0 0 0 1 2 1 0 .7 4 9 4.0 6 8 E + 0 1 2 .8 5 6 E + 0 1 p o w er hit b ound (d = 1) exponential (M 4) 3 0.071 98.835 1.912E-01 1.384E-01 exponential (M 5) Hill b 2 2 0.040 100.232 2.394E-01 1.433E-01 0.219 96.847 3.823E-01 2.336E-01 linear polynom ial, 5degree power 4 <.0001 203.577 2.076E+01 8.128E+00 4 <.0001 203.577 2.076E+01 8.128E+00 4 < .0 0 0 1 2 0 3 .5 7 7 2.0 7 6 E + 0 1 8 .1 2 8 E + 0 0 p o w er b oun d hit (p ow er = 1) a N on-constant variance m odel selected (p = < .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 H .3 .5 .2 . O utputfo r Selected M odel: H ill 6 N ation al T o x ico lo g y Program , 2006: L iver E R O D 53 W eek s 7 8 9 Hill Model. (Version: 2.14; Date: 06/26/2008) 10 I n p u t D a t a File: C : \ 5 \ 4 6 _ N T P _ 2 0 0 6 _ E R O D l i v 5 3 _ H i l l _ 1 . ( d ) 11 G n u p l o t P l o t t i n g F i l e : C : \ 5 \ 4 6 _ N T P _ 2 0 0 6 _ E R O D l i v 5 3 _ H i l l _ 1 . p l t 12 S u n M a y 02 1 5 : 0 5 : 0 2 2 0 1 0 13 14 15 0 16 17 18 T h e f o r m of t h e r e s p o n s e f u n c t i o n is: 19 2 0 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 21 22 23 D e p endent v a r i a b l e = Mea n 24 Independent variable = Dose 25 Power parameter restricted to be greater than 1 26 The v ariance is to be mode l e d as Var(i) = exp(lalpha + rho * ln(mean(i))) 27 28 Total number of dose groups = 6 29 Total number of records with missing values = 0 30 M a x i m u m number of iterations = 250 31 R e l a t i v e F u n c t i o n C o n v e r g e n c e has b e e n set to: 1e-0 0 8 32 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 33 34 35 36 Default Initial Parameter Values This document is a draftfor review purposes only and does not constitute Agency policy. H-66 DRAFT--DO NOT CITE OR QUOTE 1 lalpha 1.59547 2 rho 0 3 intercept 3.614 4 17.599 5 1.38542 6 8.70663 7 8 9 Asymptotic Correlation Matrix of Parameter Estimates 10 11 lalpha rho intercept v n k 12 13 l a l p h a 1 -0.96 -0.16 0.086 -0.057 0.041 14 15 rho -0 .96 1 0.14 -0.11 0.059 -0.045 16 17 i n t e r c e p t -0 .16 0.14 1 -0.18 0.13 0.069 18 19 v 0. 0 8 6 - 0 . 1 1 - 0 . 1 8 1 -0.72 0.84 20 21 n -0. 057 0.059 0.13 -0.72 1 -0.79 22 23 k 0. 041 -0.045 0.069 0.84 -0.79 1 24 25 26 27 Parameter Estimates 28 29 95.0% Wald Confidence Interval 30 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 31 lalpha -4.86522 0.741624 -6.31878 -3.41167 32 rho 2.26949 0.287245 1.7065 2.83248 33 i n t e r c e p t 3.62909 0.133823 3.3668 3.89138 34 v 17.9802 0.989132 16.0416 19.9189 35 n 1.4314 0.162447 1.11301 1.74979 36 k 5.58259 0.717084 4.17713 6.98805 37 38 39 40 Table of Data and Estimated Values of Interest 41 42 Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 43 44 45 08 3.61 3.63 0.486 0.379 -0.113 46 2.14 8 7.27 7.27 0.557 0.833 0.0203 47 7.14 8 14.8 14.2 1.61 1.78 0.911 48 15.7 8 17.3 18.3 1.59 2.37 -1.19 49 32.9 8 20.6 20.3 3.05 2.67 0.304 50 71.4 8 21.2 21.2 3.82 2.8 0.0606 51 52 53 54 Model Descriptions for likelihoods calculated 55 56 57 Model A1: Yij Mu(i) + e(ij) 58 Var{e(ij)} Sigma^2 59 60 Model A2: Yij Mu(i) + e(ij) 61 Var{e(ij)} Sigma(i)^2 62 63 M odel A3: Yij = Mu(i) + e(ij) 64 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 65 Model A3 uses any fixed variance parameters that 66 were specified by the user 67 6 8 M o d e l R: Yi = M u + e(i) 69 Var{e(i)} = Sigma^2 70 71 This document is a draftfor review purposes only and does not constitute Agency policy. H-67 DRAFT--DO NOT CITE OR QUOTE 1 Likelihoods of Interest 2 3 Model Log(likelihood) # Param's AIC 4 A1 -59.086537 7 132.173073 5 A2 -37.515858 12 99. 0 3 1 7 1 6 6 A3 -40.906180 8 97.812359 7 fitted -42.423278 6 96.846556 8 R -116.710291 2 237.420582 9 10 11 E x p l a n a t i o n o f T e s t s 12 13 T e s t 1: D o r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 14 ( A 2 vs. R) 15 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? ( A 1 vs A2) 16 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 17 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 18 (Note W h e n r h o = 0 the r e s u l t s of T e s t 3 a n d T e s t 2 w i l l be t h e s a m e 19 20 Tests of Interest 21 22 Test -2*log(Likelihood Ratio) Test df p-value 23 24 Test 1 158.389 10 <.0001 25 Test 2 43.1414 5 <.0001 26 Test 3 6.78064 4 0.1479 27 Test 4 3.0342 2 0.2193 28 29 The p - v a l u e for Test 1 is less t h a n .05. T h e r e a p p e a r s to be a 30 difference between response and/or variances among the dose levels 31 It s e ems a p p r o p r i a t e to m o d e l the d a t a 32 33 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 34 model appears to be appropriate 35 36 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 37 to be appropriate here 38 39 T h e p - v a l u e f o r T e s t 4 is g r e a t e r t h a n .1. T h e m o d e l c h o s e n s e e m s 40 to adequately describe the data 41 42 43 B e n c h m a r k Dose C o m p u t a t i o n 44 45 Specified effect = 1 46 47 Risk Type = Estimated standard deviations from the control mean 48 49 Confidence level = 0.95 50 51 BMD = 0.382287 52 53 BMDL = 0.233611 54 This document is a draftfor review purposes only and does not constitute Agency policy. H-68 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 H.3.5.3. F ig u r e f o r S e le c te d M o d e l: H ill Hill Model with 0.95 Confidence Level 2 15:05 05/02 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. H-69 DRAFT--DO NOT CITE OR QUOTE 1 H.3.6. National Toxicology Program, 2006: L u n g Erod 53 W e e k s 2 H.3.6.1. S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts ____________ Model a Degrees of Freedom x 2p Value AIC BMD (ng/kg-d) BMDL (ng/kg-d) Notes exponential (M 2) 4 <0.0001 316.324 8.979E+01 5.757E+01 exponential (M 3) exponential (M4) b exponential (M 5) 4 < 0 .0 0 0 1 3 1 6 .3 2 4 8.9 7 9 E + 0 1 5.7 5 7 E + 0 1 p o w er hit b oun d (d = 1) 3 0.421 255.120 8.746E-02 5.370E-02 2 0.276 256.882 6.769E-01 5.491E-02 H ill 2 0.275 256.882 1.454E+00 1.138E-01 linear 4 <.0001 315.961 8.550E+01 4.502E+01 polynom ial, 5degree 4 <.0001 315.961 8.550E+01 4.502E+01 power 4 < .0001 31 5 .9 6 1 8.5 5 0 E + 0 1 4 .5 0 2 E + 0 1 p o w er b ound hit (p ow er = 1) power, unrestricted c 3 0.037 260.794 2.688E -10 2.688E -10 unrestricted (power = 0.129) a N on-con stant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix c Alternate m odel, BM D S output also presented in this appendix 3 4 5 H .3 .6 .2 . O utputfo r Selected M odel: E xponential (M4) 6 N ation al T o x ico lo g y Program , 2006: L u ng E R O D 53 W eek s 7 8 9 Exponential Model. (Version: 1.61; Date: 7/24/2009) 10 I n p u t D a t a File: C : \ 5 \ 5 2 _ N T P _ 2 0 0 6 _ L u n g E R O D 5 3 _ E x p _ 1 . ( d ) 11 G n u p l o t P l o t t i n g Fil e : 12 Fri A p r 30 2 1 : 2 2 : 3 6 2 0 1 0 13 14 15 T b l 12, W e e k 53, L u n g M i c r o s o m e s E R O D 16 17 18 T h e f o r m of t h e r e s p o n s e f u n c t i o n b y M o d e l : 19 M o d e l 2: Y dose] = a * exp{sign * b * dose} 20 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 21 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 22 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 23 24 Note: Y[dose] is the m e dian response for exposure = dose; 25 sign = +1 for increasing trend in data; 26 sign = -1 for d e c r e a s i n g trend. 27 2 8 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 2 9 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 30 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 31 This document is a draftfor review purposes only and does not constitute Agency policy. H-70 DRAFT--DO NOT CITE OR QUOTE 1 2 Dependent variable = Mean 3 Independent variable = Dose 4 Data are assumed to be distributed: normally 5 Variance Model: exp(lnalpha +rho *ln(Y[dose])) 6 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) rho) 7 8 Total number of dose groups = 6 9 Total number of records with missing values = 0 10 M a x i m u m n u m b e r of i t e r a t i o n s = 250 11 R e l a t i v e F u n c t i o n C o n v e r g e n c e h a s b e e n s e t to: 1 e - 0 0 8 12 P a r a m e t e r C o n v e r g e n c e has b e e n set to: 1 e - 0 0 8 13 14 M L E s o l u t i o n p r o v i d e d : E x a c t 15 16 17 I n i t i a l P a r a m e t e r V a l u e s 18 19 Variable Model 4 20 21 lnalpha -0.80064 22 rho 1.47683 23 a 2.86045 24 b 0.054659 25 c 16.0581 26 d 1 27 28 29 30 Parameter Estimates 31 32 Variable Model 4 33 34 lnalpha -1.15021 35 rho 1.63127 36 a 3.06838 37 b 0.414677 38 c 13.847 39 d 1 40 41 42 Table of Stats From Input Data 43 44 Dose N Obs Mean Obs Std Dev 45 46 08 3.011 1.584 47 2.14 8 27.15 5.269 48 7.14 8 42.85 11.15 49 15.7 8 36.57 12.99 50 32.9 8 43.75 18.55 51 71.4 8 43.71 6.322 52 53 54 Estimated Values of Interest 55 56 Dose Est Mean Est Std Scaled Residual 57 58 0 3.068 1.404 - 0.1156 59 2.14 26.26 8.088 0.3116 60 7.14 40.45 11.5 0.5901 61 15.7 42.43 11.96 -1.386 62 32.9 42.49 11.98 0.2972 63 71.4 42.49 11.98 0.2894 64 65 66 67 ther models for which likelihoods are calculated: 68 69 Model A1: Yij = M u ( i ) + e (ij) 70 Var{e(ij)} = Sigma^2 71 This document is a draftfor review purposes only and does not constitute Agency policy. H-71 DRAFT--DO NOT CITE OR QUOTE 1 Model A2: Yij Mu(i) + e(ij) 2 Var{e(ij)} Sigma(i)^2 3 4 Model A3: Yij Mu(i) + e(ij) 5 Var{e(ij)} exp(lalpha + log(mean(i) rho) 6 7 M o d e l R: Yij M u + e(i) 8 Var{e(ij)} Sigma^2 9 10 11 L i k e l i h o o d s o f I n t e r e s t 12 13 Model Log(likelihood) DF AIC 14 15 A1 -135.2677 7 284.5353 16 A2 -115.6885 12 25 5 . 3 7 7 1 17 A3 -121.1517 8 258.3034 18 R -162.0902 2 328.1805 19 4 -122.5601 5 255.1202 20 21 22 Additive constant for all log-likelihoods = -44.11. This constant added to the 23 above values gives the log - l i k e l i h o o d i n c luding the ter m that does not 24 depend on the model parameters. 25 26 27 Explanation of Tests 28 2 9 T e s t 1 D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 30 T e s t 2 A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 31 T e s t 3 A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 32 33 T e s t 6a: D o e s M o d e l 4 f i t t h e d a t a ? (A3 v s 4) 34 35 36 Tests of Interest 37 38 Test -2*log(Likelihood Ratio) D. F. p-value 39 40 Test 1 92.8 10 < 0.0001 41 Test 2 39.16 5 < 0.0001 42 Test 3 10.93 4 0.0274 43 Test 6a 2.817 3 0.4207 44 45 46 The p - v a l u e for T e s t 1 is less t h a n .05. T h e r e a p p e a r s to be a 47 difference between response and/or variances among the dose 48 levels, it seems appropriate to model the data. 49 50 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 51 v a r i a n c e m o d e l a p p e a r s to be a p p r o p r i a t e . 52 53 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t to 54 consider a different variance model. 55 56 T h e p - v a l u e f o r T e s t 6a is g r e a t e r t h a n .1. M o d e l 4 s e e m s 57 to a dequately describe the data. 58 59 60 Benchmark Dose Computations: 61 62 Specified Effect = 1.000000 63 64 Risk Type = Estimated standard deviations from control 65 66 Confidence Level = 0.950000 67 68 BMD = 0.0874595 69 70 BMDL = 0.0537035 This document is a draftfor review purposes only and does not constitute Agency policy. H-72 DRAFT--DO NOT CITE OR QUOTE 1 H .3 .6 .3 . F igure fo r Selected M odel: E xponential (M4) Exponential Model 4 with 0.95 Confidence Level 2 3 4 5 H .3 .6 .4 . O utputfo r A dditional M odel Presented: Power, Unrestricted 6 N ation al T o x ico lo g y Program , 2006: L u ng E R O D 53 W eek s 7 8 9 Power Model. (Version: 2.15; Date: 04/07/2008) 10 I n p u t D a t a File: C : \ 5 \ 5 2 _ N T P _ 2 0 0 6 _ L u n g E R O D 5 3 _ P w r _ U _ 1 . ( d ) 11 G n u p l o t P l o t t i n g F i l e : C : \ 5 \ 5 2 _ N T P _ 2 0 0 6 _ L u n g E R O D 5 3 _ P w r _ U _ 1 . p l t 12 _ Fri A p r 30 2 1 : 2 2 : 4 0 2 0 1 0 13 14 15 T b l 12, W e e k 53, L u n g M i c r o s o m e s E R O D 16 17 18 T h e f o r m of t h e r e s p o n s e f u n c t i o n is: 19 20 Y[dose] = control + slope * d o s e Apower 21 22 23 D e p endent v a r i a b l e = Mea n 24 Independent variable = Dose 25 The power is not restricted 26 The v a r i a n c e is to be m o d e l e d as Var(i) = exp(l a l p h a + log(mean(i)) * rho) 27 28 Total number of dose groups = 6 29 Total number of records with missing values = 0 30 M a x i m u m number of iterations = 250 31 R e l a t i v e F u n c t i o n C o n v e r g e n c e has b e e n set to: 1e-0 0 8 This document is a draftfor review purposes only and does not constitute Agency policy. H-73 DRAFT--DO NOT CITE OR QUOTE 1 Parameter Convergence has been set to: 1e-008 2 3 4 5 Default Initial Parameter Values 6 lalpha = 4.76968 7 rho = 0 8 control = 3.011 9 slope = 24.7003 10 power = 0.132996 11 12 13 A s y m p t o t i c C o r r e l a t i o n M a t r i x o f P a r a m e t e r E s t i m a t e s 14 15 lalpha rho control slope power 16 17 l a l p h a 1 -0.96 -0.48 0.11 -0.048 18 19 rho -0.96 1 0.45 -0.15 0.053 20 21 control -0.48 0.45 1 -0.15 0.05 22 23 slope 0.11 -0.15 -0.15 1 -0.92 24 25 power -0.048 0.053 0.05 -0.92 1 26 27 28 29 Parameter Estimates 30 31 95.0% W a l d C o n f i d e n c e I n t e r v a l 32 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 33 lalpha -1.03242 0.815871 -2.6315 0.566654 34 rho 1.63031 0.239764 1.16038 2.10024 35 control 3.01793 0.518146 2.00238 4.03348 36 slope 25.144 3.39289 18.494 31.7939 37 power 0.128894 0.0448391 0.041011 0.216777 38 39 40 41 Table of Data and E s t i m a t e d Values of Interest 42 43 Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 44 45 46 08 3.01 3.02 1.58 1.47 -0.0133 47 2.14 8 27.1 30.8 5.27 9.74 -1.05 48 7.14 8 42.8 35.4 11.2 10.9 1.92 49 15.7 8 36.6 38.9 13 11.8 -0.553 50 32.9 8 43.7 42.5 18.5 12.7 0.286 51 71.4 8 43.7 46.6 6.32 13.7 -0.598 52 53 54 55 M odel Descr i p t i o n s for l i k e l ihoods calcu l a t e d 56 57 58 Model A1: Yij Mu(i) + e(ij) 59 Var{e(ij)} Sigma^2 60 61 M odel A2: Yij Mu(i) + e(ij) 62 Var{e(ij)} Sigma(i)^2 63 64 Model A3: Yij = Mu(i) + e(ij) 65 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 66 Model A3 uses any fixed variance parameters that 67 were specified by the user 68 6 9 M o d e l R: Yi = M u + e(i) 70 Var{e(i)} = Sigma^2 71 This document is a draftfor review purposes only and does not constitute Agency policy. H-74 DRAFT--DO NOT CITE OR QUOTE 1 2 Likelihoods of Interest 3 4 Model Log(likelihood) # Param's AIC 5 A1 -135.267662 7 284.535325 6 A2 -115.688533 12 2 5 5 . 3 7 7 0 6 7 7 A3 -121.151707 8 258.303413 8 fitted -125.397022 5 260.794043 9 R -162.090242 2 328.180484 10 11 12 E x p l a n a t i o n of T e s t s 13 14 T e s t 1: Do r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 15 ( A 2 vs. R) 16 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? ( A 1 vs A2) 17 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 18 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 19 (Note W h e n r h o = 0 the r e s u l t s of T e s t 3 a n d T e s t 2 w i l l be t h e s a m e 20 21 Tests of Interest 22 23 Test -2*log(Likelihood Ratio) Test df p-value 24 25 Test 1 92.8034 10 <.0001 26 Test 2 39.1583 5 <.0001 27 Test 3 10.9263 4 0.0274 28 Test 4 8.49063 3 0.03689 29 30 The p - v a l u e for Test 1 is less t h a n .05. T h e r e a p p e a r s to be a 31 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g the d o s e leve l s 32 It seems appropriate to model the data 33 34 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 35 model appears to be a p p r o priate 36 37 T h e p - v a l u e f o r T e s t 3 is l e s s t h a n .1. Y o u m a y w a n t t o c o n s i d e r a 38 different variance model 39 4 0 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 41 model 42 43 44 Benchmark Dose Computation 45 46 Specified effect = 1 47 48 Risk Type = Estimated standard deviations from the control mean 49 50 Confidence level = 0.95 51 52 BMD = 2.68823e-010 53 54 55 BMDL = 2.688 2 3 e - 0 1 0 56 This document is a draftfor review purposes only and does not constitute Agency policy. H-75 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 H.3.6.5. F ig u r e f o r A d d itio n a l M o d e l P r e s e n te d : P o w e r, U n re stric te d Power Model with 0.95 Confidence Level 2 21:22 04/30 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. H-76 DRAFT--DO NOT CITE OR QUOTE 1 H .3 .7 . N a tio n a l T o x ico lo g y P ro g r a m , 2 0 0 6 : L a b elin g In d e x 31 W ee k s 2 H . 3 . 7 . 1 . S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts ______________________ Model a Degrees of Freedom x 2p Value AIC BMD (ng/kg-d) BMDL (ng/kg-d) Notes exponential (M2) b 4 0.000 47.304 2.336E+01 1.867E+01 exponential (M 3) 4 0.000 4 7 .3 0 4 2 .3 3 6 E + 0 1 1.867E +01 p o w er hit boun d (d = 1) exponential (M 4) 3 <0.0001 53.331 1.233E+01 7.562E+00 exponential (M 5) 2 <0.0001 51.057 3.279E+01 2.055E+01 H ill 3 0.000 49.057 3.277E+01 error n upper bound hit (n = 18) linear polynom ial, 5degree power 4 <.0001 51.331 1.233E+01 7.563E+00 3 0.000 48.698 2.510E+01 1.192E+01 3 <.0001 49.826 3.238E+01 1.723E+01 a N on-con stant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 H .3 .7 .2 . O utputfo r Selected M odel: E xponential (M2) 6 N ation al T o x ico lo g y Program , 2006: L ab elin g In dex 31 W eek s 7 8 9 Exponential Model. (Version: 1.61; Date: 7/24/2009) 10 I n p u t D a t a File: C : \ 5 \ 3 8 _ N T P _ 2 0 0 6 _ H e p I n d e x _ E x p _ 1 . ( d ) 11 G n u p l o t P l o t t i n g Fil e : 12 Fri A p r 30 2 1 : 2 3 : 2 8 2 0 1 0 13 14 15 T b l 11, 3 1 w k , H e p C e l l P r o l i f e r a t i o n L a b e l i n g I n d e x 16 17 18 T h e f o r m of t h e r e s p o n s e f u n c t i o n b y M o d e l : 19 M o d e l 2: Y dose] = a * exp{sign * b * dose} 20 M o d e l 3: Y d o s e ] = a * e x p { s i g n * (b * d o s e ) ^ d } 21 M o d e l 4: Y dose] = a * [c-(c-1) * exp{-b * dose}] 22 M o d e l 5: Y dose] = a * [c-(c-1) * exp{-(b * dose)^d 23 24 Note: Y[dose] is the m e dian response for exposure = dose; 25 sign = +1 for increasing trend in data; 26 sign = -1 for d e c r e a s i n g trend. 27 2 8 M o d e l 2 is n e s t e d w i t h i n M o d e l s 3 a n d 4. 2 9 M o d e l 3 is n e s t e d w i t h i n M o d e l 5. 30 M o d e l 4 is n e s t e d w i t h i n M o d e l 5. 31 32 33 D e p e n d e n t v a r i a b l e = M e a n 34 Independent variable = Dose 35 Data are a s sumed to be distributed: n o r m a l l y 36 Variance Model: exp(lnalpha +rho *ln(Y[dose])) This document is a draftfor review purposes only and does not constitute Agency policy. H-77 DRAFT--DO NOT CITE OR QUOTE 1 The variance is to be modeled as Var(i) = exp(lalpha + log(mean(i)) rho) 2 3 Total number of dose groups = 6 4 Total number of records with missing values = 0 5 Maximum number of iterations = 250 6 Relat i v e Function C o n v e r g e n c e has bee n set to: 1e-008 7 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 8 9 MLE solution provided: Exact 10 11 12 I n i t i a l P a r a m e t e r V a l u e s 13 14 Variable Model 2 15 16 lnalpha -0.674004 17 rho 2.29189 18 a 0 . 5 7 6 3 6 3 19 b 0 . 0 2 6 6 1 7 4 20 c 0 21 d 1 22 23 24 25 Parameter Estimates 26 27 Variable Model 2 28 29 lnalpha -0.471424 30 rho 1.90298 31 a 0 . 6 1 6 5 3 9 32 b 0.0253715 33 c 0 34 d 1 35 36 37 Table of Stats From Input Data 38 39 Dose N Obs Mean Obs Std Dev 40 41 09 0.327 0.189 42 2.14 10 0.852 0.6514 43 7.14 10 0.956 0.7368 44 15.7 10 0.792 0.4617 45 32.9 10 1.333 1.123 46 71.4 10 3.846 3.08 47 48 49 Estimated Values of Interest 50 51 Dose Est Mean Est Std Scaled Residual 52 53 0 0.6165 0.4986 -1.742 54 2.14 0.6509 0.5251 1.211 55 7.14 0.739 0.5924 1.158 56 15.7 0.9182 0.7284 -0.548 57 32.9 1.421 1.103 -0.2511 58 71.4 3.773 2.795 0.08251 59 60 61 62 Other models for which likelihoods are calculated: 63 64 Model A1: Yij = M u ( i ) + e( ij) 65 V a r { e (ij)} = Sigma"2 66 67 Model A2: Yij = M u ( i ) + e( ij) 6 8 V a r { e (ij)} = S i g m a ( i ) 1"2 69 70 Model A3: Yij = M u ( i ) + e( ij) 71 V a r { e (ij)} = e x p ( l a l p h a + log(mean(i)) * rho) This document is a draftfor review purposes only and does not constitute Agency policy. H-78 DRAFT--DO NOT CITE OR QUOTE 1 2 M o d e l R: Yij = M u + e(i) 3 Var{e(ij)} = Sigma^2 4 5 6 Likelihoods of Interest 7 8 Model Log(likelihood) DF AIC 9 10 A1 -47.23498 7 108.47 11 A2 -8.679256 12 4 1 .35851 12 A3 -8.980651 8 33.9613 13 R -63.44829 2 130.8966 14 2 -19.65195 4 47.30389 15 16 17 Additive constant for all log-likelihoods = -54.22. This constant added to the 18 a b o v e v a l u e s g i v e s t h e l o g - l i k e l i h o o d i n c l u d i n g t h e t e r m t h a t d o e s not 19 d e p e n d on t h e m o d e l p a r a m e t e r s . 20 21 22 Explanation of Tests 23 2 4 T e s t 1: D o e s r e s p o n s e a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? (A2 vs. R) 25 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? (A2 vs. A1) 2 6 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs. A3) 2 7 T e s t 4: D o e s M o d e l 2 f i t t h e d a t a ? (A3 vs. 2) 28 29 30 Tests of Interest 31 32 Test -2*log(Likelihood Ratio) D. F. p-value 33 34 Test 1 109.5 10 < 0.0001 35 Test 2 77.11 5 < 0.0001 36 Test 3 0.6028 4 0.9628 37 Test 4 21.34 4 0.0002708 38 39 40 The p - v a l u e for T e s t 1 is less t h a n .05. T h e r e a p p e a r s to be a 41 d i f f e r e n c e b e t w e e n response and/or va r i a n c e s among the dose 42 levels, it seems appropriate to model the data. 43 4 4 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s 45 variance model appears to be appropriate. 46 4 7 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d 48 variance appears to be appropriate here. 49 50 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. M o d e l 2 m a y n o t a d e q u a t e l y 51 d e s c r i b e the data; y o u m a y w a n t to c o n s i d e r a n o t h e r model. 52 53 54 Benchmark Dose Computations: 55 56 Specified Effect = 1.000000 57 58 Risk Type = Estimated standard deviations from control 59 60 Confidence Level = 0.950000 61 62 BMD 23.3586 63 64 BMDL 18.6683 This document is a draftfor review purposes only and does not constitute Agency policy. H-79 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 H .3 .7 .3 . F igure fo r Selected M odel: E xponential (M2) Exponential Model 2 with 0.95 Confidence Level 2 21:23 04/30 2010 3 This document is a draftfor review purposes only and does not constitute Agency policy. H-80 DRAFT--DO NOT CITE OR QUOTE 1 H .3 .8 . V a n d en H e u v e l et a l., 1994: H e p a tic C Y P 1 A 1 M r n a E x p re ssio n 2 H . 3 . 8 . 1 . S u m m a r y T a b le o f B M D S M o d e lin g R e s u lts ________________________ Model a Degrees of Freedom x 2p Value AIC BMD (ng/kg-d) BMDL (ng/kg-d) Notes exponential (M 2) 5 <0.0001 1164.377 4.699E+03 1.729E+03 exponential (M 3) 5 < 0 .0 0 0 1 1 1 64.377 4 .6 9 9 E + 0 3 1.729E +03 p o w er hit bound (d = 1) exponential (M 4) 4 <0.0001 661.006 4.550E-01 2.643E-01 exponential (M 5) Hill b 3 <0.0001 635.327 1.516E+01 1.046E+01 3 <.0001 662.251 8.091E-01 4.844E-01 linear polynom ial, 6degree power 5 <.0001 667.554 4.953E-01 3.093E-01 1 <.0001 715.412 5.774E+03 1.204E+01 4 <.0001 669.441 5.571E-01 3.204E-01 a N on-con stant variance m odel selected (p = < 0 .0 0 0 1 ) b Best-fitting m odel, BM D S output presented in this appendix 3 4 5 H .3 .8 .2 . O utputfo r Selected M odel: H ill 6 V a n d en H eu v e l et al., 1994: H e p a tic C Y P 1 A 1 m R N A E x p ressio n 7 8 9 Hill Model. (Version: 2.14; Date: 06/26/2008) 10 I n p u t D a t a File: C : \ U s e p a \ B M D S 2 1 \ D a t a \ h i l _ V a n d e n _ m R N A _ S e t t i n g . ( d ) 11 G n u p l o t P l o t t i n g F i l e : C : \ U s e p a \ B M D S 2 1 \ D a t a \ h i l _ V a n d e n _ m R N A _ S e t t i n g . p l t 12 W e d M a y 19 1 4 : 2 5 : 0 6 2 0 1 0 13 14 15 B M D S M o d e l R u n 16 17 18 T h e f o r m of t h e r e s p o n s e f u n c t i o n is: 19 2 0 Y [ d o s e ] = i n t e r c e p t + v * d o s e A n / ( k A n + d o s e A n) 21 22 23 D e p endent v a r i a b l e = m R N A mean 24 Independent variable = d 25 Power pa r a m e t e r is not restricted 26 The v ariance is to be mode l e d as Var(i) = exp(lalpha + rho * ln(mean(i))) 27 28 Total number of dose groups = 7 29 Total number of records with missing values = 0 30 M a x i m u m number of iterations = 250 31 R e l a t i v e F u n c t i o n C o n v e r g e n c e has b e e n set to: 1e-0 0 8 32 P a r ameter C o n v e r g e n c e has bee n set to: 1e-008 33 34 35 36 Default Initial Parameter Values This document is a draftfor review purposes only and does not constitute Agency policy. H-81 DRAFT--DO NOT CITE OR QUOTE 1 lalpha 18.2064 2 rho 0 3 intercept 5.4 4 36694.6 5 0.720907 6 18830.3 7 8 9 Asymptotic Correlation Matrix of Parameter Estimates 10 11 lalpha rho intercept v n k 12 Lf) CO r Lf) oo Lf) O g Lf) CO Lf) r o oo CO o 13 l a l p h a 1 -0.41 0.37 0.7 -0.2 14 15 r h o 1 0.29 0.24 16 17 i n t e r c e p t -0.41 0.29 1 -0.11 0.13 -0.034 18 19 v 0.37 -0.11 1 0.21 0.57 20 21 n 0.7 0.13 0.21 1 -0.53 22 23 k -0.2 0.24 -0.034 0.57 1 24 25 26 27 Parameter Estimates 28 29 95.0% Wald Confidence Interval 30 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf. Limit 31 lalpha -0.28219 -1.71928 1.1549 32 rho 2.05171 0.146654 1.76427 2.33915 33 i n t e r c e p t 5.4299 1.14997 3.17599 7.68381 34 v 36598.9 13930.2 9296.23 63901.7 35 n 1.13992 0.0919476 0.959705 1.32013 36 k 2012.71 881.73 284.554 3740.87 37 38 39 40 Table of Data and Estimated Values of Interest 41 42 Dose N Obs Mean Est Mean Obs Std Dev Est Std Dev Scaled Res 43 44 45 0 13 5.4 5.43 3.61 4.93 -0.0219 46 0.1 5 7.2 5.88 5.59 5.35 0.55 47 1 12 14.8 11.7 14.9 10.8 0.991 48 10 7 12.8 91.8 4.5 89.6 -2.33 49 100 7 536 1.16e+003 320 1.21e+003 -1.37 50 1000 11 1 . 8 e + 0 0 4 1.14e+004 1.52e+004 1.26e+004 1.75 51 1e+004 5 3.67e+004 3.15e+004 2.21e+004 3.58e+004 0.323 52 53 54 55 M odel Descr i p t i o n s for l i k e l ihoods calcu l a t e d 56 57 58 Model A1: Yij Mu(i) + e(ij) 59 Var{e(ij)} Sigma^2 60 61 M odel A2: Yij Mu(i) + e(ij) 62 Var{e(ij)} Sigma(i)^2 63 64 Model A3: Yij = Mu(i) + e(ij) 65 Var{e(ij)} = exp(lalpha + rho*ln(Mu(i))) 66 Model A3 uses any fixed variance parameters that 67 were specified by the user 68 6 9 M o d e l R: Yi = M u + e(i) 70 Var{e(i)} = Sigma^2 71 This document is a draftfor review purposes only and does not constitute Agency policy. H-82 DRAFT--DO NOT CITE OR QUOTE 1 2 Likelihoods of Interest 3 4 Model Log(likelihood) # Param's AIC 5 A1 -572.470944 8 1160.941889 6 A2 -290.799287 14 6 0 9 . 5 9 8 5 7 5 7 A3 -293.809342 9 605.618684 8 fitted -325.125462 6 662.250924 9 R -603.663396 2 1211.326792 10 11 12 E x p l a n a t i o n of T e s t s 13 14 T e s t 1: D o r e s p o n s e s a n d / o r v a r i a n c e s d i f f e r a m o n g D o s e l e v e l s ? 15 ( A 2 vs. R) 16 T e s t 2: A r e V a r i a n c e s H o m o g e n e o u s ? ( A 1 vs A2) 17 T e s t 3: A r e v a r i a n c e s a d e q u a t e l y m o d e l e d ? (A2 vs A3) 18 T e s t 4: D o e s t h e M o d e l f o r t h e M e a n F i t ? A 3 vs. f i t t e d ) 19 (Note W h e n r h o = 0 the r e s u l t s of T e s t 3 a n d T e s t 2 w i l l be t h e s a m e 20 21 Tests of Interest 22 23 Test -2*log(Likelihood Ratio) Test df p-value 24 25 Test 1 625.728 12 <.0001 26 Test 2 563.343 6 <.0001 27 Test 3 6.02011 5 0.3043 28 Test 4 62.6322 3 <.0001 29 30 The p - v a l u e for Test 1 is less t h a n .05. T h e r e a p p e a r s to be a 31 d i f f e r e n c e b e t w e e n r e s p o n s e a n d / o r v a r i a n c e s a m o n g the d o s e leve l s 32 It seems appropriate to model the data 33 34 T h e p - v a l u e f o r T e s t 2 is l e s s t h a n .1. A n o n - h o m o g e n e o u s v a r i a n c e 35 model appears to be a p p r o priate 36 37 T h e p - v a l u e f o r T e s t 3 is g r e a t e r t h a n .1. T h e m o d e l e d v a r i a n c e a p p e a r s 38 to be appropriate here 39 4 0 T h e p - v a l u e f o r T e s t 4 is l e s s t h a n .1. Y o u m a y w a n t t o t r y a d i f f e r e n t 41 model 42 43 44 Benchmark Dose Computation 45 46 Specified effect = 1 47 48 Risk Type = Estimated standard deviations from the control mean 49 50 Confidence l e v e l = 0.95 51 52 BMD = 0.809125 53 54 BMDL = 0.484455 55 56 This document is a draftfor review purposes only and does not constitute Agency policy. H-83 DRAFT--DO NOT CITE OR QUOTE Mean Response 1 H .3 .8 .3 . F igure fo r Selected M odel: E xponential (M5) Hill Model with 0.95 Confidence Level 2 14:25 05/19 2010 3 4 This document is a draftfor review purposes only and does not constitute Agency policy. H-84 DRAFT--DO NOT CITE OR QUOTE [T h is p a g e in ten tio n a lly le ft b lan k .] DRAFT D O N O T C ITE O R Q U O T E M ay 2010 E xternal R ev ie w D raft APPENDIX I Effect of Background Exposure on Benchmark-Dose Modeling N O T IC E T H IS D O C U M E N T IS A N E X T E R N A L R E V IE W D R A F T . It has n ot b een form ally released b y the U .S . E n viron m en tal P rotection A g en cy and sh ou ld n ot at th is stage b e con stru ed to represent A g en cy p o licy . It is b ein g circulated for com m en t on its tech n ical accuracy and p o licy im p lication s. N ation al C enter for E n viron m en tal A ssessm en t O ffice o f R esearch and D evelop m en t U .S . E n viron m en tal P rotection A g en cy C in cin n ati, O H C O N T E N T S -- A P P E N D IX I: E ffect o f B a ck g ro u n d E x p o su re on B en ch m a rk -D o se M o d elin g L I S T O F F I G U R E S ............................................................................................................................................................................... I - ii A P P E N D I X I. E F F E C T O F B A C K G R O U N D E X P O S U R E O N B E N C H M A R K -D O S E M O D E L I N G ............................................................................................................................................................I-1 1.1. N T P , 2 0 0 6 (C H O L A N G IO C A R C IN O M A S ): U N A D J U S T E D B L O O D C O N C E N T R A T I O N S ...................................................................................................................................................I-1 1.2. N T P , 2 0 0 6 (C H O L A N G IO C A R C IN O M A S ): B A C K G R O U N D D O S E = M E A S U R E D T C D D C O N C E N T R A T I O N O N L Y ............................................................................... I - 4 1.3. N T P , 2 0 0 6 (C H O L A N G IO C A R C IN O M A S ): B A C K G R O U N D D O S E = M E A S U R E D T E Q C O N C E N T R A T I O N ( T C D D , P E C D F , A N D P C B - 1 2 6 ) ................... I - 7 1.4. N T P , 2 0 0 6 (C H O L A N G IO C A R C IN O M A S ): B A C K G R O U N D D O S E = 2 x M E A S U R E D T E Q C O N C E N T R A T I O N ( T C D D , P E C D F , A N D P C B - 1 2 6 ) ................ I - 1 0 1.5. N T P , 2 0 0 6 (C H O L A N G IO C A R C IN O M A S ): B A C K G R O U N D D O S E = 1 0* M E A S U R E D T C D D C O N C E N T R A T I O N .............................................................................................. I - 1 3 1 .6 . R E F E R E N C E ..................................................................................................................................................................I - 1 5 L IST O F F IG U R E S I - 1 . N T P , 2 0 0 6 : U n a d j u s t e d b l o o d c o n c e n t r a t i o n s ( c h o l a n g i o c a r c i n o m a s ) ........................................... I - 3 I-2. N T P , 2 0 0 6 (ch olan giocarcin om as): B ack grou n d d o se = m easured T C D D c o n c e n t r a t i o n o n l y ................................................................................................................................................................... I - 6 I-3. N T P , 2 0 0 6 (ch olan giocarcin om as): B ack grou n d d o se = m easured T E Q c o n c e n t r a t i o n ( T C D D , P e C D F , a n d P C B - 1 2 6 ) ................................................................................................. I - 9 I-4. N T P , 2 0 0 6 (ch olan giocarcin om as): B ack grou n d d o se = 2 * m easured T E Q c o n c e n t r a t i o n ( T C D D , P e C D F , a n d P C B - 1 2 6 ) .............................................................................................. I - 1 2 I-5. N T P , 2 0 0 6 (ch olan giocarcin om as): B ack grou n d d o se = 10* m easu red T C D D c o n c e n t r a t i o n ............................................................................................................................................................................ I - 1 5 This document is a draftfor review purposes only and does not constitute Agency policy. I-ii DRAFT--DO NOT CITE OR QUOTE 1 A P P E N D IX I. E F F E C T O F B A C K G R O U N D E X P O S U R E O N B E N C H M A R K -D O S E 2 M O D E L IN G 3 4 5 I.1 . N T P , 2 0 0 6 (C H O L A N G IO C A R C IN O M A S ): U N A D J U S T E D B L O O D 6 C O N C E N T R A T IO N S 7 8 9 Multistage Cancer Model. (Version: 1.7; Date: 05/16/2008) 1 0 Input Data File: C:\Usepa\BMDS21\Data\msc_NTP_2006_carcin_Setting.(d) 11 Gnuplot Plotting File: C:\Usepa\BMDS21\Data\msc_NTP_2006_carcin_Setting.plt 1 2 Wed Apr 14 12:59:57 2010 13 14 15 BMDS Model Run 16 17 1 8 The form of the probability function is: 19 2 0 P[response] = background + (1-background)*[1-EXP( 2 1 -betai*dose^1-beta2*dose^2-beta3*dose^3)] 22 2 3 The parameter betas are restricted to be positive 24 25 2 6 Dependent variable = cholang 2 7 Independent variable = bl_nom 28 2 9 Total number of observations = 6 3 0 Total number of records with missing values = 0 3 1 Total number of parameters in model = 4 3 2 Total number of specified parameters = 0 3 3 Degree of polynomial = 3 34 35 3 6 Maximum number of iterations = 250 3 7 Relative Function Convergence has been set to: 1e-008 3 8 Parameter Convergence has been set to: 1e-008 39 40 41 4 2 Default Initial Parameter Values 43 Background = 0 44 Beta(1) = 0 45 Beta(2) = 0 4 6 Beta(3)= 2.44609e-005 47 48 4 9 Asymptotic Correlation Matrix of Parameter Estimates 50 51 ( *** The model parameter(s) -Background -Beta(1) -Beta(2) 5 2 have been estimated at a boundary point, or have been specified by 5 3 the user, 5 4 and do not appear in the correlation matrix ) 55 5 6 Beta(3) 57 5 8 Beta(3) 1 59 60 This document is a draftfor review purposes only and does not constitute Agency policy. I-1 DRAFT--DO NOT CITE OR QUOTE 1 2 Parameter Estimates 3 4 95.0% Wald Confidence 5 Interval 6 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf 7 Limit 8 Background 9 Beta(1) 1 0 Beta(2) 11 Beta(3) 0 0 0 2.30992e-005 * * * * * * * * * * * * 12 13 * - Indicates that this value is not calculated. 14 15 16 1 7 Analysis of Deviance Table 18 19 Model Log(likelihood) # Param's Deviance Test d.f. P-value 2 0 Full model -55.408 6 2 1 Fitted model -55.7584 1 0.700706 5 0.9829 2 2 Reduced model -96.9934 1 83.1708 5 <.0001 23 24 AIC: 113.517 25 26 2 7 Goodness of Fit 2 8 Scaled 29 Dose Est. Prob. Expected Observed Size Residual 30 31 0.0000 0.0000 0.000 0.000 49 0.000 32 2.5600 0.0004 0.019 0.000 48 -0.136 33 5.6900 0.0042 0.195 0.000 46 -0.443 34 9.7900 0.0214 1.072 1.000 50 -0.070 35 16.6000 0.1003 4.913 4.000 49 -0.434 36 29.7000 0.4540 24.063 25.000 53 0.259 37 3 8 Chi^2 = 0.48 d.f. = 5 P-value = 0.9930 39 40 4 1 Benchmark Dose Computation 42 4 3 Specified effect = 0.01 44 4 5 Risk Type = Extra risk 46 4 7 Confidence level = 0.95 48 49 BMD = 7.57754 50 51 BMDL = 4.13907 52 53 BMDU = 8.42931 54 5 5 Taken together, (4.13907, 8.42931) is a 90 two-sided confidence 5 6 interval for the BMD 57 5 8 Multistage Cancer Slope Factor 0.002416 59 60 This document is a draftfor review purposes only and does not constitute Agency policy. 1-2 DRAFT--DO NOT CITE OR QUOTE Fraction Affected Multistage Cancer Model with 0.95 Confidence Level 1 12:59 04/14 2010 2 3 F igure I-1. N T P , 2006: U n ad ju sted blood concentrations 4 (cholangiocarcinom as). This document is a draftfor review purposes only and does not constitute Agency policy. I-3 DRAFT--DO NOT CITE OR QUOTE 1 I.2. N T P , 2006 ( C H O L A N G I O C A R C I N O M A S ) : B A C K G R O U N D D O S E = M E A S U R E D 2 TCDD CONCENTRATION ONLY 3 4 5 6 Multistage Cancer Model. (Version: 1.7; Date: 05/16/2008) 7 Input Data File: C:\Usepa\BMDS21\Data\msc_NTP_2006_carcin_Setting.(d) 8 Gnuplot Plotting File: C:\Usepa\BMDS21\Data\msc_NTP_2006_carcin_Setting.plt 9 Fri Apr 16 15:47:08 2010 10 11 1 2 BMDS Model Run 13 14 15 The form of the probability function is: 16 1 7 P[response] = background + (1-background)*[1-EXP( 18 -betai*dose^1-beta2*dose^2-beta3*dose^3)] 19 2 0 The parameter betas are restricted to be positive 21 22 2 3 Dependent variable = cholang 2 4 Independent variable = bl_TCDDadj 25 2 6 Total number of observations = 6 2 7 Total number of records with missing values = 0 2 8 Total number of parameters in model = 4 2 9 Total number of specified parameters = 0 3 0 Degree of polynomial = 3 31 32 3 3 Maximum number of iterations = 250 3 4 Relative Function Convergence has been set to: 1e-008 3 5 Parameter Convergence has been set to: 1e-008 36 37 38 3 9 Default Initial Parameter Values 40 Background = 0 41 Beta(1) = 0 42 Beta(2) = 0 4 3 Beta(3) = 2.43074e-005 44 45 4 6 Asymptotic Correlation Matrix of Parameter Estimates 47 48 ( *** The model parameter(s) -Background -Beta(1) -Beta(2) 4 9 have been estimated at a boundary point, or have been specified by 5 0 the user, 5 1 and do not appear in the correlation matrix ) 52 5 3 Beta(3) 54 5 5 Beta(3) 1 56 57 58 5 9 Parameter Estimates 60 This document is a draftfor review purposes only and does not constitute Agency policy. I-4 DRAFT--DO NOT CITE OR QUOTE 1 95.0% Wald Confidence 2 Interval 3 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf 4 Limit 5 Background 6 Beta(1) 7 Beta(2) 8 Beta(3) 0 0 0 2.29144e-005 * * * * * * * * * * * * 9 1 0 * - Indicates that this value is not calculated. 11 12 13 1 4 Analysis of Deviance Table 15 16 Model Log(likelihood) # Param's Deviance Test d.f. P-value 1 7 Full model -55.408 6 1 8 Fitted model -55.771 1 0.726 5 0.9815 1 9 Reduced model -96.9934 1 83.1708 5 <.0001 20 21 AIC: 113.542 22 23 2 4 Goodness of Fit 2 5 Scaled 26 Dose Est. Prob. Expected Observed Size Residual 27 28 0.0640 0.0000 0.000 0.000 49 -0.001 29 2.6240 0.0004 0.020 0.000 48 -0.141 30 5.7540 0.0044 0.200 0.000 46 -0.449 31 9.8540 0.0217 1.084 1.000 50 -0.082 32 16.6640 0.1006 4.930 4.000 49 -0.442 33 29.7640 0.4535 24.035 25.000 53 0.266 34 3 5 Chi^2 = 0.49 d.f. = 5 P-value = 0.9924 36 37 3 8 Benchmark Dose Computation 39 4 0 Specified effect = 0.01 41 4 2 Risk Type = Extra risk 43 4 4 Confidence level = 0.95 45 46 BMD = 7.59785 47 48 BMDL = 4.19355 49 50 BMDU = 8.45188 51 5 2 Taken together, (4.19355, 8.45188) is a 90 two-sided confidence 5 3 interval for the BMD 54 5 5 Multistage Cancer Slope Factor 0.00238461 56 57 This document is a draftfor review purposes only and does not constitute Agency policy. 1-5 DRAFT--DO NOT CITE OR QUOTE Fraction Affected Multistage Cancer Model with 0.95 Confidence Level 1 15:47 04/16 2010 2 F igure I-2. N T P , 2006 (cholangiocarcinom as): B ack grou n d dose = m easu red 3 T C D D con cen tration only. 4 This document is a draftfor review purposes only and does not constitute Agency policy. 1-6 DRAFT--DO NOT CITE OR QUOTE 1 I.3. N T P , 2006 ( C H O L A N G I O C A R C I N O M A S ) : B A C K G R O U N D D O S E = M E A S U R E D 2 T E Q C O N C E N T R A T I O N (TCDD, P E C D F , A N D PCB-126) 3 4 5 6 Multistage Cancer Model. (Version: 1.7; Date: 05/16/2008) 7 Input Data File: C:\Usepa\BMDS21\Data\msc_NTP_2006_carcin_Setting.(d) 8 Gnuplot Plotting File: C:\Usepa\BMDS21\Data\msc_NTP_2006_carcin_Setting.plt 9 Fri Apr 16 15:50:00 2010 10 11 1 2 BMDS Model Run 13 14 15 The form of the probability function is: 16 1 7 P[response] = background + (1-background)*[1-EXP( 18 -betai*dose^1-beta2*dose^2-beta3*dose^3)] 19 2 0 The parameter betas are restricted to be positive 21 22 2 3 Dependent variable = cholang 2 4 Independent variable = bl_TEQadj 25 2 6 Total number of observations = 6 2 7 Total number of records with missing values = 0 2 8 Total number of parameters in model = 4 2 9 Total number of specified parameters = 0 3 0 Degree of polynomial = 3 31 32 3 3 Maximum number of iterations = 250 3 4 Relative Function Convergence has been set to: 1e-008 3 5 Parameter Convergence has been set to: 1e-008 36 37 38 3 9 Default Initial Parameter Values 40 Background = 0 41 Beta(1) = 0 42 Beta(2) = 0 4 3 Beta(3) = 2.40088e-005 44 45 4 6 Asymptotic Correlation Matrix of Parameter Estimates 47 48 ( *** The model parameter(s) -Background -Beta(1) -Beta(2) 4 9 have been estimated at a boundary point, or have been specified by 5 0 the user, 5 1 and do not appear in the correlation matrix ) 52 5 3 Beta(3) 54 5 5 Beta(3) 1 56 57 58 5 9 Parameter Estimates 60 This document is a draftfor review purposes only and does not constitute Agency policy. 1-7 DRAFT--DO NOT CITE OR QUOTE 1 95.0% Wald Confidence 2 Interval 3 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf 4 Limit 5 Background 6 Beta(1) 7 Beta(2) 8 Beta(3) 0 0 0 2.25556e-005 * * * * * * * * * * * * 9 1 0 * - Indicates that this value is not calculated. 11 12 13 1 4 Analysis of Deviance Table 15 16 Model Log(likelihood) # Param's Deviance Test d.f. P-value 1 7 Full model -55.408 6 1 8 Fitted model -55.7969 1 0.777718 5 0.9784 1 9 Reduced model -96.9934 1 83.1708 5 <.0001 20 21 AIC: 113.594 22 23 2 4 Goodness of Fit 2 5 Scaled 26 Dose Est. Prob. Expected Observed Size Residual 27 28 0.1900 0.0000 0.000 0.000 49 -0.003 29 2.7500 0.0005 0.023 0.000 48 -0.150 30 5.8800 0.0046 0.210 0.000 46 -0.460 31 9.9800 0.0222 1.109 1.000 50 -0.104 32 16.7900 0.1013 4.962 4.000 49 -0.455 33 29.8900 0.4525 23.981 25.000 53 0.281 34 3 5 Chi^2 = 0.53 d.f. = 5 P-value = 0.9909 36 37 3 8 Benchmark Dose Computation 39 4 0 Specified effect = 0.01 41 4 2 Risk Type = Extra risk 43 4 4 Confidence level = 0.95 45 46 BMD = 7.63793 47 48 BMDL = 4.29872 49 50 BMDU = 8.4964 51 5 2 Taken together, (4.29872, 8.4964 ) is a 90 two-sided confidence 5 3 interval for the BMD 54 5 5 Multistage Cancer Slope Factor 0.00232627 56 57 This document is a draftfor review purposes only and does not constitute Agency policy. I-8 DRAFT--DO NOT CITE OR QUOTE Fraction Affected Multistage Cancer Model with 0.95 Confidence Level 1 15:50 04/16 2010 2 F ig u re I-3. N T P , 20 0 6 (ch o la n g io ca rcin o m a s): B a ck g ro u n d d o se = m ea su red 3 T E Q co n cen tra tio n (T C D D , P eC D F , an d P C B -1 2 6 ). This document is a draftfor review purposes only and does not constitute Agency policy. 1-9 DRAFT--DO NOT CITE OR QUOTE 1 I.4. N T P , 2006 ( C H O L A N G I O C A R C I N O M A S ) : B A C K G R O U N D D O S E = 2x 2 M E A S U R E D T E Q C O N C E N T R A T I O N (TCDD, PEC D F , A N D PCB-126) 3 4 5 6 Multistage Cancer Model. (Version: 1.7; Date: 05/16/2008) 7 Input Data File: C:\Usepa\BMDS21\Data\msc_NTP_2006_carcin_Setting.(d) 8 Gnuplot Plotting File: C:\Usepa\BMDS21\Data\msc_NTP_2006_carcin_Setting.plt 9 Fri Apr 16 15:51:30 2010 10 11 12 BMDS Model Run 13 14 15 The form of the probability function is: 16 17 P[response] = background + (1-background)*[1-EXP( 18 -betai*dose^1-beta2*dose^2-beta3*dose^3)] 19 2 0 The parameter betas are restricted to be positive 21 22 23 Dependent variable = cholang 2 4 Independent variable = bl_TEQ2x 25 2 6 Total number of observations = 6 2 7 Total number of records with missing values = 0 2 8 Total number of parameters in model = 4 2 9 Total number of specified parameters = 0 3 0 Degree of polynomial = 3 31 32 33 Maximum number of iterations = 250 3 4 Relative Function Convergence has been set to: 1e-008 35 Parameter Convergence has been set to: 1e-008 36 37 38 3 9 Default Initial Parameter Values 40 Background = 0 41 Beta(1) = 0 42 Beta(2) = 0 43 Beta(3) = 2.3568e-005 44 45 4 6 Asymptotic Correlation Matrix of Parameter Estimates 47 48 ( *** The model parameter(s) -Background -Beta(1) -Beta(2) 4 9 have been estimated at a boundary point, or have been specified by 5 0 the user, 51 and do not appear in the correlation matrix ) 52 53 Beta(3) 54 55 Beta(3) 1 56 57 58 59 Parameter Estimates 60 This document is a draftfor review purposes only and does not constitute Agency policy. I-10 DRAFT--DO NOT CITE OR QUOTE 1 95.0% Wald Confidence 2 Interval 3 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf 4 Limit 5 Background 6 Beta(1) 0* 0* * * * * 7 Beta(2) 0 8 Beta(3) 2.20268e-005 * * * * * * 9 10 * - Indicates that this value is not calculated. 11 12 13 14 Analysis of Deviance Table 15 16 Model Log(likelihood) # Param's Deviance Test d.f. P-value 17 Full model -55.408 6 18 Fitted model -55.8382 1 0.860456 5 0.973 19 Reduced model -96.9934 1 83.1708 5 <.0001 20 21 AIC: 113.676 22 23 2 4 Goodness of Fit 25 Scaled 26 Dose Est. Prob. Expected Observed Size Residual 27 28 0.3800 0.0000 0.000 0.000 49 -0.008 29 2.9400 0.0006 0.027 0.000 48 -0.164 30 6.0700 0.0049 0.226 0.000 46 -0.477 31 10.1700 0.0229 1.145 1.000 50 -0.137 32 16.9800 0.1022 5.009 4.000 49 -0.476 33 30.0800 0.4509 23.898 25.000 53 0.304 34 35 Chi^2 = 0.59 d.f. = 5 P-value = 0.9884 36 37 3 8 Benchmark Dose Computation 39 4 0 Specified effect = 0.01 41 4 2 Risk Type = Extra risk 43 4 4 Confidence level = 0.95 45 46 BMD = 7.69856 47 48 BMDL = 4.45212 49 50 BMDU = 8.56376 51 5 2 Taken together, (4.45212, 8.56376) is a 90 % two-sided confidence 53 interval for the BMD 54 55 Multistage Cancer Slope Factor 0.00224612 56 This document is a draftfor review purposes only and does not constitute Agency policy. I-11 DRAFT--DO NOT CITE OR QUOTE Multistage Cancer Model with 0.95 Confidence Level Fraction Affected 1 15:51 04/16 2010 dose 2 3 F ig u re I-4. N T P , 20 0 6 (ch o la n g io ca rcin o m a s): B a ck g ro u n d d o se = 2 x 4 m ea su red T E Q co n cen tra tio n (T C D D , P eC D F , an d P C B -1 2 6 ). This document is a draftfor review purposes only and does not constitute Agency policy. I-12 DRAFT--DO NOT CITE OR QUOTE 1 I.5. N T P , 2006 ( C H O L A N G I O C A R C I N O M A S ) : B A C K G R O U N D D O S E = 10x 2 MEASURED TCDD CONCENTRATION 3 4 5 6 Multistage Cancer Model. (Version: 1.7; Date: 05/16/2008) 7 Input Data File: C:\Usepa\BMDS21\Data\msc_NTP_2006_carcin_Setting.(d) 8 Gnuplot Plotting File: C:\Usepa\BMDS21\Data\msc_NTP_2006_carcin_Setting.plt 9 Fri Apr 16 15:55:37 2010 10 11 12 BMDS Model Run 13 14 15 The form of the probability function is: 16 17 P[response] = background + (1-background)*[1-EXP( 18 -betai*dose^1-beta2*dose^2-beta3*dose^3)] 19 2 0 The parameter betas are restricted to be positive 21 22 23 Dependent variable = cholang 2 4 Independent variable = bl_TEQmax 25 2 6 Total number of observations = 6 2 7 Total number of records with missing values = 0 2 8 Total number of parameters in model = 4 2 9 Total number of specified parameters = 0 3 0 Degree of polynomial = 3 31 32 33 Maximum number of iterations = 250 3 4 Relative Function Convergence has been set to: 1e-008 35 Parameter Convergence has been set to: 1e-008 36 37 38 3 9 Default Initial Parameter Values 40 Background = 0 41 Beta(1) = 0 42 Beta(2) = 0 43 Beta(3) = 2.29823e-005 44 45 4 6 Asymptotic Correlation Matrix of Parameter Estimates 47 48 ( *** The model parameter(s) -Background -Beta(1) -Beta(2) 4 9 have been estimated at a boundary point, or have been specified by 5 0 the user, 51 and do not appear in the correlation matrix ) 52 53 Beta(3) 54 55 Beta(3) 1 56 57 58 59 Parameter Estimates 60 This document is a draftfor review purposes only and does not constitute Agency policy. I-13 DRAFT--DO NOT CITE OR QUOTE 1 95.0% Wald Confidence 2 Interval 3 Variable Estimate Std. Err. Lower Conf. Limit Upper Conf 4 Limit 5 Background 6 Beta(1) 0* 0* * * * * 7 Beta(2) 0 8 Beta(3) 2.13264e-005 * * * * * * 9 10 * - Indicates that this value is not calculated. 11 12 13 14 Analysis of Deviance Table 15 16 Model Log(likelihood) # Param's Deviance Test d.f. P-value 17 Full model -55.408 6 18 Fitted model -55.8994 1 0.982747 5 0.9639 19 Reduced model -96.9934 1 83.1708 5 <.0001 20 21 AIC: 113.799 22 23 2 4 Goodness of Fit 25 Scaled 26 Dose Est. Prob. Expected Observed Size Residual 27 28 0.6400 0.0000 0.000 0.000 49 -0.017 29 3.2000 0.0007 0.034 0.000 48 -0.183 30 6.3300 0.0054 0.248 0.000 46 -0.499 31 10.4300 0.0239 1.195 1.000 50 -0.181 32 17.2400 0.1035 5.072 4.000 49 -0.503 33 30.3400 0.4488 23.785 25.000 53 0.336 34 35 Chi^2 = 0.68 d.f. = 5 P-value = 0.9840 36 37 3 8 Benchmark Dose Computation 39 4 0 Specified effect = 0.01 41 4 2 Risk Type = Extra risk 43 4 4 Confidence level = 0.95 45 46 BMD = 7.78193 47 48 BMDL = 4.65224 49 50 BMDU = 8.65638 51 5 2 Taken together, (4.65224, 8.65638) is a 90 % two-sided confidence 53 interval for the BMD 54 55 Multistage Cancer Slope Factor 0.0021495 56 57 This document is a draftfor review purposes only and does not constitute Agency policy. I-14 DRAFT--DO NOT CITE OR QUOTE Multistage Cancer Model with 0.95 Confidence Level 1 15:55 04/16 2010 2 3 F ig u re I-5. N T P , 20 0 6 (ch o la n g io ca rcin o m a s): B a ck g ro u n d d o se = 10* 4 m ea su red T C D D co n cen tra tio n . 5 I.6 . R E F E R E N C E 6 NTP (National Toxicology Program). (2006a) NTP technical report on the toxicology and carcinogenesis studies o f 7 2,3,7,8-tetrachlorodibenzo-p-dioxin (T C D D ) (C A S N o. 1746-01-6) in fem ale Harlan Sprague-D aw ley rats (Gavage 8 Studies). Natl T oxicol ProgramTech Rep 521. Public Health Service, N ational Institute o f Health, U .S. Department 9 o f H ealth and Human Services, R esearch Triangle Park, NC. This document is a draftfor review purposes only and does not constitute Agency policy. I-15 DRAFT--DO NOT CITE OR QUOTE