Document OEbdVqwGVVEyVZy69KyoMn5LX

Electronically Served 9/22/2017 10:33 PM Hennepin County, MN 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN STATE OF MINNESOTA DISTRICT COURT FOR THE COUNTY OF HENNEPIN FOURTH JUDICIAL DISTRICT CIVIL ACTION NO. 27-CV-10-28862 STATE OF MINNESOTA, e t al. v. 3M COMPANY EXPERT REPORT OF PHILIPPE GRANDJEAN, MD, DMSc PREPARED ON BEHALF OF PLAINTIFF STATE OF MINNESOTA 22 September, 2017 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN TABLE OF CONTENTS I. INTRODUCTION.........................................................................................................1 A. Qualifications.......................................................................................................1 B. Materials relied upon.......................................................................................... 2 C. Exhibits............................................................................................................... 3 D. Updates and reservation...................................................................................... 3 E. Compensation..................................................................................................... 3 F. Previous service as expert at deposition or trial during last 4 years................... 4 II. SUMMARY OF OPINIONS....................................................................................... 4 III. BACKGROUND ON PFC PRODUCTION AND CONTAMINATION................... 5 A. An abbreviated history of PFC production......................................................... 5 B. Widespread, persistent environmental contamination........................................ 5 C. Environmental contamination in Minnesota....................................................... 7 D. Drinking water contamination............................................................................ 8 E. Other contaminated sources affecting humans..................................................10 IV. HUMAN EXPOSURE TO PERFLUORINATED COMPOUNDS...........................13 A. Early knowledge of human exposure................................................................13 B. Detection and distribution of PFCs in humans..................................................16 C. Serum analyses from Minnesota residents........................................................18 D. Serum analyses from other populations.............................................................19 E. Conclusions on PFC exposure in Minnesota residents..................................... 20 V. METHODOLOGICAL ISSUES IN RISK EVALUATION...................................... 21 A. Interpretation of epidemiology studies............................................................. 21 B. Toxicity and interpretation of data................................................................... 24 VI. HEALTH EFFECTS FROM PERFLUORINATED COMPOUNDS....................... 25 A. Early warning signs.......................................................................................... 25 B. Early epidemiology studies (1970s to 1990s)................................................... 26 C. Early toxicology studies (1970s to 1990s)........................................................ 30 D. Further 3M-influenced research (1990s to early 2000s)................................... 31 E. Growth of PFC research................................................................................... 34 F. Public knowledge on PFCs............................................................................... 35 G. Recent key reviews, studies and resources ....................................................... 37 VII. ADVERSE HEALTH EFFECTS AT INDIVIDUAL ENDPOINTS........................ 39 A. Immunotoxicity and autoimmunity .................................................................. 39 B. Reproductive toxicity........................................................................................ 49 C. Endocrine disruption......................................................................................... 57 D. Thyroid hormones and related diseases ............................................................ 62 E. Insulin and diabetes .......................................................................................... 64 F. Neurobehavioral functions................................................................................ 67 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN G. Liver toxicity.................................................................................................... 69 H. Risk factors for cardiovascular disease............................................................. 74 I. Carcinogenicity................................................................................................. 77 V III. PFHXS, PFBA, AND OTHER SHORT-CHAIN PFCS............................................ 87 IX. RISK ASSESSMENTS AND CURRENT LIMITS FOR PFC EXPOSURE............ 90 A. Drinking water limits........................................................................................ 91 B. Setting drinking water health limits.................................................................. 95 C. Applications of benchmark dose calculations.................................................. 97 D. Comparison of existing limits with recent research information...................... 98 E. Proposed limits based on recent epidemiological evidence............................. 99 F. Underestimation of hazards posed by PFCs....................................................100 X. AFFIRMATION........................................................................................................101 EXHIBIT A ABBREVIATIONS.......................................................................................102 EXHIBIT B PHILIPPE GRANDJEAN, M.D. CV.........................................................104 EXHIBIT C LIST OF GRANDJEAN PUBLICATIONS FROM RECENT 10 YEARS .109 EXHIBIT D CITED PUBLICATIONS.............................................................................123 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN I. INTRODUCTION My name is Philippe Grandjean. I have been asked by counsel for the State of Minnesota to provide, from an epidemiological perspective, an evaluation of the human health risks associated with environmental PFCa contamination from 3M's manufacturing and disposal operations in Minnesota. A. Qualifications I earned my M.D. and D.M.Sc. degrees from the University of Copenhagen, Denmark, in 1974 and 1978, respectively. I serve as Adjunct Professor of Environmental Health at the Harvard School of Public Health (since 2003) and as Professor and Chair of Environmental Medicine at the University of Southern Denmark (since 1982). I previously served for a brief period (1980-1982) as the Director of the Department of Occupational Medicine at the Danish National Institute of Occupational Health. Former positions in the U.S. include Research Fellow and Senior Fulbright Scholar, Mount Sinai School of Medicine in New York (1978-1979), and Adjunct Professor of Neurology and Environmental Health, Boston University Schools of Medicine and Public Health (1994-2002). As part of my employment as a civil servant in Denmark, I have served for more than 30 years as the Consultant in Toxicology to the Danish Health Authority. In the latter capacity, I have reviewed and commented on case reports, research studies, and proposed regulations on environmental chemicals. I also serve on the Scientific Committee of the European Environment Agency (EEA) and on the European Advisory Committee on Health Research of the World Health Organization (WHO). My research in environmental epidemiology focuses on the health effects of exposures to environmental chemicals, including perfluorinated alkylate substances (PFCs or PFASs).aMost of my efforts have concentrated on the effects of environmental pollutants on early human development. This research has been financed by public sources, mainly the National Institutes of Health and other U.S. agencies, the European Commission, and the Danish Research Agency. I have published about 500 scientific papers, of which most are research articles in international scientific journals with peer review. My h-index in the Web of Science data base is greater than 60. Seven of my articles published in the last 10 years have earned the attribute "Highly Cited Paper," i.e., they received enough citations to place them in the top 1% of published papers in the field. This list includes an article on PFAS immunotoxicity published in the Journal of the American Medical Association (JAMA) in 2012. I have also authored or edited 20 books, including textbooks on environmental health and risk assessment. a Terminology concerning perfluorinated compounds has evolved. The term PFC had been used to refer generally to perfluorinated compounds. Recently, the scientific community has with more precision started to settle on the term PFAS to refer to the narrower family of perfluorinated alkylate substances, which family includes PFOA, PFOS, PFBA and most of the other chemicals relevant in this matter. For ease of reference, and for consistency with some of the historical documentation in this case, I will, for the most part, use the term PFCs to refer to the family of PFASs, and will use more specific terms when the situation warrants. A table of the abbreviations I use in this report is attached as Exhibit A. 1 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN I am regularly invited as speaker at international conferences and other scientific events. Regarding PFCs, I was invited to give a special presentation at the meeting of the (U.S.) National Advisory Environmental Health Sciences Council (at the National Institute of Environmental Health Sciences) in 2012, and also that year at a meeting of the Emerging Chemicals Workgroup, U.S. Environmental Protection Agency (EPA). Both presentations were on the immunotoxicity of PFCs. In the fall of 2016, I was invited to give a special presentation on PFOA at the committee meeting of the United Nations Stockholm Convention. I am (Founding) Editor-in-Chief of the open-access scientific journal, Environmental Health (since 2002), which ranks among the upper 25% ofjournals in the field. I also serve or have served on editorial boards of about a dozen journals within medicine, environmental science, and toxicology. As editor and as reviewer for other major journals, I frequently evaluate manuscripts on environmental epidemiology and toxicology. I have served on, sometimes chaired, or acted as rapporteur for, expert committees under the auspices of the WHO, the International Agency for Research on Cancer (IARC), EPA, the European Commission, the European Food Safety Authority (EFSA), and other organizations. During my six-year membership of an EFSA expert panel, I participated in developing the opinion on `Perfluorooctane sulfonate (PFOS), perfluorooctanoic acid (PFOA) and their salts' [1] and the `Guidance of the Scientific Committee on Use of the benchmark dose approach in risk assessment' [2]. EFSA is currently developing updated opinions on PFCs, and I have been recently approached by EFSA and invited to serve as ad hoc expert to help finalize these assessments. I have previously served as an expert witness in the U.S. regarding mercury pollution. In that regard, I wrote an expert report and testified in deposition and at trial in the case Maine People's Alliance v. Holtrachem Manufacturing Co. in 2002 and again in 2014. I also wrote an expert report and testified in deposition at the request of the U.S. Department of Justice, regarding pollution from coal-fired power plants, in 2008 (United States et al., v. Cinergy Corp et al., District Court for the Southern District of Indiana, Indianapolis Division) and later contributed to an amicus statement to the appeals court. I also served as expert witness at the request by a Danish lower court (1999) and subsequently by the superior court (2007 and 2011) regarding adverse health effects of indoor use of a wood treatment product containing a pesticide. A copy of my most recent CV is attached as Exhibit B. A list of publications which I authored or co-authored, including those of the past 10 years, is attached as Exhibit C. B. Materials relied upon For the purposes of this report, I have relied in part on my own epidemiological research and publications concerning PFCs. I also have reviewed the epidemiological literature concerning studies by others on the human health risks associated with exposure to PFCs. In this report, citations to studies are in brackets in the text, [x], with these endnote references listed in Exhibit D. Among other sources of information, such as the reports from the C8 panel [3], I have relied upon the most recent version of ATSDR's draft ToxProfile [4], the evaluation of 2 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN immunotoxicity b y the N T P [5], the assessment of carcinogenicity b y I A R C [6], and recent reviews [7-10]. In addition to epidemiological studies, I have considered certain supporting toxicological information from laboratory animal studies and in vitro models, but I do not necessarily provide a complete review of all such supporting evidence, as m y focus is h u m a n health. Wh i l e summarizing available documentation, I also outline the emergence over time of the know le dg e on h u m a n P F C exposures and associated risks. I have had access to reports of certain studies wh i c h w e re commissioned or sponsored b y 3 M , to documents filed b y 3 M with the E P A , and to documents provided in connection with the discovery phase of the court proceedings. I c o m m e n t as to certain of the 3 M documents, particularly those wh i c h reflect epidemiological or toxicological studies, from a scientific perspective. A s part of m y report, based on 3 M documents available to me, I characterize h o w 3 M 's kn ow le dg e on h u m a n health effects of P F C s has evolved, so that I can assess the extent to w h i c h these risks m a y have been underestimated or disregarded in the past. I reference documents produced b y 3 M in discovery, as I had access to the discovery record, but they are exceedingly n u m e r o u s and I have not personally reviewed all of them. Wh i l e preferentially citing studies published in peer-reviewed journals, and relying on m y o w n research experience, I cite the additional sources that I consider most relevant. Published literature is referred to b y nu m b e r s in square brackets, and the unpublished materials are referred to b y superscript letters listed in the footnotes. C. Exhibits I m a y use as exhibits part or all of any of the documents or papers cited in this report including this report itself; graphs or tables d r a w n from data in any of those documents or papers; any d o c u m e n t helpful as foundation for or illustration of m y testimony; any do cu m e n t considered or cited or relied u p o n b y any other expert in this case; any d o c u m e n t used as an exhibit or produced in discovery b y any party or non-party or expert; or any d o c u m e n t needed to respond to or rebut testimony b y any witness including any other expert. D. Updates and reservation T h e opinions expressed in this report are m y o w n and are based on the data, documents, and facts available to m e at the time of writing. Should additional relevant or pertinent information b e c o m e available, I reserve the right to supplement the discussion and findings in m y report. I also reserve the right to respond to any opinions on similar topics by other experts in this matter, and to respond to any criticism or c o m m e n t on m y opinions. E. Compensation I a m being compensated at the rate of $250 per hour for m y time, wh i c h is m y customary rate for matters of this type. M y compensation does not depend in any w a y on the content of m y opinions. 3 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN F. Previous service as expert at deposition or trial during last 4 years Maine People's Alliance and Natural Resources Defense Council, Inc. v. Holtrachem Manufacturing Company, LLC and Mallinckrodt US LLC, Case No. 1:00-cv-00069JAW, United States District Court for the District of Maine. II. SUMMARY OF OPINIONS The major opinions expressed in this report, which are described below in more detail along with more minor and supportive opinions, can be generally summarized as follows: PFCs pose a substantial present and potential hazard to human health. In particular, it is my opinion, based on the weight of the epidemiological evidence and supporting toxicity evidence, that PFCs pose a substantial present and potential hazard to at least human immune system functions, reproductive functions including adverse effects to the next generation, endocrine functions, thyroid functions, liver functions, and cardiovascular functions, and by causing or increasing the risk of cancer. Both PFOA and PFOS show convincing associations with these outcomes. This opinion mainly relies on available evidence on PFOS and PFOA, with somewhat less documentation on PFHxS. To date, PFBA has been less studied than certain PFCs such as PFOA and PFOS, in part because of PFBA's short apparent elimination half-life (T'A) in blood serum. However, a recent study shows that PFBA is retained significantly in human kidneys and lungs, rather than blood, and serum concentrations are therefore not reliable indicators of retention in body organs. Moreover, even early 3M-sponsored studies showed adverse effects from PFBA in laboratory animals in the liver, thyroid system, cholesterol levels, and negative developmental effects on the next generation. Given the similar structure of PFBA to other more-studied PFCs particularly at the end group of the chain, given the adverse human health effects that have been shown as to PFOA and PFOS, and given the adverse effects already shown as to PFBA, it is my opinion that PFBA too poses a substantial present and potential hazard to human health. Adverse health effects have been documented in epidemiological studies at background exposure levels. Recent and currently deveoping scientific insight is likely to justify more protective drinking water limits in the future. Thus, PFC contamination of drinking water within current limits may not be considered safe in the future. Adverse effects of PFCs were identified in animal studies commissioned by 3M as far back as the 1970s, but the company did not pursue these findings to characterize the nature of the hazards. Early studies of worker health inappropriately sought to explain away any possible associations with ill health. Minutes were edited to remove language that suggested the presence of a health hazard. Further, documentation of the transfer of PFCs via maternal milk was not revealed to the public. By 2000, when 3M announced the phase-out of PFC production, it announced that "the presence of these materials at these very low levels does not pose a human health or environmental risk," but this statement4 4 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN was at odds with the information already known to 3M. As late as 2016, a 3M Medical Director made a similar misleading statement.bIndependent research more recently, and at a substantial delay, began to examine adverse effects of PFCs, and risks to human health have been identified at very low exposure levels. III. BACKGROUND ON PFC PRODUCTION AND CONTAMINATION A. An abbreviated history of PFC production By way of brief background, my understanding of the history of PFC production is basically as follows. Perfluorinated compounds (PFCs) have a wide range of applications, and they have been in use for over 60 years. PFCs were first manufactured by 3M in Cottage Grove, Washington County in Minnesota, in 1947 or shortly thereafter. Perfluorooctane sulfonyl fluoride (POSF) was a main product for producing other PFC compounds, especially PFOA, but PFOS, PFBA and other PFCs were also produced, along with a variety of branched isomers, mostly with 4 to 9 carbons. Some of the PFC products (such as FOSA) degrade to PFOS. Global environmental dissemination became publicly known by about 2000 (but as is mentioned below was known by 3M many years earlier). A phase-out of commercial PFOS and PFOA production from the end of 2002 was announced by 3M in 2000. The academic research community (outside 3M) became aware of the environmental dissemination of PFCs following reports on world-wide dissemination and the legal proceedings regarding the contamination of the Upper Ohio River Valley. These experiences inspired an increased scientific interest in PFCs, especially during the most recent 10 years, as revealed by a growing number of scholarly papers published on PFC contamination, exposures and adverse effects in experimental models and epidemiological studies. B. Widespread, persistent environmental contamination Also by way of background, I outline my basic understanding of PFC environmental contamination. Major physicochemical properties of PFCs were characterized in publications as early as 1951 [11]. Standard chemical handbooks listed PFC vapor pressures and water solubilities at least by the 1970s (although they may not have been accurate). Many PFCs (or their salts or precursors) are somewhat water soluble and can potentially leach through soil to reach the groundwater, while some compounds have a sufficient vapor pressure to allow their dissemination via the atmosphere. These properties must have been known at 3M in the 1970s or earlier than that.5 b 3M_MN04422606. 3M plays down impact of toxic chemicals affecting Williamtown water. ABC, 15 March, 2016. 5 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN PFCs do not occur naturally. Many PFCs show high thermal, chemical and apparent biological inertness, properties that made them useful for certain industrial purposes, though at the same time also rendering these compounds an environmental hazard due to the potential for persistence and bioaccumulation [12]. The carbon-fluorine bond is strong, thus making the PFCs virtually indestructible in the environment and in the human body. However, as a 3M-supported report claimed in late 1997, organofluorine compounds are "generally viewed as recalcitrant because of their lack of chemical reactivity" [13]. Although most PFCs are oleophobic and therefore do not accumulate in fatty tissues (in contrast to dioxin and other persistent halogenated compounds), especially PFOS is now known to bioaccumulate in aquatic and marine food chains. In collaboration with researchers at Michigan State University, 3M's Robert Howell in 1997 reported on the persistence of organofluorine compounds in the environment [13]. Accordingly, and belatedly, a 3M memo from 1998 concludes that "we must begin to capture and incinerate the waste water process streams containing residuals as soon as possible."c Internal 3M reports from 1998 (apparently not published) also confirmed the occurrence of PFOS in bald eagles from Minnesota and Michigan,d and in marine food chains involving seals, sea lions, killer whales and porpoises.6As a major, and sometimes sole producer of PFCs, 3M and its products would clearly be suspected as a source of this contamination, whether close to or far from the production facilities. Long-range aqueous transport -- e.g., via the Mississippi River, groundwater and ocean currents -- allows PFCs in their soluble anionic forms to reach remote locations. In addition, several precursor compounds can be metabolized to PFOA and PFOS in the environment or in humans. The global dissemination of PFCs also occurs through atmospheric transport of volatile precursors, which are metabolized into the more persistent PFCs that are then deposited. Pollution of terrestrial and aquatic food chains resulted in high concentrations of PFOS in vulnerable species even in remote ecosystems, including, for example, polar bears [14, 15]. In late 1998, 3M environmental scientist Richard Purdy wrote to his 3M colleague, Georjean Adams, and suggested that his food chain risk assessment demonstrated a significant risk that should not be kept confidential. In another email, Purdy referred to a plan to examine environmental dissemination and then concluded: "For 20 years the division has been stalling the collection of data needed for evaluating the environmental impact of fluorochemicals. PFOS is the most onerous pollutant since PCB and you want to avoid collecting data that indicates that it is probably worse."g c 3MA00905854. Accumulation and Disposal of Waste Water at Decatur. page 3MA00905854. d3MA01402884. Occurrence of PFOS in wildlife (Part 1/ Eagles and Albatrosses), page 3ma01402885. e 3MA00468422. Pioneer Food Chain Risk Assessment of PFOS, page 3MA00468422. f 3M_MN00053763. Risk to the Environment Due to the Presence of POFS. page 3M_MN00053763. I have not determined if and when such reporting took place. g 3MA01373218.pdf. Page 03 166322. 6 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN On the basis of these studies and other evidence, the concerns about ignored environmental risks made Dr. Purdy seek to resign in 1999 due to his "profound disappointment in 3M's handling of the environmental risks associated with the manufacture and use of PFOS and its precursors, such as EtFOSE and MeFOSE)."h In his apparent letter of resignation, Dr. Purdy said: "3M told those of us working on the fluorochemical project not to write down our thoughts or have email discussions on issues because of how our speculations could be viewed in a legal discovery process. This has stymied intellectual development on the issue, and stifled discussion on the serious ethical implications of decisions." Dr. Purdy had worked on PFCs at least from 1993, but apparently, none of his work was published, given that I could locate no relevant scholarly publications under his name. He did give a presentation at the SETAC 2001 conference, where he discussed environmental risks from PFCs, and notes prepared on behalf of 3M challenged his presentation and suggested ways in which 3M should counter Dr. Purdy's conclusions in regard to environmental toxicity of PFOS and PFOA.j Dr. Purdy called PFOS the most "insidious pollutant since PCB" (i.e., polychlorinated biphenyls, industrial chemicals that were banned in the US in 1979 and included among the "dirty dozen" in 2001 when the Stockholm Convention was launched by the United Nations). Dr. Purdy also noted about PFOS that "it is probably more damaging than PCB because it does not degrade." He described his experience at 3M: "At almost every step, I have been assured that action will be taken - yet I see slow or no results. I am told the company is concerned, but their actions speak to different concerns than mine. I can no longer participate in the process that 3M has established for the management of PFOS and precursors. For me it is unethical to be concerned with markets, legal defensibility and image over environmental safety." A summary of 3M's food-chain findings was submitted to the EPA in 2003.k C. Environmental contamination in Minnesota I understand that, not surprisingly, PFC contamination has occurred in the environment remotely from the production site and at particularly elevated concentrations in areas proximate to 3M's facilities and disposal sites. The PFC releases have affected not only 3M workers but also many residents, especially in the East Metro area and nearby communities. I defer to other experts for the State of Minnesota on the extent and levels of the contamination, but my working understanding is summarized in part as follows: For an unknown period - beginning decades ago - residents in the East Metro area and nearby communities have consumed drinking water and, quite likely, freshwater fish and farm products contaminated with PFCs. The total number of residents exposed to this pollution over time likely numbers tens of thousands. A systematic study from 2008 showed that serum PFOA concentrations in residents averaged about 10-fold above national averages, and individual results ranged up to 100 fold above the national average. As this study was designed to be representative of the exposed local population, a sizable number of current and past Minnesota residents in the h 1999.03.28 - 3MA10065677.pdf 13MA10065677. Resignation letter from Richard Purdy to 3M. page 3MA10065678. j 3MA02600383.pdf. SETAC - PFOS Report. November 15, 2001. kAR226-1486. Environmental and health assessment of perfluorooctane sulfonic acid and its salts. 7 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN East Metro area have been exposed to very substantial amounts of PFOA and related PFCs. Serum concentrations are known to decrease only slowly. Background serum concentrations are now decreasing, and a 2010 re-examination of residents of affected communities confirmed this trend [16]. However, the cumulated PFC burdens in past and present residents will remain above national levels for many years to come and will depend on the extent to which continued exposures are minimized. D. Drinking water contamination 1. PFC disposal practices It is my understanding that in Minnesota, waste from the 3M production processes was disposed both at on- and off-site locations. Sources of contamination include 3M's Cottage Grove manufacturing facility. Disposal facilities used by 3M from about 1956 onward included the Oakdale Disposal Site, the Woodbury Disposal Site, and the Washington County Landfill. It is likely that some of this waste contained PFOS, PFOA and PFBA as well as precursor substances/ In 1966, DuPont, which purchased PFCs from 3M to make Teflon, among other things, decided that wet Teflon scrap should no longer go to a municipal land-fill in the Upper Ohio River Valley but instead should be kept on the plant for "disposal at sea at a future date."m This decision appears to reflect an early recognition of the risk of environmental discharge. 2. PFCs discovered in drinking water sources It is my understanding that, after the disposition of PFC waste became known at the Minnesota Pollution Control Agency (MPCA), a method for chemical analysis of water for PFOA and PFOS was established. Soon after that, PFCs were identified in samples from nearby groundwater monitoring wells and in soil samples. Subsequently, PFC contamination was found in the groundwater further away in both Washington and Dakota Counties. In Oakdale, the average PFOA concentration in the municipal water was 0.57 pg/L (or 570 ng/L), i.e. before the special water treatment was built in 2006. Much higher levels were found close to the production facility, with maximum measured water concentrations in groundwater wells at Cottage Grove of 619 pg/L (PFOA), 318 pg/L (PFBA), 26 pg/L (PFOS), and 40 pg/L (PFHxS) [17]. One pg (microgram) is one millionth of a gram, and this unit is one thousand times greater than 1 ng (nanogram), which is one billionth of a gram. As a general tendency, lower concentrations were found at greater distances from the sources. For comparison, the current EPA limits for PFOS and PFOA in water are 0.07 pg/L, and the 2009 MDH Health Risk Limits (HRLs) for PFOA and PFOS in water were both 0.3 pg/L. In 2011, MDH developed limits for perfluorobutane sulfonate (PFBS) and l 3M_MN03521906. FC-PFOA-PFOS East Metro. mAR226.1445, memo from J.E. Higginbothaim, 10/28/66, p. 00079 8 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN perfluorobutyrate (PFBA) at 7 pg/L. In May of 2017, MDH released its updated limits of 0.027 pg/L and 0.035 pg/L for PFOS and PFOA, respectively. While still relying on the EPA risk evaluation, the lowered limits were considered to better protect the fetus and the breastfed infant. Most efforts to date to characterize the extent of the pollution in Minnesota have focused on drinking water exceeding the MDH HRLs for PFOA and PFOS, as complemented by biomonitoring studies of PFC concentrations in serum of residents consuming this water. However, as I discuss and evaluate below in the section on exposure limits, the available information on water contamination needs to be considered in light of what can be reasonably assumed, given the current insight into the health impact from exposures to PFC contamination even at background levels. Although some older water analyses of questionable quality have been identified [18], it is not known precisely when the groundwater contamination began, but I understand that the plume of contaminated ground water likely reached private wells and community water supplies decades ago. The State has determined that, by now, over 100 square miles of groundwater have been contaminated by 3M's disposal of PFCs, and the source of residential drinking water for tens of thousands of Minnesotans has thus been affected by the PFC waste disposal. The area of contamination includes four major aquifers; namely, the St. Peter, Prairie du Chien, Jordan, and Franconia aquifers. These four aquifers serve as the sole source of drinking water for approximately 125,000 or more Minnesotans who reside in the Twin Cities area. Four of eight Oakdale municipal wells were discovered to be contaminated with PFOA and PFOS above MDH 2017 Health Based Values for drinking water, thus affecting up to 6,655 households supplied with the municipal water. Also, several hundred private wells in nearby communities were found to be contaminated. In 2008, the estimated population in Oakdale was 27,249 in 10,803 households in 2006, and Lake Elmo had 7,695 inhabitants in 2,738 households [18]. Thus, a very substantial population has apparently been exposed to the PFC contamination. As PFCs are not removed from drinking water by standard treatment processes, a large granular activated carbon system was installed on the two most contaminated wells in Oakdale, and these wells were then used preferentially. This system is known to remove longchained PFCs, while other PFCs may be more difficult to remove [19]. After lowering the limits for PFOA and PFOS, MDH announced that approximately 120 private wells, primarily in Lake Elmo and Cottage Grove, were known to be contaminated with PFOA or PFOS above the 2017 MDH limits." As the water contamination likely began many years ago, the exposed population also includes former residents. Current residents have also been exposed at levels MDH does not consider protective, because they did not have an alternative water supply provided until limits were lowered again in 2017. In short, the population exposed to hazardous water contamination is therefore larger than previously assessed. nhttp://www health.state.mn.us/news/pressrel/2017/water052317.html 9 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN PFBA contamination also was and is present in Minnesota drinking water sources. Indeed, it may be more widespread due to the greater mobility of this compound, and because this PFC seems to pass carbon filters more easily. PFBA water contamination has been discovered in Cottage Grove, St. Paul Park, Newport, Woodbury, Hastings and South St. Paul in 2006-2007 [18]. In a parallel setting in the Upper Ohio River Valley (West Virginia and Ohio), drinking water wells as far as 20 miles away were contaminated by releases from a Dupont industrial facility using PFCs from 3M. PFOA was detected in public water supply wells in this vicinity at levels up to 44 pg/L and in private wells up to over 13 pg/L [17]. Further in this report, the extensive health data recently collected by the C8 Panel from the Upper Ohio River Valley community [3]will feature prominently. E. Other contaminated sources affecting humans 1. Fish I understand that effluents from the 3M Cottage Grove facility and from the off site waste deposit sites led to discharges into the Mississippi River and some Minnesota lakes [18]. State and EPA scientists have documented PFC pollution of Lake Elmo and approximately 139 miles of the Mississippi River from St. Anthony Falls in Minneapolis (Hennepin County) downstream to the La Moille Dam (Lock and Dam No. 6), south of Winona. I have not searched the 3M documents to determine when 3M became aware of the PFC contamination of fish. PFC-contamination of waterways results in biomagnification in aqueous food chains, with PFOS apparently causing greater accumulation than PFOA or PFHxS PFHxS [20 22]. Increased PFC concentrations have been found in fish in the greater metropolitan area and in Mississippi River Pool 2. For example, in bluegill filets from fish samples taken in Minnesota from Lake Calhoun, and from four downstream locations on the Mississippi River, median PFOS concentrations ranged between 50 and 275 ng/g [22]. In another study of fish from Mississippi Pool 2, concentrations in bluegill, freshwater drum, smallmouth bass, and white bass fillet tissue ranged from 3.2 to 757 ng/g, with an overall mean PFOS concentration for all fish fillet of 46 ng/g wet weight [23]. Consumption of a meal sized portion (200 g, or 6.5 ounces) of this fish would then lead to average exposures of 150 ng/kg body weight (BW) for an adult weighing about 60 kg. Even though such dinners might not happen daily or weekly, the magnitude of the exposure is noteworthy, particularly in light of evidence that many fish are heavily contaminated. For example, in a 2012 study of PFCs in Mississippi River Pool 2, the highest concentration of PFOS in a fish was measured at 6,160 ng/g in a carp in section 4 of Pool 2, near the 3M Cottage Grove facility.o Water at the previous MDH limit of 0.3 pg/L would contribute about 10 ng/kg BW per day (assuming about 2 L of water consumption for a 60-kg adult). Thus, fish from contaminated waters can be an important additional source of intake. Based on similar ohttps://www.pca.state mn.us/sites/default/files/c-pfc1-21.pdf 10 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN calculations, the MDH has issued fish consumption advice for bluegill, carp, largemouth bass, northern smallmouth bass, walleye, and white bass from Mississippi pool 2 and for some species also for the Minneapolis chain of lakes [18, 24]. 2. Breast milk Recent data support that U.S. children have higher serum concentrations than adults [25, 26]. As a potential contributor, evidence has accumulated that PFCs are excreted in human milk [27]. Analyses of paired samples of maternal serum, cord serum, and maternal milk have demonstrated that PFCs are transferred through the human placenta and via human milk [28, 29]. During prolonged breast-feeding, the cumulated postnatal exposure to PFCs through lactation can be much higher compared to prenatal exposure, especially for PFOA [29], thereby causing additional exposures during sensitive early development. From serial blood samples in young children, a recent study showed that extended breastfeeding could increase an infant's blood concentrations of PFCs to several-fold above the mother's [30]. These observations are in accord with a 3M-sponsored study on EtFOSE and FC95 (PFOS) carried out in goats, as indicated by an email to Dr. Larry Zobel in early 1998.p The first studies on transfer into milk were apparently carried out in 1993, as a document from that time states: "There is some preliminary evidence that in lactating goats PFOS is transferred to milk. It is likely that lactating human females would also transfer PFOS to milk."qAgain, it appears that 3M decided not to publish the results of this study, and it further appears that 3M did not follow up with analyses of human milk to establish the potential implications of this discovery. In a two-generation study involving rats published by 3M's Dr. Butenhoff and coauthors in 2004, pup weights in the highest dose group were consistently decreased throughout the lactation period [31]. This study also found a slight but statistically significant decrease in lactation index (percent viability from day 5-22 post-partum) for Fi pups. It is not clear whether these observations were due directly to toxicity from PFCs in the milk, or indirectly via deficient lactation, and such concerns were apparently not of interest to the authors. The transfer of PFCs through maternal milk in animal models was finally recognized by 3M colleagues in a 2005 publication [32]. Accordingly, the recent publications on PFC transfer via human milk [27, 30] do not represent a discovery, but rather a re-discovery of a mechanism identified by 3M much earlier, in 1993, in studies that 3M did not disclose publicly. The relevance of the 3M results on PFC transfer via lactation in goats must be considered in a wider perspective. Due to growing concerns about lactational transfer of dioxins, PCBs, and some pesticides, toxicological studies began to pay attention to this pathway, and in 1976 the World Health Organization began a global monitoring program to measure persistent pollutants in human milk [33]. Since a thorough review handbook on human milk contamination and health risks was published in 1991, the p 3MA00906295. Fluorochemical Projects Underway in the Environmental Laboratory. Page 3MA00906295. q3MA10037365, PFOS Disposition, custodian Reagen, William K. Page 3MA10037367. 11 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN concern has been regularly raised [34]. Perhaps this evidence inspired studies at 3 M on transfer of the non-lipophilic P F C s via milk. Additional to P F C s in maternal milk, infants fed formula prepared with contaminated drinking water also could be substantially exposed to PFCs. Observed ratios between P F O A concentrations in drinking water and h u m a n serum and between concentrations in h u m a n serum and milk suggest that milk concentrations m a y approximate those occurring in contaminated drinking water [7]. Thus, Minnesota infants from contaminated communities likely had highly elevated exposures n o matter they w e re breastfed or bottle-fed. A contributing factor is that infants and children have a higher fluid consumption co mp a r e d to adults on a B W basis. Accordingly, P F C burdens will increase m o r e rapidly than in adults, and exposures to house dust and other domestic sources can further increase this difference [17, 25]. O f related concern, several studies support that U.S. children have higher serum concentrations than adults [25, 26]. This tendency is also reflected b y s e r u m - P F C concentrations in other populations, e.g., in m y o w n t e a m 's study of Faroese children [35]. M o s t of these studies did not consider the carry-over of P F C s from breastfeeding. Given that P F C s are eliminated in h u m a n milk [27], w o m e n will decrease their o w n serum concentrations b y each pregnancy and the duration of lactation [36]. Although advantageous to the w o m e n , children will absorb greatly increased P F C doses, and they are likely m o r e vulnerable to toxic effects due to sensitive organ development processes at early life stages, as I shall discuss further below. I note that the accumulated evidence on exposure via breastmilk as well as transplacental passage, in M a y of 2 0 1 7 led the M D H to revise d o w n w a r d the limits for P F O A and P F O S in drinking water [37]. 3. Consumer goods and other sources of exposure Other P F C exposure pathways likely contribute to background exposures that affect residents of Minnesota and the U.S. population in general. For example, P F C s are used in a variety of c o ns um er goods, including certain textiles [41]. P F C s from textiles m a y contribute to levels in house dust, wh i c h m a y be a particular source affecting children [7, 41]. Several of the exposure sources m a y be interrelated, as local production and point sources m a y contribute to h u m a n exposures via m o r e than one pathway. Air pollution from a production facility can result in inhalation exposure, and dust deposition will contribute to freshwater and soil contamination. T h e latter m a y lead to groundwater contamination. F o o d m a y be contaminated via use of drinking water for preparation of soup, gravy, and other foods. D u e to the stability of the PFCs, releases from sources in Minnesota also contribute to global cycling of these substances, thus adding to the so-called background contamination levels, e.g., in marine fish, s o m e of w h i c h are also marketed in Minnesota. Accordingly, at least s o m e of the so-called background exposure in the Minnesota population likely originates from production and emissions of thousands of tons of P F C s or precursor c o m p o u n d s from 3 M facilities. 12 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 4. Summary on exposures In conclusion, many thousands, probably tens of thousands of residents in the East Metro area and nearby communities have been exposed to PFCs from environmental pollution that has particularly affected drinking water and freshwater fish. The contamination was revealed publicly about 2005, but likely reached residents much earlier, as much as 20, 25, or more years ago. As the PFCs are excreted in human milk, infants would be at particular risk, whether they were breastfed or received substitute made with contaminated water. IV. HUMAN EXPOSURE TO PERFLUORINATED COMPOUNDS A. Early knowledge of human exposure By way of background, I first outline the early information on human exposures to PFCs, including information in relation to 3M. Apparently, the first public report that suggested that the general population had been exposed to fluorine-substituted organic compounds appeared in Nature in 1968 [42], where Dr. Donald Taves from the University of Rochester showed that serum proteins (from his own blood) contained an elevated amount of non-extractable fluorine that could not be ascribed to fluoride from drinking water or food. In the following years, it became clear that industrial organofluorides were a likely source. The occurrence of PFCs in the human body was further characterized when fluoride in blood samples was found to be partially bound to organic compounds of unknown structure and that they were bound to albumin [42, 43]. As would have been predicted by the known chemical properties of the PFCs, laboratory animal studies soon confirmed that PFCs are easily absorbed, both from oral intake and inhalation [4, 17, 44]. Even some dermal absorption has been documented following cutaneous contact [45], but generally is believed to be negligible (apart from certain occupational settings), as compared to the other environmental exposure pathways. It appears that 3M started measuring worker blood samples for organic fluorine by about 1976. In 1979, they measured five workers' serum PFOS concentrations and found the levels to be 100-fold higher than the expected level. Total organic fluorine was 4.1-10.1 ppm (or mg/L), which is equivalent to 4100-10,100 ppb (or pg/L), and PFOS represented 55% - 80% of that in the worker samples. By comparison, nine different studies published between 1972 and 1989 suggested average serum organic fluorine concentrations about 30-40 ppb (or pg/L) in the general population/ The latter is similar to the general background average serum-PFOS concentration of 30.4 pg/L in the NHANES samples from the general U.S. population in 1999 2000 [46]. A further report from Dr. Taves and his colleagues in 1976 extended their studies and showed that there was widespread contamination of human tissues with organofluorine r AR226-0548.pdf, Perfluorooctane Sulfonate: Current Summary of Human Sera, Health and Toxicology Data, Table II.1, page 000026 13 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN c o m p o u n d s wh i c h likely derived from commercial sources such as P F O A (or P F O S , as it s e em ed later on) [47]. Ba ck g r o u n d serum concentrations of these c o m p o u n d s w e re estimated to range from about 10 pg/L to about 130 pg/L (again corresponding to the later analyses of N H A N E S blood samples carried out b y C D C ) . Dr. W.S. Guy, one of the authors of this later report, informed 3 M b y telephone in 1975 that the study w a s going to be published soon at the annual A m e r i c a n Chemical Society conference. In this regard, 3 M 's J.D. LaZerte noted that the authors w e re "attempting to locate the source of the organic fluorocompound and thought that S C O T C H G A R D might be the source." In response, J.D. LaZerte noted that he advised Dr. G u y "not to speculate." In other words, w h e n fluorocarbons similar to those produced b y 3 M w e re found to be widespread in h u m a n blood, 3 M advised against assuming that 3 M products could be a source. Bu t as the major if not sole producer of such compounds, 3 M and 3 M products constituted an obvious and likely source. Dr. G u y 's finding, reported in 1976 [47], w a s an important milestone. In fact, one could have expected it to have triggered an extensive examination b y 3 M of the fate of P F C s emitted from 3 M production facilities and waste disposal sites, and from dissemination via co ns um er products. I have found n o evidence that 3 M seriously considered the potential implications of the research findings at the time or soon thereafter. In the late 1970s, at the time w h e n Dr. Taves and his co-workers w e re identifying organofluorine c o m p o u n d s in blood samples, landmark cases of environmental pollution were unfolding, one of t h e m L o v e Canal close to the Niagara Falls in upstate N e w Y o r k [48]. A t this site, highly stable organochlorine compounds, such as hexachlorocyclohexanes and chlorinated benzenes were leaking from a chemical dump. T h e organofluorine c o m p o u n d s being studied by Dr. Taves, including P F C s from 3 M , shared several properties with the organochlorine c o m p o u n d s in L o v e Canal, i.e., a l o w vapor pressure, s o m e solubility in water, substantial stability and resistance to break-down, and bioaccumulation. In addition, both the P F C s and the chlorinated c o m p o u n d s were deposited at public waste sites and w e re released in sewage water. Given the attention paid to chlorinated c o m p o u n d s already fifty years ago and onwards, a producer, like 3 M , of highly stable organofluorine c o m p o u n d s with physicochemical properties suggesting a potential for environmental dissemination and bioaccumulation, w o u l d be expected to trigger a careful assessment of the safety of the PFCs. T h e 3 M documents reveal that similar concerns we re raised 20 years later b y 3 M 's Dr. Richard Purdy, but that the warnings were not taken seriously, even this recently (and therefore led to Dr. P u r d y 's letter of intended resignation). After the institution of regular blood testing at 3 M 's production facility in 1976, the 3 M medical service team noticed, b y September of 1984, an increasing trend in worker organic fluorine concentrations in blood and advised "w e m u st v i e w this present trend with serious concern [...] exposure opportunities are providing a potential uptake of fluorochemicals that exceeds excretion capabilities of the body."1It is not clear from the file h o w this s 3MA01326445. Chronology - Fluorochemicals in Blood. page 3MA01326445, Confidential Information. t AR226-0483.pdf, 3M Internal Correspondence re Organic Fluorine Levels, from DE Roach to PF Riehle, dated August 31, 1984. page 003586. 14 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN recommendation was developed, nor how it was received, but health examinations apparently became part of the routine (see Section VI.B). One action taken by 3M was to search to see if it could find any blood samples that were free of organofluorine compounds. When that was unsuccessful, the following conclusion was proposed in an internal document: "It is in the interest of 3M to strengthen the evidence of non-industrial sources of organic fluorine in nornal human blood."u 3M initiated efforts beginning in 1993 to show that organic fluorine in blood could be from entirely natural sourcesv*This too obviously was a futile effort. As analytical methods became more sophisticated in the 1990s (with LC-MS instrumentation), specific compounds, such as PFOA, were identified and determined in serum samples from exposed workers. In May of 1998, 3M reported to the EPA the finding of PFOS in blood samples from the general population.TM In September of that year, Battelle, a 3M contractor, assessed data quality on 12 data sources provided by 3M on PFOS concentrations from non-occupationally exposed populations. Battelle concluded that in most cases there was not enough written documentation of the study design and sampling protocols to judge whether the resulting data were statistically valid, a conclusion that seems overly skeptical and judgmental. The only data source that was approved by Battelle to be statistically defensible was the data set of 1998 pooled serum from U.S. blood banks (geometric mean 28.2 ppb or pg/L). xThese results became available via the EPA in 2003, and a follow-up study from 2000-2001 was published and showed that PFOA and other PFCs were indeed present in virtually all subjects tested [49]. This way, the suspicions raised in the 1970s by researchers from the University of Rochester were finally recognized. Subsequently, the Centers for Disease Control found that almost all blood samples collected from a representative sample of Americans above 12 years of age during the 1999-2000 cycle of the National Health and Nutrition Examination Survey (NHANES), contained PFOA, PFOS and/or nine other long-chain PFCs [46, 50]. Subsequent NHANES studies have continued to document the wide occurrence of PFCs in the U.S. population. Overall, it appears to have taken 3M about 25 years to accept and publicly disclose the dissemination of PFCs into the environment, thereby affecting very large populations. Still, even after PFCs were discovered in local drinking water, a 3M executive in 2007 claimed that the water was safe: "Based on science, there are no health risks from PFBA, u 3M_MN03112528.pdf. In search of non-industrial sources of organic fluorine in normal human blood from the general public. v 3M_MN031125328. In search of Non-Industrial Sources of Organic Fluorine in Normal Human Blood from the General Public. wAR226-0622.pdf, 12/9/98 Letter form W. Weppner (3M) to Charles Auer (EPA) re agenda for 12/14/98 meeting, page 1. x AR226-0036.pdf, Working Memorandum on Data Quality Assessment. page 001728 and page 001722; and 226 1486. 15 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN PFOA or PFOS at levels found in the environment," 3M's vice president for environmental, health and safety operations saidy B. Detection and distribution of PFCs in humans Although most studies have relied on analyzing serum or plasma PFC concentrations (which are very similar), those measures may not accurately reflect the amounts of PFCs retained in the body overall. For example, PFCs may not show the same partition in blood, with some of them being bound to red cells, rather than serum proteins [51]. Thus, certain PFCs, such as PFOSA and PFHxA, occur only to a small extent in serum or plasma, as they bind to the red blood cells. Moreover, although PFOS and PFOA are both PFC compounds with the same length of the carbon chain, in the blood, a greater fraction of PFOS than of PFOA occurs in plasma [51]. Some PFCs are rarely detected in human blood or serum, as they accumulate in internal organs. Thus, when PFC retention in tissues was determined at 99 autopsies, PFBA was the PFC most commonly detected in kidney and lung and also showed the highest concentrations with medians of 263 and 807ng/g, respectively [52]. In liver and brain, PFHxA showed the maximum levels with medians of 68.3 and 141ng/g, respectively. These organofluorine substances are rarely detected in serum samples, yet substantial accumulation in target organs occurs. Thus, the absence of detectable amounts of these PFCs in a blood sample should not be taken as an indication that no exposure has occurred or that any uptake has been eliminated. With this caveat, serum-PFC analyses are still useful for exposure assessments for some of the PFCs. Thus, the Centers for Disease Control and Prevention (CDC) have focused on serum samples collected in connection with the NHANES [50]. These data show that PFOS, PFOA, and PFHxS continue to be detectable in virtually all Americans. There seems to be little difference in adults associated with age, but men have higher PFC concentrations than women (likely to be due, at least in part, to pregnancies and lactation), and some ethnically-related differences also exist that are likely due to dietary differences. Median concentrations in serum about the time that PFOS production was phased-out were about 30 ng/mL (PFOS), 5 ng/mL (PFOA), and 2 ng/mL (PFHxS) [50]. The average for 2007-2008 for PFOS had decreased to 14 ng/mL, while PFOA and PFHxS had changed only little [46, 53]. The 2008 data from NHANES served as national average data for comparison with the biomonitoring study carried out by MDH in the affected area that year [54]. PFBS was mostly below the detection limit, and PFBA was not reported. The long-chain PFCs are persistent in the human body and therefore stay in the blood and in organs for a long time. They may be excreted in urine, though only slowly, and the elimination half-lives are several years. This means that a continuing exposure results in accumulation of the PFCs in the body. The concentration in a blood sample therefore reflects exposures that happened during the last several years. On the other hand, when exposure stops, then elimination will be slow, usually with a constant fraction being excreted each month, each year. First-order toxicokinetics are usually assumed, i.e., that constant fractions of serum y 3M_MN03012886. Early media reports - House Hearing. 16 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN concentrations are eliminated over time [55, 56]. The time it takes for half the compound to be eliminated is called the (biological or elimination) half-life. From serial analyses of serum samples from former 3M production workers after retirement, half-lives for long-chain PFCs have been estimated to be ~3years (PFOA), ~5years (PFOS), while most short-chain PFCs have a relatively short T / in serum of ~3days (PFBA) [57]. However, the serum half-life for PFHxS is unusually long, i.e., ~9 years. If the exposure is not completely eliminated during the follow-up, the body burden will decrease more slowly (if at all). However, calculations based on serum concentrations may be erroneous, as very high PFC concentrations in various organs show that "deep" compartments, e.g., in the liver may have much longer retention times [52]. Accordingly, these half-life calculations are not precise. Also, the correlation between PFC chain length and half-life is not absolute in that, for example, the six-carbon chain (C6) PFHxS has one of the longest half-lives, while eight-carbon chain (C8) PFCs such as PFOA and PFOS have somewhat shorter half-lives (see above). Animal studies suggest that, at high exposure levels, the renal excretion may become concentration-dependent as the tubular reabsorption mediated by organic anion transporters becomes saturated [58], and variability in kidney function between subjects may also play a role. Thus, while toxicokinetic modeling may depend on the range of blood concentrations and the individual variability associated with model parameters [56], reasonable results have been obtained modeling PFC kinetics using a first-order model [55, 59]. This approach has been used by the C8 Panel to estimate serum concentrations both before and after blood sampling. Since 1999-2000, serum PFOS concentrations in NHANES have shown a significant downward trend, presumably because of discontinuing industrial production of PFOS in the U.S. In contrast, the PFOA concentrations during 1999-2000 were higher than during any other time period examined, but then remained essentially unchanged during 2003-2008 [53]. In Red Cross blood donors, declines in geometric mean concentrations from 2000-2001 to 2010 were 76% for PFOS and 48% for PFOA (48%) [60]. The decline in PFOS suggested a population halving time of 4.3 years, which agrees with the geometric mean serum elimination T / of 4.8 years reported in individuals [57]. However, blood donors are not ideal for population studies and for assessment of the kinetics of the PFCs that occur in the blood. At each blood donation, a fraction of the body burden of the PFCs is removed, and repeated blood donation can therefore increase the elimination of PFCs from the blood compartment. A shortened elimination half-life in blood donors was demonstrated in six family members with elevated PFC exposures [61]. Likewise, local residents classified as frequent blood donors had much lower serum concentrations of PFOS, PFOA and PFHxS than the non-donor population [16]. Similarly an Australian study showed that regular venesection treatment of hemochromatosis patients resulted in approximately 40% lower serum concentrations of PFOA, PFOS, and PFHxS [62]. Laboratory animals, especially the rat, show different retention patterns and even sex-related differences in elimination rates. Most laboratory animals have much shorter elimination half-lives than humans. In interpreting toxicology studies, the specific accumulation patterns must therefore be considered. An appropriate way of ensuring proper interspecies comparisons often relies on serum concentrations (or tissue concentrations), rather than dose levels. 17 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN C. Serum analyses from Minnesota residents T h e first data on serum P F O A and P F O S concentrations in residents of the affected communities, b e y o n d 3 M employees, w e re obtained b y M D H while the drinking water w a s substantially contaminated. T h e y s h o w e d average serum P F O A and P F O S concentrations of 54 n g / m L and 37 ng/mL, respectively, in 75 adult residents from Oakdale, and 30 n g / m L and 16 n g / m L in 26 adults from L a k e E l m o [18]. O f note is that 10 Oakdale children w e re included in this study, and that one of t h e m had the highest serum concentrations of 180 n g / m L ( P F O A ) and 155 n g / m L ( P F O S ) observed a m o n g all participants. T h e a n o n ym iz ed data file did not include important parameters, such as duration of breastfeeding. Also, due to the m e t h o d for selection of participants, these data are best characterized as a case series that is not necessarily representative for the residents of the affected communities. T w o years later, a follow-up study w a s carried out to examine changes in serumP F C s since the first specimens w e re collected [54, 63]. T h e study covered households served by the Oakdale municipal water supply and households with contaminated private wells in L a k e E l m o and Cottage Grove. Eligible subjects aged 20 years or older in the households selected had lived at their current residence prior to January 1, 2005, i.e., before the P F C s we re detected in the water. T h e serum concentrations averaged (range in parenthesis) 15.4 (1.6-177) n g / m L for P F O A ; 35.9 (3.2-448) n g / m L for P F O S ; and 8.4 (0.3-316) n g / m L for P F H x S . A total of 164 of the participants from 20 08 w e re re-examined on this occasion, and serum concentrations had declined on average b y 2 1 % ( P FO A) ; 2 6 % (PFOS); and 1 3 % ( P F H x S ) [63]. A s s u m i n g elimination half-lives of 3, 5, and 9 years, for the three PFCs, average reductions b y 2 5 % , 3 7 % , and 1 4 % w o u l d be expected if all exposures had ceased at the time of the first blood sampling. T h e s a m e pattern w a s evident in 2014, w h e n further follow-up w a s conducted [64]. Thus, although s o m e continuing exposures to P F O A and P F O S m a y continue, from water and other sources, it appears that the exposure to these particular PFCs, at least as regards their presence in serum, has been significantly decreasing. T h e decreases are in approximate accordance with our k n o w le dg e on elimination half-lives (see previous section). Ho we ve r, continued exposure to contaminated drinking water with P F C concentrations be l o w the M D H limits likely prevented serum concentrations from decreasing at the m a x i m a l rate. Thus, even several years after the drinking water w a s filtered, the average serum concentrations are still several-fold higher than U S averages. These studies we re not designed to identify and d o c u m e n t high-level exposures, but the m a x i m u m serum concentrations found s h o w that excesses of about 100-fold ( P F O A and P F H x S ) and10-fold ( P F O S ) above U.S. average levels remained in this c o m m u n i t y (although m a n y residents m a y have m o v e d a w a y or died in the meantime). For confidentiality reasons, the M D H reports do not contain any further detail, e.g., whether subjects with the highest levels of exposure had resided longer in the communities, but an age-dependent increase w a s noted. T h e highest P F O A and P F O S levels are similar to findings in the M i d - O h i o Valley, see section I V . D [65]. A s the publicly available summaries do not provide m u c h detail [54, 63, 64], I have examined the M D H database from the first study to carry out s o m e complementary statistical analyses. Thus, in the questionnaire used, the study participants we re asked about the n u m b e r of years of residence in their current h o m e (they we re apparently not asked about previous residences). T h e s e r u m - P F O A concentration s h o w e d a stronger association with 18 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN duration of residence than did PFOS and PFHxS (both of which may not depend as much on water intake as does PFOA). The correlation coefficient is 0.33 (a perfect correlation would have a correlation of 1, while no correlation would be 0). Although this coefficient is unlikely a matter of chance (p value below 0.001), there is much scatter. For example, the highest serum-PFOA concentration occurred in a 50-year resident (only three other subjects had been at the current residence longer). However, in another subject, resident for 46 years, the serum-PFOA was barely increased compared to the background. A range of factors must have contributed to this variability beyond the simple duration of possible exposure, such as the actual water contamination level and its possible changes over the years, the daily water consumption from the tap, and intake from other sources such as fish. Regarding the latter, the participants were asked how often they consumed fish or shellfish caught from lakes and rivers in the local area, and only 11 of 161, who responded to the question in 2010, said yes, possibly because they were sports fishers (or from a household with one). Those who consumed local freshwater fish, had about 40% higher serum concentrations of both PFOA and PFOS, and tripled levels of PFHxS (differences in actual concentrations were more similar). Although the latter PFC has not been found to bioaccumulate in fish from the Mississippi, there may be other waters in the Twin City area that may be particularly polluted with this substance, thus possibly leading to this substantial increase in serum levels in fish consumers. Because of the long elimination half-life of PFHxS, exposures many years ago may still be reflected in elevated serum concentrations. Otherwise, the highest reported serum-PFC concentrations have been encountered in production employees, and 3M worker concentrations of PFOA have ranged up to 100 pg/mL, or 100,000 ng/mL, with more recent values being lower [66]. Analysis for perfluorobutanoic acid (PFBA) was attempted for serum samples from 127 former employees and 50 current employees of the 3M Cottage Grove Facility, but concentrations were below the detection limit in 73% of the former employees and 68% of the current employees, and only 4% of the samples contained PFBA above 2 ng/mL, with maximum concentrations of 6.2 ng/mL for the former employees and 2.2 ng/mL for the current employees [57, 67]. Despite the short serum-based half-life, several former employees and local community residents had measurable concentrations of PFBA in their blood, although this may have been due to continuing exposure, e.g., through contaminated drinking water. In light of PFBA mainly binding to certain tissues, such as the kidney and lungs (see Section IV.B), non-detectable concentrations are not diagnostic, but measurable serum concentrations certainly suggest substantial exposures. D. Serum analyses from other populations 1. Mid-Ohio River Valley In the similar setting in the Upper-Ohio River Valley in West Virginia, where PFOA concentrations in drinking water in affected communities averaged about 3,500 ng/mL, the median serum-PFOA concentration in exposed residents was 374 ng/mL [65], about 100-fold above concentrations normally found in Americans. Based on these data, a water-PFOA concentration of about 1000 ng/L can be assumed to result in a serum concentration of 100 ng/mL (or 100 pg/L), i.e., a ratio of approximately 1:100 during steady state [68]. However, note 19 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN should be taken that this approximate ratio covers average numbers for residents with different histories of water consumption and with variable durations of residence in the area. This average ratio between PFOA in drinking water and in serum of participants is consistent with the data from 98 East Metro residents with private wells [63], if the expected post-exposure decline in serum concentrations is taken into account. Thus, an average serumPFOA level of 25 ng/mL will result from a contamination of 250 ng/L in the drinking water. Short-term residence would result in lower serum concentrations, while high daily water intakes would increase the accumulation in the body. Assuming a background level in serum of 4 ng/mL at the time, a PFOA concentration in drinking water at about 10 ng/L (or 0.01 pg/L) will contribute an average of about 20-25% of the total exposure to this compound, while a PFOA concentration in drinking water of 100 ng/L can be estimated to increase the serum PFOA concentration by 250%, with drinking water in this case contributing over 70% of the total exposure [7]. Thus, like in the Mid-Ohio River Valley, water contamination is a very substantial source of PFC exposures in Minnesota, even if the previous MDH water limit of 0.3 pg/L (300 ng/L) was not exceeded. (As discussed in Section IX, I conclude that MDH's prior values were not sufficient to protect against adverse effects, and, indeed, MDH lowered limits in 2017.) 2. International comparisons Internationally, where they have been measured, average serum concentrations appear to be fairly similar to those obtained from NHANES by CDC, though with higher levels, e.g., in Inuit populations [69]. Our own studies in the Faroe Islands have shown maternal and child serum concentrations of PFOA, PFOS, and PFHxS [35] that roughly correspond to U.S. levels [53], although the Faroese (like the Greenlanders) appear to be mainly exposed to PFOS and PFHxS through their marine diet [70], rather than through drinking water. I note that caution must be applied when carrying such comparisons between different populations, as, e.g., long term breastfeeding can cause elevated concentrations in infants and lowered levels in the mothers, and repeated blood donations can also lower blood-PFC concentrations (see Section IV.C). E. Conclusions on PFC exposure in Minnesota residents PFOA and PFOS are long-chain PFCs that are very stable in the body and also have long half-lives in serum concentrations. PFHxS is shorter chain, with six carbons, but has a longer half-life in the serum than PFOA and PFOS. PFBA is more quickly eliminated from the blood but is retained in the kidneys and the lungs; the tissue half-life is unknown, but can be assumed to be fairly long. These PFCs therefore accumulate and remain in the body up to many years after cessation of the exposure. For an unknown period that may have lasted decades, residents in the affected area have consumed drinking water contaminated with PFOA and other PFCs. Locally caught freshwater fish and, to some extent, local farm produce mostly likely added to the excess exposures. Due to the contamination of major aquifers serving East Metro and nearby communities, it is likely that tens of thousands of residents have been exposed to excess amounts 20 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN of P F O A and related PFCs, and long-term residents m a y have been exposed to elevated concentrations for decades. A s recently as 2010, serum P F O A and P F H x S concentrations in residents in the affected communities s h o w u p to 100-fold excesses above national averages, and P F O S concentrations 10-fold above U.S. average levels. T o the extent that residents outside East Me t r o have c o n s u m e d drinking water, or fish, contaminated b y P F C s originating from 3 M facilities, the exposed population is likely substantially greater than hitherto assumed. Higher serum concentrations w e re seen in subjects w h o had resided in their present h o m e for longer time, and in subjects w h o ate freshwater fish from local lakes. A small n u m b e r of analyses s h o w high concentrations in serum from children, and elevated levels m a y originate from breastfeeding. P F B A and P F H x A accumulate in organ tissue, and serum concentrations therefore do not reflect the exposure. Exposure patterns change with time. Eight major producers (including 3 M ) participate in the phase-out of P F C emissions, through an agreement with the E P A , with the aim of reducing and then eliminating global facility emissions and product content of P F O A and related chemicals. Ho we ve r, even ifthese c o m m i t m e n t s are or have been met, environmental dissemination of and h u m a n exposure to P F C s are anticipated to continue for the foreseeable future due to the persistence of the P F C s already disseminated in the environment, their redistribution and their continued formation from precursor c o m p o u n d s [71]. Although serum P F C concentrations are k n o w n to be decreasing in the U.S. population, the cumulated P F C burdens of residents (past and present) of the affected communities will remain above national levels for m a n y years to come. V. METHODOLOGICAL ISSUES IN RISK EVALUATION A. Interpretation of epidemiology studies Because it w o u l d be unethical to conduct experimental studies in h u m a n s with substances that m a y cause cancer or other serious effects, the m a i n h u m a n epidemiological evidence on the P F C s c o m e s from observational studies of occupational cohorts and from c o m m u n i t y studies of subjects exposed at different background levels such as through contaminated drinking water. 1. Occupational studies T h e existing occupational studies mainly regard males and are generally cross sectional, with only f e w providing follow-up over several years. In s o m e studies, blood analyses provide information on accumulated exposure levels at the point of time wh e r e the blood w a s drawn. Information from cross-sectional or prospective studies of worker populations exposed to highly increased levels of P F O A and other P F C s is useful, but can be c o mp le x to interpret. For example, follow-up studies of workers can s h o w an overall mortality deficit [72 74], i.e., that the exposed workers lived longer than expected for the general population. Further, an easily available comparison group, e.g., the U.S. general population m a y not be appropriate, if 21 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN a disease, such as prostate cancer, occurs at a different (lower) rate in the background population of Minnesota, as compared to the country as a whole. In addition, deficits in mortality or morbidity should not necessarily be interpreted as a sign that PFC exposure is beneficial to health, but rather that the exposed population, at least at the time of first hire, was in a better health condition and with better longevity prospects than the background population, which includes the non-working population, some of whom may suffer from chronic disease or otherwise have an increased mortality risk. Such selection bias is well established. Although it does not affect all health outcomes uniformly, the healthy worker effect demands critical assessment of relative risks, not just those that may show a statistically significant excess, but all outcomes [75]. A 1989 letter with detailed tables from Dr. Jack Mandel to 3M's Dr. Larry Zobel emphasized that deaths among former Chemical Division and Chemolite employees occurred in 41 states.z Both total cancer and prostate cancer were elevated in the two groups when compared to Minnesota rates. The findings are similar to an article by Drs. Gilliland and Mandel published in 1993 [72]. Partial adjustment may be achieved by using occupational comparison groups that have not been exposed to PFCs or other hazards. Another local occupational group might avoid "healthy worker" selection bias and geographical differences, but one must then make sure that the comparison group is not exposed to some other toxicants. Exactly this issue was relevant in regard to the control group of 3M workers from St. Paul that does not appear to have been unexposed,aa as further discussed below. As an optimal comparison group may not be available, it may be preferable to show the results from the exposed population in comparison with more than one reference population and to identify the strengths and weaknesses of the comparisons. 3M's Drs Geary Olsen and Larry Zobel were clearly aware of these issues when they commented, in 2006, on Dr. Bruce Alexander's draft version of a manuscript that was later published with a revised title, with Dr. Olsen as a co-author [76]. The 3M colleagues criticized the use of "non-exposed" 3M workers as a control group (with a very low cancer mortality) and emphasized the importance of the general population, although the healthy worker effects would bias the results.bb Tables were produced to show the impact of using different comparison groups.cc In the end, the published article highlighted the comparison with general population rates but also included the internal comparison group with several caveats [76]. As a further concern, mortality statistics are limited by numbers of deaths and provide little information about diseases that are rare, occur mainly in the elderly or that are not reflected well by mortality data that may not include all contributing conditions at the time of death. As an alternative, some studies use clinical pathology tests or other diagnostic means obtained at a particular point in time, to uniformly establish disease or risk markers, although such data may sometimes be complex to interpret in terms of long-term health implications. 2. General population studies z 3MA00632313 Letter of April 6, 1989 from Dr. Jack Mandel to Dr. Larry Zobel. aa 3MA00058525. Memo to Clifford W. Hanson from F.A. Ubel. April 3, 1978. bb 3MA02557490. Letter to Dr. Bruce Alexander, July 14, 2006. cc 3MA00632317 and 3MA00632315 prepared Dr. Jack Mandel and sent to Dr. Larry Zobel on April 6, 1989. 22 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN Highly relevant information regarding environmental health risks often originate from prospective studies of cohorts generated within the general population. For PFCs, this can m e a n either local communities affected b y contamination of drinking water and other emissions from production facilities or from m o r e general (background) environmental exposures. M a n y studies are cross-sectional, but the validity of s e r u m - P F C concentrations as long-term exposure biomarkers can be supported b y the long elimination half-life of PFCs. Nonetheless, prospective studies often can be m o r e useful, as changes in exposures can be better linked to changes in health. Also, birth cohorts are very important, as they can reveal impacts of exposures during early stages, but such data are often limited in numbers, and long-term follow-up studies are expensive and of course take a long time. Thus, while m a n y h u m a n studies have focused on gainfully em pl o y e d adult males, only a small n u m b e r of published articles relate to developmental exposure and vulnerable subgroups such as pregnant w o m e n and children. T h e limited extent of prospective information on exposed birth cohorts is unfortunate, as it is not just the dose that can matter but also the timing of the dosing in regard to the developmental stage of the subjects [77]. A s illustrated b y laboratory animal data, developmental exposure to P F O A induces effects that are not necessarily seen in response to exposures during adulthood [78]. M a n y studies, also in exposed communities, are limited b y uncertain chronic P F C exposure information (in part because early exposure measurements or estimates, w h e n available, m a y have been inaccurate). Studies with prospective information on exposure levels are few. Again, data regarding prenatal or childhood exposures remain scant at this point. A further concern is that exposures are usually mixed, and it m a y be difficult to distinguish between effects attributable to particular PFCs. 3. Bias toward the null In the field of epidemiology, there is a well-known and often misleading bias toward the null, of wh i c h epidemiologists (and readers of epidemiology reports) need to be careful, especially w h e n public health is at stake. Studies that do not s h o w a statistical significance are sometimes called "negative," although this is misleading. A better w o r d in n o n informative. Joint analyses of several such studies m a y well s h o w a significant difference or trend. Table 1 highlights c o m m o n causes of bias toward the null in epidemiological studies, i.e., reasons that a study might not s h o w the existence of a risk that indeed is present, though hidden due to the bias. These are all relevant to reports b y 3 M authors. Wh i l e biases in the opposite direction also exist, they are usually of m u c h less significance [79]. 23 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN Table 1. Causes of bias toward the null in epidemiology studies [79]. Inadequate statistical p o w e r in small studies Lost cases and inadequate follow-up for long-term effects E x p o s e d or otherwise inappropriate comparison (control) group Exposure misclassification Insensitive or imprecise o u tc om e measures Failure to adjust for confounders with effects in the opposite direction Disregarding vulnerable subgroups 5 % probability level to minimize risk of false positives (Type I error) 2 0 % probability level to minimize risk of false negatives (Type II error) Pressure to avoid false alarm*10 T h e bias toward the null is particularly problematic wh e r e h u m a n health is concerned; scientists and public health officers therefore often assess and rely on the direction or weight of the evidence and not solely on statistical significance, as it m a y take a very long time to obtain complete and irrefutable proof. Thus, observational studies will rarely if ever provide a 100 percent proof, and it is always possible for s o m e o n e critical of the weight of the evidence to raise s o m e type of doubt seeking to require additional or improved data before a conclusion can be d r a w n [80, 81]. It is important to repeat that the presence of uncertainties often tends to cause underestimations of actual risks, not the opposite, and this issue is of importance especially regarding substances that have not yet been studied in the detail desired. Again, m a n y unfortunate past errors in regard to industrial chemicals have s h o w n that initial assessments were erroneous and led to an underestimation of the true risks [82]. In the present report, while considering the extent of possible biases, m y conclusions are stated in terms of assessing whether P F C s pose a substantial present and potential hazard to h u m a n health. In m a n y instances, the existing evidence of a hazard is m u c h stronger than that, but I understand this to be the applicable legal standard. I shall also rely on the findings regarding the P F C contamination in W V - O H , wh e r e "probable link" assessments were requested b y the court and then provided b y the C 8 scientific panel in its reports. M y evaluation as an expert therefore considers the uncertainties involved, the plausibilities and w h a t could possibly be know n, given the study opportunities and methodologies applied. B. Toxicity and interpretation of data It is partly because of potential limitations of epidemiology studies, in particular w h e n randomized clinical trials are not possible or appropriate, that I also examine the toxicological evidence from animal studies, as is c o m m o n in epidemiology, to evaluate the plausibility of the epidemiological findings and whether or not they are supported b y the experimental evidence. Ho we ve r, the toxicology studies that w o u l d be m o st relevant w e re not necessarily carried out. A s expressed in an internal 3 M email exchange on a proposed 24 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN reproductive toxicity study, the considerations whether to carry out the study should include the "possibility of finding repro effects w h e n conducting such a study."dd Regarding PFCs, one potential critique that has been leveled b y 3 M scientists [83], particularly against certain animal models, is that effects of P F O A m a y be speciesdependent and contingent to, e.g., activation of a peroxisome proliferator-activated receptor (PPAR). This family of receptors consists of three isotypes ( P P A R a , P, and y) that regulate cellular differentiation and metabolism of lipids and carbohydrates. Although these receptors are present in h u m a n s as well, they differ in their expression in different species, thus potentially complicating extrapolation from s o m e rodent studies to h u m a n health [7]. Ho we ve r, as concluded in a recent risk assessment, both h u m a n and m o u s e P P A R - a l p h a are activated by P F O A in vitro, and the results do not clearly indicate that h u m a n P P A R - a l p h a is less sensitive than rodent P P A R - a l p h a in the in vitro systems [10]. Also, certain studies have used specially- bred rodents - s o m e of t h e m genetically modified not to express a particular receptor under study - to control for any such possible effect. VI. HEALTH EFFECTS FROM PERFLUORINATED COMPOUNDS I outline in this section the general development over time of PFC-related research, in particular at 3 M . I also touch u p o n key studies and reviews to outline h o w information w a s obtained over time and h o w know le dg e w a s pursued (or w a s n 't) over time. I highlight 3M-derived information, whether from formal publications or court files. A. Early warning signs Given that 3 M w a s an innovator in P F C production and the primary if not sole manufacturer in the U.S. for m a n y years, it is not surprising that m u c h of the early examination of P F C properties w a s undertaken or sponsored b y 3 M . W h a t does appear to be remarkable is h o w little and h o w late 3 M 's know le dg e on environmental dissemination and toxicity w a s publicly disclosed and h o w little this information inspired the c o m p a n y to conduct in-depth studies to reveal and understand any PFC-associated risks. A s already mentioned, the first public report that suggested that the general population had been exposed to fluorine-substituted organic c o m p o u n d s appeared in Nature in 1968 [42], wh e r e Dr. D o n a l d Taves from the University of Rochester s h o w e d that serum proteins contained an elevated a m o u n t of non-extractable fluorine that could not be ascribed to inorganic fluoride from drinking water or food. A s also mentioned above, it appears that 3 M started measuring worker blood samples for organic fluorine b y about 1976. In 1979, they measured five workers' serum P F O S concentrations and found the levels to be 100-fold higher than the level that w o u l d be expected. A t about the s a m e time, a further report from Dr. Taves and his colleagues s h o w e d that there w a s widespread contamination of h u m a n tissues with organofluorine c o m p o u n d s wh i c h likely derived from commercial sources such as P F O A (or P F O S , as it s e em ed dd3M_MN03436924.pdf. Page 3M_MN03436925. 25 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN later on) [47]. 3M's J.D. LaZerte noted that he instructed Dr. Taves's team "not to speculate" that Scotchgard was the source (see Section IV.A). After the institution of regular blood testing at 3M's production facility in 1976, the 3M medical service team noticed, at least by September of 1984, an increasing trend in worker organic fluorine concentrations in blood.ee Health examinations were apparently routinely carried out at 3M, although it is not clear when formal and systematic records were kept. One additional action taken by 3M was to search for blood samples that were free of organofluorine compounds. 3M also initiated could be from entirely natural esfofuorrctessb. fefgBinonthinegffionrt1s9w93erteo usnhsouwcctehsastfuolr,gbanuitcreflvueoarliende in blood a strategy to pursue studies that might benefit company interests. Among warning signs as early as the late 1960s and 1970s of potential adverse health effects from PFCs were complaints from consumers and doctors that Scotchgard could cause adverse effects, such as bronchospasm.gg 3M denied any link between use of their product and toxic effects and instead referred to trichloroethane as a potential risk, although the LD50 for the latter is 2-3 orders of magnitude greater than for the fluorochemicals. B. Early epidemiology studies (1970s to 1990s) As a general matter, information on adverse health effects of PFCs originated with early studies of exposed workers, as complemented by some experimental toxicology studies[7, 9, 17], I shall again mainly focus on developments at 3M to the extent that I have been able to identify relevant information. While a published scientific article in 1980 reported on the presence of organic fluorine compounds in serum from exposed workers [84], and drinking water contamination in Ohio and West Virginia by PFOA was discovered and later publicly disclosed [71], PFC industries, such as 3M and DuPont, conducted some evaluations of potential health effects over time, but the results of these investigations were generally in the form of internal reports that were not published in the peer reviewed literature [71]. According to internal 3M documents, 3M consulted, in the late 1970s, with their external expert reviewer, Dr. Harold C. Hodge. Dr. Hodge had served as the first President of the Society of Toxicology and was department chair at the University of Rochester, NY. Among his research interests was the toxicity of inorganic fluoride [85]. At a meeting in 1978, Dr. Hodge told 3M's medical director, Dr. F.A. Ubel, that physical examination results of employees should be compared with controls. "There appears to be indications of liver change from the physical*8 ee AR226-0483.pdf, 3M Internal Correspondence re Organic Fluorine Levels, from DE Roach to PF Riehle, dated August 31, 1984. page 003586. ff 3M_MN03112538. In search of Non-Industrial Sources of Organic Fluorine in Normal Human Blood from the General Public. 88 3M_MN01235364. Letter from Mrs. David Sweeney on February 22,1969 concerning SCOTCHGARD Brand Fabric Protector; 3M_MN01314621. Letter from Dr. Leo Johns to 3M company on August 14, 1967; 3M_MN01314624. Letter from Mrs. Russel H. Taber to 3M company on June 7, 1966. page 3M_MN01314625. 26 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN examination results. In terms of indicators of liver disorder, there are [sic] a higher percentage at Chemolite than at Decatur and the organically bound fluorine level at Chemolite is correspondingly higher."hh At this time, Dr. Hodge indicated that a potential hazard was present. When reviewing the data on fluorocarbons in blood the following year, Dr. Hodge commented regarding a POSF-based compound: "The study of levels of FC-807 [a PFOS precursor, PG] or its metabolites is of utmost importance in determining possible future problems. It should be determined if FC-807 or its metabolites are present in man, what level they are present, and the degree of persistence (half-life) of these materials. If the levels are high and widespread and the half-life is long, we could have a serious problem."" Interestingly, in a subsequent document referring to this conversation^ Dr. Hodge's comment was edited, and the last sentence from previous version was deleted, i.e., that "we could have a serious problem." It is not clear whether Dr. Hodge had agreed with the deletion of his conclusion. A 1978 memo written by Dr. F.A. Ubel, 3M's medical director, refers to a conversation with Drs Leonard Schuman and (then) Mr. Jack Mandel, where a follow-up study of Chemolite workers was discussed. The two would act as 3M consultants without the involvement of their university. The memo says that the two researchers would respect confidentiality. I have been unable to find any records of the findings from such study. A joint meeting was held in 1979 between 3M and DuPont, i.e., 3M's fluorochemical exposure committee and DuPont's Eugene Berman and some colleagues. The latter group had reviewed the documentation of organofluorine compounds in blood samples from workers and decided that it was not of sufficient importance to notify the EPA: "The committee reviewed the information on fluorochemicals in blood and came to the same conclusions as 3M--that since there were no adverse health effects, there was no reason for an 8E notification under TSCA."kkIt is not clear how the committee decided that "there were no adverse health effects," and I would disagree from a medical perspective. The decision recalls the statement by Dr. R.A. Kehoe, a medical consultant to the lead industry, who adopted an approach to the hazards of lead exposure that essentially dismissed evidence of potential harm to human health or the need for action to protect the public on the basis that the harm had not been conclusively proven. Kehoe's approach later became known as the infamous "show-me" rule, according to which "it took 2 years to put lead into gasoline and 60 years to take it out" [86]. In the intervening period, exposure to lead contamination was linked to extensive harm, and the global benefit from phasing out lead additives to gasoline has been estimated to be $2.45 trillion/year [87]. As has seen on many occasions, it is easy (and often irresponsible) to raise doubts only about research results which are pointing toward a risk, and financial stakes can easily inspire such exaggerated doubt-raising [80]. Minutes from the 1979 meeting said: "DuPont was asked if they had carried out any chronic studies on fluorochemicals in the past and if they planned any in the future. In both hh 3MA00967742. Minutes of meeting with H.C. Hodge (August 23, 1978). page 3MA00967744. 11 3MA00967775. Meeting Minutes - Meeting with H.C. Hodge (April 26, 1979) page 3MA00967780. jj 3MA00592803. Meeting Minutes - Meeting with H.C. Hodge (June 7, 1979). page 3MA00592806. kk 3M_MN00050981. Meeting Minutes on July 20, 1979. page. 3M_MN00050983 27 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN cases the answer was negative. Fluorochemicals have a low priority in their chronic testing program. They would not carry out such studies unless they were forced to by regulations. (...) 3M was asked if we had considered 8E notification. DuPont was told that we had considered it several times, but felt that since we have seen no adverse human health effects and our chemicals had no widespread potential for bioaccumulation, we did not feel we had a reportable case." ll 3M's conclusion of "no adverse human health effects" would obviously depend on how hard they looked, as acute toxicity would not be expected, and chronic or delayed adverse effects can only be ruled out by conducting detailed studies. Likewise, "no widespread potential for bioaccumulation" is counter to predictions from physicochemical properties, and it is not clear how 3M justified the conclusion (which was later shown to be erroneous). 3M either closed its eyes to the evidence, or chose purposefully not to find it, or being generous to 3M, it seems possible that 3M may have mistakenly relied on the absence of evidence, despite the old dictum that "the absence of evidence is not evidence of absence," which later became famous in U.S. politics. 3M had an excellent opportunity to gather evidence emanating from health surveillance of exposed workers, such as those employed at the 3M Chemolite (Cottage Grove, Minnesota) manufacturing plant. A cross-sectional study of worker health was summarized by the 3M medical officer in 1982. There was a high prevalence of high blood pressure, and elevation of cholesterol, changes that the authors believed to be more likely due to lifestyle, not occupational exposures.TMTM It is not clear how this conclusion was reached, as the fact that other factors may have contributed to the outcomes does not mean that the occupational exposure had no impact. In particular, elevated cholesterol - as an indication of liver toxicity - was soon found to be a major effect of PFC exposure and should not have been ignored in 1982. The most extensive early epidemiology study was conducted in connection with the thesis project carried out by Frank Gilliland, MD from the University of Minnesota School of Public Health. This study provided surveillance data on employees who voluntarily participated in the program at the Chemolite plant. Although Gilliland identified several associations that suggested adverse health effects, 3M later undermined the findings from the 1990 medical examination data stating that they could not be repeated in subsequent 1993 and 1995 medical surveillance examinations. This was also the conclusion of an article with 3M's Dr. Geary W. Olsen as the first author and published in the Journal of Occupational & Environmental Medicine in 1998; Dr. Gilliland was a co-author on that paper.1TMOne potential reason for the difference between the conclusions by Dr. Gilliland and Dr. Olsen is that Dr. Olsen in the later article relied on measured PFOA concentrations, while Dr. Gilliland had access to measured total organic fluorine only, thus perhaps including other PFCs that may have contributed to the changes u 3MA10035034Phone Conversation Report # Disclosure Of Info On Levels Of Fluorochemicals In Blood / Note Dated 19790730. mmAR226-0481.pdf, 3M Internal Correspondence re 1981 Chemolite Health Evaluations, from DE Roach to PF Riehle, dated August 12, 1982. page 003581. nn AR226-0474.pdf, Geary W. Olsen, et al., An Epidemiological Investigation of Reproductive Hormones in Men with Occupational Exposure to Perfluorooctanoic Acid, 40 JOEM 614-619 (July 1, 1998), with Summary of study. page 003468. Also as Olsen, Geary W., et al. "An epidemiologic investigation of reproductive hormones in men with occupational exposure to perfluorooctanoic acid." Journal o f occupational and environmental medicine 40.7 (1998): 614-622. 28 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN observed in the 1990 examinations.00 In a letter to Dr. Geary Olsen dated 23 October 1996, Dr. Gilliland (by then an Assistant Professor at the University of New Mexico) raised concerns about the follow-up study and asked that the manuscript should reflect the viewpoint of all the co investigators: "The overall focus might best be directed toward addressing the specific research hypotheses of the study rather than refuting the preliminary 1990 study."pp This apparent shift in focus is concerning, as it suggests that 3M was seeking to bias the results in order to undercut the earlier study. The letter suggests that Dr. Gilliland did not have access to the new raw data, but a memo from Dr. Geary Olsen indicates that supplementary data analyses were carried out,qq although it is difficult to judge the conclusions drawn by Dr. Olsen without access to the tables referred to, as they were not in the published paper [88]. The article listed Dr. Olsen as the first author, Dr. Gilliland second, followed by other 3M colleagues, including Dr. J. Mandel. It is not clear whether and how Dr. Gilliland's concerns were taken into account. The blood concentrations showed an apparent logarithmic distribution, but a log transformation, which one would normally expect, was apparently not used. Also, while there were concerns about some differences in the methods, only cross-sectional analyses were included. The two follow-up studies in 1993 and 1995 included 111 and 80 workers, 68 of them being the same, and 94 subjects examined in 1993 also participated in the 1990 study carried out by Dr. Gilliland. The published article refers to a half-life of 18-24 months for PFOA in serum as reason for not using repeated measurements methods to examine the data [88], although this reasoning does not invalidate the use of a stronger "repeated measurements" analysis. Dr. Gilliland's supervisor at the University of Minnesota, Dr. Jack S. Mandel, had previously as a 3M consultant co-authored other articles with other 3M colleagues, and he has also coauthored publications that were supported by other industry sources. Dr. Mandel is now employed with the Exponent consulting firm. Dr. Gilliland pursued an academic career and is now a Professor at the University of Southern California. Another early source of information relates to the registered mortality of 3M employees considered to have been exposed to PFCs (mainly PFOA). Film plant workers were used as the control group for these occupational studies, although it was known that they had at least some exposure and therefore constituted a very imperfect control group (see Section V.A.1).rr According to Alan De Waard, nearly all (95% plus) new employees on the Decatur site started in the chemical plant, and many started in the fluorochemicals drying jobs.ss This issue of exposed control groups was apparently not mentioned or taken into account in the published articles written or sponsored by 3M; the lack of an unexposed control group would tend to dilute oo AR226-0474.pdf, Geary W. Olsen, et al., An Epidemiological Investigation of Reproductive Hormones in Men with Occupational Exposure to Perfluorooctanoic Acid, 40 JOEM 614-619 (July 1, 1998), with Summary of study. page 003474 pp 3MA00630893. Letter from Dr. F. Gilliland to Dr. G. Olsen, 23 October 1996. qq3MA00652081. Memo to file, 15 January 1998. rr 3M_MN00000789, Fluorochemical Steering Committee Minutes. ss 3MA00000675, Re: FC Communications. 29 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN the findings and bias the results toward the null. This problem is similar to control animals being exposed to contaminated feed in a toxicology test. M a n y studies also suffered from selection bias. T h e 3 M health and mortality studies focused on production plant workers exposed to PFCs. B u t at the time, it w a s already well k n o w n that workers exposed to toxic substances might erroneously appear to be in better health, with lower mortality than the background population due to the "healthy worker" effect, i.e., that the em pl o y e d population is on average in better health than the population at large [75, 89]. In addition, a cohort mortality study w a s not carried out in 3 M 's A n t w e r p plant due to "its recent construction in the 1970s,"ttand former employees that we re not covered b y 3 M insurance we re also not part of the study.uu Also the explanatory p o w e r of the statistical models w a s low, only 2 % - 3 0 % . vv Systematic evidence from 3 M 's occupational medicine records w a s analyzed only from the 1990s and then reached the public at a delay, and only in parts, and without detailed reference to the caveats. Such studies w e re apparently not a matter of priority. Again, discussions we re held at 3 M whether Dr. Gilliland's data constituted a T S C A 8(e) submission, i.e., whether they referred to a h u m a n health hazard.ww It appears n o such report w a s ever submitted. C. Early toxicology studies (1970s to 1990s) Despite being k n o w n and in commercial use since about 1950, little w a s published in scientific journals on P F C toxicology until the 1980s. Internal 3 M documents s h o w that c o m p a n y scientists had been aware of the health risks of P F O A as early as the 1960s [90]. Publications on the solubility of P F O A salts w e re publicly available in the 1960s (although they m a y not have been accurate). Immunotoxicity w a s docu me nt ed in a study available to the E P A in the late 1970s, but this o u tc om e w a s apparently not further pursued for m a n y years. O n e difficulty w a s that analytical chemistry did not reliably determine organic fluorides in organic matrices until the late 1970s. W h e n the magnitude of worker exposure to and absorption of P F O A b e c a m e evident and impossible to ignore any longer, a series of studies w e re initiated b y 3 M , including an in vitro mutagenicity assay, 28-day oral studies in rats and mice, and 90-day oral studies in rats and mo nk ey s.xx 3MA01407039. An Epidemiology Investigation of Clinical Chemistries, Hematology and Hormones in Relation to Serum Levels of Perfluorooctane Sulfonate among Male Fluorochemical Production Employees. page 3MA01407050. uu 3MA00514877. Appendix III Robust Summaries of Toxicology, Epidemiology, and Health Studies. Appendix III, III-65. page 3MA00514877 vv 3MA01407039. An Epidemiology Investigation of Clinical Chemistries, Hematology and Hormones in Relation to Serum Levels of Perfluorooctane Sulfonate among Male Fluorochemical Production Employees. page 3MA01407103. 3M_MN00000789, Fluorochemical Steering Committee Minutes. 3MA02512309. Animal Toxicity Studies with Ammonium Perfluorooctanoate. 30 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN In a review of 3M-sponsored subacute toxicity studies by 1979, a conclusion on PFOS was that "FC-95 was the most toxic of the three compounds studied and certainly more toxic than anticipated. [...] Unless there are adequate data through human epidemiological evaluations that can reasonably assure relative safety of these compounds following long term exposure, lifetime rodent studies should be undertaken as soon as possible."yy However, 3M apparently did not submit any study to the EPA until 1987, which was a two-year chronic study in rats; several pre-1976 studies were submitted to the EPA only in 2000.zz Likewise, epidemiology information, for example of 3M workers, was slow to emerge publicly. Of note, 3M and Dupont agreed in 1979 about the low priority of toxicology studies and the lack of a TSCA requirement to report findings to EPA. Early PFOA toxicology studies commissioned by 3M were summarized in 1980 [44], and the liver was highlighted as a target organ, while effects on the immune (or reticuloendothelial) system were also reported. The studies reviewed had been conducted in two stages, first with a focus on skin and eye irritation, acute oral toxicity, and one-hour inhalation studies. The study reports were not submitted to the EPA until 2000.aaa Later on, a 1988 pilot feeding study in rats for 28 days revealed that "[t]he primary lesion consisted of focal to multifocal cytoplasmic enlargement (hypertrophy) of hepatocytes among animals fed 300 ppm or less, and multifocal to diffuse enlargement of hepatocytes among animals fed 1,000 ppm or more."bbb Subsequently, in 1997, it was reported that the rat feed contained PFCs, and control groups showed the presence of PFOS in livers of control rats used in toxicology studies; the source was traced to fishmeal in the feed.ccc With PFOS present in the rodent feed, the control animals were not free of PFC exposure, as should have been intended.dddFor interpretation of low-dose toxicity, the presence of a true (unexposed) control group is crucial, as an underestimation of dose-related effects may otherwise occur. Effects on lipid metabolism and thyroid function were among those studied at non-3M research institutions in the 1980s [91, 92]. Induction of peroxisome proliferation in the liver was well established by 1990 and resulted in a substantial number of publications on mechanistic aspects. D. Further 3M-influenced research (1990s to early 2000s) Scientists working with the 3M company began to publish studies on PFC toxicology from the late 1990s [88], although far from all studies at 3M were published, and often with a delay. In 1998, a 3M report to the EPA reported the widespread occurrence of PFCs in serum samples from blood banks, and environmental contamination and the persistence of the yy 3MA00593073. FC-95, FC-143 and FM-3422 - 90 Day Subacute Toxicity Studies Conducted at IRDC - Review of Final Reports and Summary. zz 3MA00016315. September 8 2000. aaa 3MA00016315. September 8 2000. bbb 3M_MN00433331(FC-143). September 29 1988. ccc 3MA01470049. Email attachment found in Custodian Olsen, Geary W. (CB429_1208981). ddd3MA00176550. Post-pioneer Risk Assessment of Adverse Effects in Marine Mammals from PFOS in the Food Chain, #1. page 3MA00176550. 31 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN PFCs in the human body became an increasing concern [71]. In April 2000, 3M submitted to the EPA a study that showed deaths among monkeys exposed to fairly low levels of PFOS.eee Soon after that, 3M announced that it was phasing out the production of PFOS, stating that the chemical had been "detected broadly at extremely low levels in the environment and people." Yet 3M qualified the findings by stating: "All existing scientific knowledge indicates that the presence of these materials at these very low levels does not pose a human health or environmental risk" (http://www.chemicalindustryarchives.org/dirtysecrets/scotchgard/4.asp). Dr. Geary W. Olsen from 3M was an author of more than 30 articles published in 1998 and onward. However, the influence of 3M on PFC research goes beyond in-house studies and authoring original reports and reviews. The company also funded or supported research elsewhere, whether by contract laboratories or academic institutions. A total of eight such studies were published in 2010-2012. Thus, 3M scientists and their collaborators contributed a substantial proportion of recent scientific articles available on the PFCs, especially in the early years of this millennium. This support and influence has some troublesome aspects. The most comprehensive mortality study of 3M workers compared deaths at Cottage Grpve with those at the unexposed St. Paul plant from 1947 to 2002. In the thesis report by Dr. Raleigh, exposure estimates were generated from air measurements, and dividing the workers into six different levels of exposures, a dose-dependent risk appeared for prostate cancer, although not statistically significant. The author concluded that the results supported previous findings of a prostate cancer risk. However, in the published report [93] that was co authored by Dr. Olsen from 3M, emphasis was on comparisons with the general population, the absence of a trend in quartile (not sextile) exposure groups, and the lack of apparent association of prostate cancer was said to be in agreement with findings in other studies. The Discussion section of the published article notes that family history of prostate cancer may play a role. Indeed, a study published the same year [94] showed a PFAS-associated prostate cancer risk in subjects from the general population who had a family history of the disease. This issue has apparently not been considered in 3M-supported research. We have previously seen that 3M had a clear interest in depicting its products as toxic only at very high doses and that any findings that suggested adverse effects at lower exposures were challenged and mostly kept away from the public. While this type of bias is probably common in the chemical industry [95], it is a problem that 3M support or sponsorship of published research may not always have been disclosed, but rather seems to have been downplayed and perhaps hidden. A note from a meeting in 2008 suggests that the findings by a consulting company "should be published, probably by a ghost writer."ggg It is of course difficult to determine the extent to which this approach was used by 3M. A report from July 2012 mentions "3M perfluorochemical studies in progress," supported by research contracts. One contract for "Danish birth outcome study" was awarded to a private entity named International Epidemiology Institute, jointly with UCLA and Aarhus*8 eeeAR226-0145. 26-Week capsule toxicity study with perfluorooctane sulfonic acid potassium salt (PFOS; T-6295) in cynomolgus monkeys, p. 002636. fff 3M MN03059185.pdf. Cancer mortality in 3M chemical workers. December 2013. 888 3M_MN00863024. Plastics Europe Fluoropolymers Committee / 081009 Ad-hoc Tox WG Minutes. page 3M_MN00863026. 32 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN University in D e n m a r k (most of the results from these studies did not lead to firm conclusions). O n e of the publications listed b y 3 M provides the following information on indirect support: "T h e original study w a s supported b y the International Epidemiology Institute (IEI), wh ic h received funding from the 3 M C o m p a n y " [96]. O f further interest, the blood samples were analyzed for P F O S and P F O A b y 3 M 's laboratory in St.Paul. A Danish cancer case-control study w a s also funded b y a 3 M contract to the participating institutions, including the IEI [97], and it provided the following statement: "T h e 3 M C o m p a n y had n o role in the design of the study, interpretation of the results, or writing of the paper. T h e authors declare they have n o competing financial interests." Nonetheless, the study questionably concluded that s e r u m - P F C concentrations appeared not to be related to a cancer risk - despite signs of elevated risk for prostate cancer despite the short follow-up time. A m o n g 3M-supported studies mentioned in a 20 08 report is a study from Stockholm, wh e r e P F O S immunotoxicity surprisingly could not be replicated b y oral dosage [98]. T h e article, published in 2010, refers to 3 M support b y an unrestricted research grant and also thanks D a v e J. E h r e s m a n from 3 M for serum analyses (there is n o 3 M co-authorship). But the published article only superficially mentions possible limitations that could explain the apparent absence of immunotoxicity. A s another example, 3 M authors [83] cited a publication to support their v i e w that cellular receptors vary between animal species and P F C toxicity. T w o of the 3 M authors are acknowledged in the publication referred to [99] for "critical review and helpful suggestions," but there is n o information on funding or conficts of interest. Analytical quality is essential in these types of research, and m o st publications in the field report quality assurance data. Ho we ve r, in regard to the 3M-supported Danish national birth cohort, wh e r e blood samples w e re sent b y regular mail, those that w e re collected in the winter s h o w e d very substantial differences between paired samples ranging from -77 to + 3 8 % for individual P F C analyses; in mo st cases concentrations w e re lower in the delayed and transported samples[100]. Such substantial imprecision, a s su me d to be r a n d o m apart from varying with season, will likely cause a bias toward the null, thereby underestimating the P F C impact on the outcomes [101]. T h e study authors have apparently not reported on the implications of this serious analytical imprecision and their interpretation of the problem in regard to their published reports on lacking or minimal exposure-related adverse effects. In m y view, it is clear that 3 M inspired and influenced P F C research and that publications on 3M-supported projects sometimes disclosed 3 M authors, sometimes not. Analyses of research bias have docu me nt ed that publications reporting on industry-supported studies tend to be biased toward conclusions beneficial to the sponsor's interests, even if a potential conflict of interest is declared. This has been found in regard to pharmaceuticals [102], nutrients [103], tobacco [104], and toxicants [105]. There are indications that such bias is present here in regard to studies sponsored b y 3 M . Part of the bias is that only selected reports, if any, have been m a d e available to the public. A n additional source of bias could be due to the choice of control groups and reference populations as well as the statistical approach to data analysis. In 2004, the E P A filed a suit against D u P o n t alleging that "D u P o n t had - over a 20 year period - repeatedly failed to submit information on adverse effects (including information on liver e n z y m e alterations and birth defects in offspring of female D u P o n t 33 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN workers)" [90]. Although no systematic study has been carried out by the EPA regarding 3M and PFCs, the evidence available suggests that 3M aimed at concealing or explaining away, or delaying unwelcome conclusions, and that 3M-commissioned or 3M-sponsored research was less likely to provide results that would be against the company's interests. In 2006, the U.S. EPA reached a settlement agreement with 3M to resolve 3M's reporting violations under the Toxic Substances Control Act, according to which 3M did not admit to a violation, but agreed to pay a penalty in excess of $1.5 million for 244 separate violations.hhh E. Growth of PFC research Even though PFCs have been produced for over 60 years, independent publication on PFC toxicity only began in earnest about 10 years ago [106]. The broader scientific community, therefore, is still at an early stage of understanding about how human exposure to these compounds affects health. For example, chronic toxicity studies have been published only based on rats [7]. A formal cancer bioassay is also missing. In addition, insufficient attention had been paid to exposures during sensitive developmental stages. I note in particular that few epidemiology studies have focused on exposures during infancy or prenatally, although early development must be considered a highly vulnerable period that must be taken into regard when determining exposure limits [107]. It is remarkable that such studies were apparently not carried out under industry auspices in the several decades of commercial production. Since the first reports in scholarly journals that revealed widespread global occurrence of PFOS in wildlife [108] and the detection of PFCs in blood from the general population [109] were published about 2000, the scientific literature on the environmental and toxicological aspects of PFCs has increased, and the annual number of publications on the PFCs is said to exceed 400 [110]. Still, by comparison, the quantity is less than for many other chemicals for which there are human health concerns. The Web of Science lists a total of more than 34,000 publications when searching for titles using the words `perfluorinated' or `perfluoro*'. However, only about 10% of these publications relate to environmental sciences, toxicology, or public, environmental, and occupational health. Only 60 articles were published before 1990 and 179 during the 1990s. Many of the early studies focused on PFCs other than those relevant to this report, many of them mainly related to chemical research. Thus, most of the published articles on human health risks from PFCs are fairly recent. To understand the growth of health-related knowledge on PFCs in further detail, I examined the coverage of the main substances in 78 major journals within the fields of environmental science, toxicology and public health during the first 10 years of this millennium [106]. These are the journals in which my colleagues and I would most frequently publish our findings, and they also prominently figure in the reference list of this report. I used the unique Chemical Abstracts Service (CAS) number for PFOS and PFOA and extracted the articles from the SciFinder data base. The 120,000 journal articles published in these journals during 2000 2009 included a total of 271 articles that, at least to some extent, covered PFOS and a slightly hhhhttps://www.epa.gov/enforcement/3m-company-settlement; https://www.epa.gov/sites/production/files/documents/3m-consentagr.pdf; https://www.epa.gov/sites/production/files/documents/3m-auditagreement.pdf. 34 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN higher number (363) on PFOA (most of the articles being the same). More than half of these articles were from 2008 and 2009, thus indicating a recent and substantial increase in research activity. Of the 120,000 articles published in these journals during 2000-2009, only 37 referred to PFBA. For comparison, the twenty most mentioned environmental chemicals (e.g., toxic metals, PCBs, and PAHs) were each covered in over 2,000 articles during this period, and lead alone was dealt with by close to 1,000 articles each year. Thus, in comparison, PFCs were not priority compounds for publicly-published environmental and toxicology research, at least up to 2007. As an indication of increased attention, the Web of Science lists a total of 2,310 PFC publications for the years 2010-2016 in the same relevant fields, i.e., an average of 370 per year, a substantial increase. When using the National Library of Medicine database (PubMed), the search items `perfluorinated' or `perfluoro*' return about 10,000 articles for all years, of which one out of four is retained when limiting to `humans'. Of these, 450 are classified as referring to children and 181 to infants, as of March, 2017. Again, most of these publications relate to medical technology uses, or are merely descriptive, thus only to a limited extent dealing with adverse human health effects from long-term environmental exposures. Accordingly, the intensive focus on PFCs in scientific publications happened during the most recent 10 years, thus slowly emerging decades after the first discoveries of PFC toxicity at 3M. Also, the reports from the court-mandated C8 studies, described below, are also very recent and mainly relied on cross-sectional study designs, although fortunately on large population groups. The evidence at hand is therefore fairly recent and unlikely to represent the full toxicological perspective, such as those that may occur at a delay, and some adverse effects and vulnerable subpopulations may not yet have been identified. The occurrence of adverse effects at chronic exposure to low PFC levels still needs to be explored in greater detail, especially regarding the long-term effects of developmental exposures. As has been seen on numerous occasions [82], the evidence available today may therefore underestimate the true extent of the PFC toxicity. F. Public knowledge on PFCs Even though PFCs had been produced for over 60 years, little was published on adverse health effects of PFCs before 1990. As reviewed above, early scientific studies carried out under 3M auspices were in some cases not published (such as the study of the transfer into milkiii) or only published after a delay of up to about 25 years (as in the case of some of the toxicology studies). Some of the findings led to apparent internal controversies that may have prolonged the delay in publication or the decision not to publish, such as Dr. Gilliland's study of worker's immune functions.jjj Some reports were held back and not submitted to the EPA after1 1113MA10037365, PFOS Disposition, custodian Reagen, William K, page 3MA10037367. jjj CB429_1167824, Custodian: DiPasquale, Thomas J., Peripheral Blood Lymphocyte Count in Men Occupationally Exposed to Perfluorooctanoic Acid. page 3MA00323876 to 3MA00323890. 35 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 3M internally, but without proper evidence, decided to conclude that "there were no adverse health effects.',kkk At least one medical report on adverse health effects in 3M workers was never published. Thus, Dr. Frank Gilliland sought to publish his results on immune dysfunctions in 3M workers. In December 1993, Dr. Jeff Mandel, Dr. Gilliland's adviser from 3M's Medical Department, provided 3M's FCSC with a copy of Dr. Gilliland's draft manuscript entitled "Peripheral Blood Lymphocyte Paper."mDr. Mandel indicated, "We're working with [Dr. Gilliland] regarding some of the wording." After the FCSC reviewed the Gilliland lymphocyte paper, it was not published, and the results were not shared with the EPA, or with state regulators at the time.mmmI have not identified any other relevant studies that were carried out, although I have been unable to identify systematically any additional unpublished reports of medical interest in addition to the cases of delayed or incomplete reporting. Overall, 3M appeared to set overly strict requirements before allowing conclusions on the existence of adverse effects, and it seemed that 3M used outside experts to elicit uncertainty about conclusions considered unwanted. Some of the exposure patterns and adverse health issues that have been recently reported can be seen as a consequence of the physicochemical properties of the PFCs that were known already in the 1950s [11]. I have been able to locate only few documents from 3M that clarify how the knowledge on environmental fate and human exposures developed over time. However, that a substantial risk to human health was present must have been realized at 3M long before Dr. Purdy summarized his concerns in the 1990s. Certainly, very substantial information must have been available to 3M by the time of the announcement in 2000 that a phase-out of commercial PFC production would be completed by 2002. The company reported two years later that environmental contamination in Minnesota had occurred. Although an increased number of biomedical articles on PFCs has appeared during the most recent decade or so, the adverse effects at low exposure levels have probably not been fully characterized, especially regarding vulnerable populations and developmental exposures. Prudent conclusions therefore must consider that the evidence so far is not completely developed, especially where the existing evidence is pointing toward probable harm. An expert group appointed by the National Research Council (NRC) refers to this erroneous conclusion as the "untested chemicals assumption" [111]. From reviewing the background evidence, it is clear to me that this consideration was not considered by 3M, but that the company in fact attempted to weaken evidence that was unwelcome. kkk 3MA10035034. Phone Conversation Report # Disclosure Of Info On Levels Of Fluorochemicals In Blood / Note Dated 19790730. m3MA00323875-890. mmmCONT003896. Plaintiffs' memorandum in support of their motion to amend their complaint to add a claim of punitive damages. page CONT003912. 36 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN G. Recent key reviews, studies and resources I mention here certain recent key reviews and studies. In m y discussion of particular endpoints, in the next section, additional reviews and studies are referred to. 1. C8 Science Panel General population studies addressing P F C s mainly have been cross-sectional, but important research data have e m er ge d from the M i d - O h i o River Valley population, wh e r e P F C contamination of drinking water occurred. T h e final conclusions of the C 8 Science Panel, as submitted to the Court, refer to the probable links of the P F C contamination and plausible adverse effects regarding cancer and several other important health conditions. T h e C 8 Panel carried out several large-scale studies, although m o s t of t h e m focused on P F O A . T h e Panel concluded that P F O A exposure w a s probably linked to six important health conditions, including t w o types of cancer [8]. In s o m e w h a t greater detail, the W e s t Virginia Circuit Court in 2005 approved a class action Settlement A g r e e m e n t in a lawsuit about releases of P F O A from D u P o n t 's production facility in W o o d County, W e s t Virginia. T h e Settlement created a Science Panel of three epidemiologists that w a s to conduct research in the c o m m u n i t y to evaluate probable links between P F O A exposure and h u m a n disease. In addition, a C 8 Health Project w a s established to collect data from Class M e m b e r s through questionnaires and blood testing. This c o m m u n i t y health study includes approximately 70,000 O h i o and W e s t Virginia residents with at least one year of exposure to drinking water contaminated with P F O A from about 50 ng/L to over 3 0 00 ng/L. Data on serum P F O A concentrations provide information on the relationships between external dose from drinking water and the internal dose, i.e., the serum concentration, and a variety of biological changes. T h e m e d i a n s e r u m - P F O A concentration for all participants w a s 28 ng/mL, and the m e di an in the highest decile (the subjects with the highest 1 0 % of exposures) w a s 48 2 ng/mL. These data, and the conclusions released b y the Science Panel constitute an important basis for the present report. Ba s e d on the results from these studies and an evaluation of the literature, the Science Panel delivered reports on `probable links,' as su mmarized in the final report from 2012. T h e C 8 Panel determined that exposure to P F O A had Probable Links to adverse effects on the following h u m a n health conditions (Table 2). Table 2. Adverse h u m a n health conditions, wh e r e `probable links' to P F O A exposure w a s identified b y the C 8 Panel [112]. Ulcerative Colitis Pregnancy-Induced Hypertension/Preeclampsia Thyroid Disease H i g h Cholesterol Kidney Cancer Testicular Cancer 37 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN Although the reports to the Court were not peer-reviewed at the time and only provide a brief summary of the new study results, most of the evidence has since then appeared in peer-reviewed scientific journals, which will be referred to below. 2. Other major assessments The PFCs have been the focus of a variety of evaluations carried out by regulatory agencies, such as the Agency for Toxic Substances and Disease Registry (ATSDR) and the U.S. Environmental Protection Agency (EPA). In addition, reviews on particular aspects have been generated by the National Toxicology Program (NTP) on immunotoxicity and by the World Health Organization's International Agency for Research on Cancer (IARC) on cancer risks. These sources refer to a wider range of studies than the present report, where the focus is on human health regarding exposures occurring in Minnesota. Human epidemiological studies have certain inherent limitations, as the evidence by nature is observational and often without well-characterized exposures over time and without unexposed control groups, as is possible in toxicological laboratory studies. This means that each subject's exposure must be characterized from, say, a chemical analysis of a blood sample, which then acts as a proxy variable and reflection of the true exposure, though without indicating the exact daily intake. Likewise, the human epidemiology studies can document only correlations between exposure measurements and particular outcomes, and such correlations may not always be causal.nnn A more common complexity is that exposures usually are mixed, i.e., that we are exposed to more than one PFC at a time. For example, subjects with high PFOS exposures usually have elevated exposures also to other PFCs. Even if serum-PFOS concentrations show clear associations with some adverse effects, then it is difficult to rule out any causative role contributed by other PFCs that may or may not show statistically significant associations with the outcome. However, due to the contamination in the Upper Ohio River Valley, a unique situation occurred, where PFOA was the main PFC in the exposure, and strong evidence on PFOA effects could therefore be extracted. In addition, some populations may be exposed to PFCs to variable degrees and in different relative proportions, and statistical methods can then help separating the likely impacts linked to individual PFCs. In the absence of human experiments on PFC toxicity, which would of course be unethical, agencies often choose to rely on experimental studies in laboratory animals to generate risk assessments and to reach conclusions on safe exposure limits for single PFCs. While this has been a long-term tradition for these purposes, the present report evaluates the weight of the epidemiological evidence on adverse effects of PFC exposures in the light of supporting toxicity evidence to determine whether these particular PFCs pose a substantial present and potential hazard. " " Dr. Deanna Luebker of 3M acknowledged that the determination that epidemiological findings demonstrate a causal relationship between the presence of a PFC in a subject and an adverse effect involves a degree of judgment and is supported by various factors, including evidence of "enough data so the numbers are strong enough," statistical evaluation, data comparison, sufficient sample sizes and evidence of the incidence and commonality of a given effect, evidence of an increase in effect at higher exposure. Luebker Dep. Tr. (Aug. 29, 2017) 49:9-50:15. 38 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN VII. ADVERSE HEALTH EFFECTS AT INDIVIDUAL ENDPOINTS A s stated above, it is m y opinion that P F C s pose a substantial present and potential hazard to at least h u m a n i m m u n e system functions, reproductive functions including adverse effects to the next generation, endocrine functions, thyroid functions, liver functions, cardiovascular functions, and b y causing or increasing the risk of cancer. Below, I discuss the different h u m a n health hazards one b y one. T h e first section is on i m m u n e system dysfunctions, as m u c h evidence is n o w available and because these effects have not been dealt with in detail in reviews b y regulatory agencies. T o s o m e degree, this is true also in regard to reproductive toxicity and endocrine disruption, while other organ systems and cancer have been dealt with in substantial detail elsewhere, so that m y coverage can be briefer. In each of the following subsections, I discuss the epidemiological evidence that I rely on, summarize the supporting toxicological evidence, and lastly discuss possible mechanisms, and additional studies or potential criticisms relating to the endpoint in question. I have m a d e a reasonably comprehensive review of the epidemiological evidence, and have em pl o y e d a weight of the evidence approach, as is c o m m o n l y accepted in the scientific c o m m u n i t y in reviewing studies on a particular topic. A. Immunotoxicity and autoimmunity It is m y opinion, based on the weight of the epidemiological evidence, and supporting toxicity evidence, that P F C s pose a substantial present and potential hazard to h u m a n i m m u n e system functions. T h e i m m u n e system is crucial in fighting co mm un ic ab le diseases. It is also crucial in detecting and eliminating cancer cells. In addition, the i m m u n e system is involved in allergic disease and in autoimmunity. A s the adaptive i m m u n e system is p r o g r a m m e d during early development, immunotoxicity assessment is particularly relevant in subjects with P F C exposures during early life [113]. A s discussed above, P F C s are excreted in h u m a n milk, and breastfed children m a y thus be particularly at risk. T h e i m m u n e system is a sensitive target for P F C toxicity, perhaps the most sensitive, as illustrated b y studies of deficient antibody responses to routine vaccinations in children exposed to PFCs. This approach w a s r e c o m m e n d e d b y an international s y m p o s i u m in 1999 [114] and has been used to characterize im munotoxic effects of, e.g., polychlorinated biphenyls (PCB s) and dioxins [115-118]. Children w h o are highly exposed to immunotoxicants m a y be unable to generate e n ou gh antibodies to provide protection against the infectious diseases against wh i c h they are vaccinated. Responses to vaccinations in terms of concentrations of specific antibodies can therefore be used to assess i m m u n e dysfunctions. T h e National Toxicology Pr o g r a m ( N T P ) concluded in 2 0 1 6 that P F O A and P F O S are likely, or "pr es um ed to be," h u m a n immunotoxicants [5]. N T P uses the term "pr es u m e d " to denote the level of evidence just b e l o w "k n o w n , " and stronger than "suspected." In addition, autoimmunity, including ulcerous colitis, is a docu me nt ed adverse effect [119]. T a k e n as a whole, P F C exposure at levels similar to or b e l o w those reported from the East Me t r o area are associated with a range of immuno to xi c effects. 39 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN A s I discuss further below, prospective studies of birth cohorts have s h o w n dramatic negative effects of P F C s regarding children's response to routine immunizations, thus demonstrating that these substances can adversely impact the development of the adaptive i m m u n e system in early childhood. A reduced or flat antibody concentration response to vaccinations has been observed even in adults of the general population at elevated s e r u m - P F C levels. Such effects are linked to an increased occurrence of infectious diseases. Thus, the implications can be substantial. T h e m e c h a n i s m of action is not yet clear in terms of h o w P F C s m a y affect allergy, a u t o i m m u n e disease (such as ulcerative colitis, as suggested b y the C 8 Panel), and other i m m u n e functions, e.g., eliminating cancer cells. 1. Epidemiological evidence M y review of available epidemiological studies demonstrates a strong link between P F C exposure and adverse effects on h u m a n i m m u n e system functions. a. I w a s the principal investigator of a study that found significant adverse impacts of P F C exposure on indicators of vaccination efficacy in children. T h e first study, wh i c h w a s based on 656 births in the Faroe Islands followed 587 of the children through to age 7 years and found that a doubling in exposure to P F O S and P F O A w a s associated with an overall decrease b y about 5 0 % in the antibody concentration [35, 120]. A t the s a m e time, a substantial n u m b e r of children at age 7 had such a l o w antibody concentration that they had n o long-term protection against the targeted diseases despite a total of four vaccinations. T o understand the significance of this finding, a brief explanation of routine immunizations is helpful. B y inoculating an antigenic c o m p o n e n t of a pathogenic organism, the b o d y develops a defense in terms of antibodies. T h e aim is to protect against future development of the disease. Vaccination is considered one of the cornerstones of m o d e r n disease prevention. W h e n a child receives the scheduled T d (tetanus-diphtheria) vaccinations, for example, he or she will generate sufficient capacity for antibody production to avoid getting infected b y tetanus and diphtheria. If a person's antibody concentration is too low, his or her i m m u n e system will likely be unable to fight off the disease. T h e antibody response to childhood immunizations is of clinical relevance and reflects major i m m u n e system functions, and in addition is a feasible parameter to use in population studies [121]. Thus, study subjects have all received the s a m e doses of antigen (in the form of the vaccines) at the s a m e ages, and examinations can then be scheduled at similar ages, i.e., at similar intervals after the mo st recent vaccination [114]. O u r particular study focused on the fishing c o m m u n i t y of the Faroe Islands, wh e r e residents with frequent intake of marine food have increased exposures to marine contaminants, such as the P F C s [70]. A major advantage of these studies is that the population is fairly h o m o g e n e o u s and that participation rates at follow-up remain high. W e have followed a Faroese birth cohort of 656 singleton births through to adolescence [122]. A m o n g P F C s in maternal pregnancy serum, P F O S s h o w e d the strongest negative correlations with antibody concentrations in their children at age 5 years (after three vaccinations within the first year after birth), wh e r e a doubling in exposure w a s associated with a difference of - 4 1 % (p = 0.0003) in the diphtheria antibody concentration. P F C s in the child's 40 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN serum at age 5 likewise s h o w e d clear, negative associations with antibody levels, especially at age 7 (two years after a booster vaccination at age 5 years), and a doubling in P F O S and P F O A concentrations w a s associated with differences in antibody levels between -24 and - 3 6 % (joint effect of -49%, p = 0.001, in a structural equation model) in the overall antibody concentration. For doubled concentrations at age 5, P F O S and P F O A s h o w e d odds ratios (ORs) between 2.4 and 4.2 for falling b e l o w a clinically protective antibody level of 0.1 I U / m L for tetanus and diphtheria at age 7. W e concluded that developmental exposure to P F C s is associated with humoral i m m u n e system deficits in h u m a n s [35]. It is worth noting that the P F O S and P F O A levels in maternal pregnancy serum and the child's serum at age 5 that we re measured in this study s h o w e d concentrations similar to, or lower than, those d o c u me nt ed in prior studies in the U.S. [46]. O f particular concern is the finding that several children at age 7 years (two years after the age-5 diphtheria and tetanus vaccination booster) had antibody levels against diphtheria and/or tetanus b e l o w the clinically protective level of 0.1 I U / m L [35, 123]. This m e a n s that the children had n o long-term protection against the diseases - despite a total of four vaccinations. W e calculated the O R s for a doubling in the child's age-5 s e r u m - P F O A concentration as a predictor of having an antibody concentration b e l o w 0.1 I U / m L at age 7 years. T h e O R s for tetanus w a s 4.2 ( 9 5 % CI, 1.5-11.4) and for diphtheria w a s 3.3 ( 9 5 % CI, 1.4-5.5). Bo th were significant at a p level <0.01. P F O S s h o w e d lower O R s (marginal significance), and certain other measured PFCs, such as P F H x S , had O R s b e l o w 2 and non-significant (we determined serum concentrations only of the major P F C s in this study). O u r regression analyses also s h o w e d P F O A to be a strong predictor of lowered antibody concentrations. W h e n w e used a structural equation mo d e l that allowed us to c o mb in e the t w o s e r u m - P F O A measurements at ages 5 and 7 years [120], w e find that a doubled s e r u m - P F O A concentration w a s associated with a change in the age-7 antibody concentration of -38.2% ( 9 5 % CI: -56.1; -13.0) for tetanus and -34.7% ( 9 5 % CI, 52.5; -10.2) for diphtheria. W h e n w e adjusted for the other PFCs, the regression coefficients we re -29.6% and -26.9 % , respectively, i.e., virtually unchanged [120]. Likewise, adjustment for the elevated P C B exposure in the Faroes did not materially affect the calculations, as w o u l d be expected due to the poor correlation between the t w o [35]. b. T h e findings of our above study are consistent with a smaller study carried out in N o r w a y on a subgroup from the national birth cohort. In 50 3-year-old children, inverse correlations w e re found between the moth er's P F C exposure during early pregnancy and decreased antibody levels in their children against four different childhood vaccinations, with rubella showing a statistically significant decrease at higher exposures to P F H x S , P F O A , P F O S , and P F N A [35]. This study also found that increased concentrations of P F O A , P F N A , and P F H x S w e re linked to statistically significant increases in the incidence of their children suffering from c o m m o n cold and from gastroenteritis. O f importance, in these children, elevated P F C exposures w e re linked to both lower antibody concentrations and m o r e frequent infections. c. In a m o r e recent Faroese birth cohort, s e r u m - P F A S concentration profiles during infancy were estimated based on the duration of breastfeeding, and the calculations were validated b y comparison with measured s e r u m - P F C concentrations at age 18 months. A t the lower P F C exposures, inverse associations with age-5 serum concentrations of antibodies against tetanus and diphtheria vaccines w e re similar to those seen in the previous cohort. Concentrations estimated for ages 3 and 6 m o nt hs s h o w e d strong inverse associations with antibody 41 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN concentrations at age 5 years, i.e., m o r e than four years later. particularly for tetanus. These associations w e re stronger than those seen for P F C concentrations at ages 18 m o nt hs and 5 years and therefore support the notion that the developing adaptive i m m u n e system is particularly vulnerable to im mu no to xi c exposures during infancy. This finding also m e a n s that studies relying on serum concentrations that do not reflect ages at peak vulnerability will likely underestimate the true effects. d. U.S. colleagues relied on the N H A N E S to extract data on a total of 1,831 adolescents aged 12-19 years between 1999 and 2 0 0 6 [124]. In cross-sectional comparisons, a doubling of the concomitant s e r u m - P F O S concentration s h o w e d a 1 3 % decrease in rubella and 6 % in m u m p s antibody concentrations, while measles did not s h o w a clear association. In the authors' wording, the findings suggest a less robust response to vaccination or greater w a n i n g of vaccine-derived immunity over time. This study is of interest, as it focused on childhood vaccines other than the diphtheria and tetanus toxoids. c. In addition to routine childhood immunizations, m a n y people receive immunizations for the flu, often on an annual basis and for a specific flu variant, such as the Avia n flu or the Swine flu. P F C exposure has been s h o w n to be linked to decreased flu vaccine effectiveness. Thus, a study carried out in connection with the C 8 studies encompassed 411 adults, w h o s e serum samples w e re analyzed before and about three w e e k s after flu ( A / H 3 N 2 ) vaccination [125]. Thus, the elevated s e r u m - P F O A concentrations w e re associated with a w e a k e n e d vaccine antibody response also in adults. f. P F C exposure has also been s h o w n to be linked to decreased effectiveness of boosters of vaccines first received in childhood. In a study of 12 healthy adult volunteers, increased P F C exposure w a s associated with flatter changes in the serum concentrations of tetanus and diphtheria antibodies. Following the booster vaccination, antibody responses widely differed during the first 10 days, with t w o subjects appearing not to respond at all, and the steepness of the antibody concentration increase w a s inversely associated with the concomitant serum P F C concentrations, especially for P F O S and PFHxS[126]. g. This decreased short-term response to vaccines is in agreement with the similar PFC-associated decreased responses seen in 4 1 9 of the Faroese children at age 5 years [35, 123]. W h e n re-examined about a m o n t h after the booster vaccination against diphtheria and tetanus, lower antibody responses w e re found at increased concurrent exposures to P F O S , P F O A , and P F H x S , while prenatal exposures appeared to play a lesser role in regard to the response to the booster. Although past exposures w e re u n k n o w n , the results suggest that adverse effects on short-term vaccine responses in adults m a y be primarily determined b y the current exposure, independent of age. h. Other studies have also linked P F C exposure to adverse impacts on the b o d y 's ability to fight off various c o m m o n diseases including colds, fevers and gastroenteritis. Thus, a study of 359 Danish children from the O d e n s e Child Cohort found that increased maternal serum concentrations of P F O A and P F O S at the end of the first trimester w a s significantly associated with a higher frequency of fever and s y m p t o m s in their children. T h e study followed the cohort of 359 children at ages 1-3 years b y monitoring the frequency of fever and associated s y m p t o m s every 2 w e e k s for a year (via text messages). T h e n u m b e r of days with 42 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN fever >38.5C (>101.3F), and also in combination with nasal discharge or cough, w a s significantly increased in association with increases in the maternal serum concentrations of P F O A and P F O S [127]. These findings are in accordance with the m u c h smaller N o r w e g i a n study already mentioned [128]. i. A s part of the studies of the Danish National Birth Cohort, maternal early pregnancy serum from randomly selected 1400 w o m e n and their offspring we re analyzed b y 3 M for P F O S and P F O A [129]. Hospitalizations for infection of the offspring w e re identified b y the linkage to the National Hospital Discharge Register, through to age 11 years. Diagnoses, such as airway infection, appendicitis, middle ear infection w e re merged, and n o clear pattern w a s observed w h e n results w e re stratified b y child's age at infection. In addition to relying only on exposures during early gestation, a recent study raised doubt about the validity of the chemical analyses [100]. Nonetheless, this study w a s highlighted b y 3 M authors w h o criticized our 2 0 12 J A M A article [83], as c o m m e n t e d on below. j. M o s t recently, a study of a large Japanese birth cohort recorded physicians' diagnosis of c o m m o n infectious diseases - including otitis media, pneumonia, respiratory syncytial virus infection, and varicella - u p to 4 years and reported higher incidence rates at elevated prenatal exposures to P F O S and P F H x S [130]. Like the O d e n s e Birth Cohort study, it focused on the m o st relevant preschool ages, did not exclude cases that w e re not hospitalized, and relied on valid prenatal exposure measurements. k. In adults, a study conducted b y the C 8 Science Panel based on the health examinations concluded in an interim report that increased P F O A exposure w a s associated with lower serum concentrations of total IgA, IgE (in females only), though not I g G [112]. Thus, using total and non-specific immunoglobulin concentrations, this study is at least partially supportive of adverse i m m u n e effects from P F O A exposure. T h e result concerning I g G concentrations should be interpreted with s o m e caution because the C 8 study examined total I g G immunoglobulins (whereas our study, A.1.a. above, focused on concentrations of specific I g G antibodies directed against vaccine antigens), thus substantially limiting the sensitivity of the C 8 study. l. P F C s also have been found to be linked to certain forms of a u t o i m m u n e disease, in wh i c h the b o d y 's i m m u n e system attacks its o w n tissues. This link is demonstrated by t w o studies conducted b y C 8 Science Panel epidemiologists, the first being an occupational study of 3,713 workers, w h o s e P F O A exposures w e re evaluated. Using a ten-year lag, the occurrence of ulcerous colitis and, without a lag, rheumatoid arthritis s h o w e d significant associations by greater disease frequencies at elevated P F O A exposures [131]. These results w e re also reflected b y the C 8 Panel conclusions, wh e r e the C 8 Panel stressed a probable link between P F O A exposures and ulcerative colitis. m. T h e second study concerned the general population in the M i d - O h i o River Valley, wh e r e 151 cases of ulcerous colitis w e re identified in connection with the medical examinations. W i t h a p value less than 0.0001, higher s e r u m - P F O A concentrations predicted a greater risk of developing the disease [119]. 43 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN n. Similarly, in the C 8 Panel's study of the >50,000 residents of the M i d O h i o River Valley, certain i m m u n e function parameters we re measured. Specifically, antinuclear antibody ( A N A ) concentrations in serum w e re used as a screening parameter for a u t o i m m u n e disease, such as rheumatoid arthritis. There w a s an increasing trend with s e r u m - P F O A concentrations. In contrast, the inflammation marker, C-reactive protein, fell with increasing P F O A . In each case the pattern w a s repeated in the s a m e w a y for males and females [112]. Although su mmarized in the reports from the Panel, n o peer-reviewed article has been located. o. Allergies m a y also be related to P F C immunotoxicity, as reported by studies linking P F C exposure to increased development of allergies in children [132]. First, a study of 2 4 4 Taiwanese children found that increased cord-blood concentrations of P F O A and P F O S correlated with elevated cord-blood IgE in boys [133]. T h e immunoglobulin IgE is usually increased in allergic subjects, but the predictive value of elevated cord-blood I g A in regard to subsequent development of allergy or atopy is limited [134]. Further, a study of 343 Japanese births reported an inverse association between P F C concentrations and cord serum IgE concentrations [135], thus revealing opposite tendencies in the t w o studies. p. Using m o r e reliable clinical data, a study of the Faroese birth cohort born in 1997-2000 included data on allergy and asthma at ages 5 to 13 years [136]. Tw e n ty -t wo of the 559 children had not been vaccinated against M M R , and a m o n g those, higher serum concentrations of the five P F A S s at age 5 years (but not prenatally) w e re associated with increased odds of asthma at ages 5 and 13. Ho we ve r, the associations we re reversed a m o n g M M R - v a c c i n a t e d children. Wh i l e P F A S exposure m a y impact i m m u n e system functions, this study suggests that M M R vaccination might be a potential effect modifier. q. A recent study from China assessed pulmonary function parameters in children with and without asthma [137]. Children with asthma had higher P F C exposures, and their pulmonary functions w e re poorer the higher the P F C exposure. It therefore seems that P F C exposure m a y contribute to the development and worsening of asthma. r. W h e n white blood cell populations are affected b y exposure to an immunotoxicant, differential counts of the peripheral blood cells m a y be a useful measure. In about 50 5-year-old Faroese children, the current s e r u m - P F A S s concentrations we re associated with higher basophil counts [138]. Although clear associations we re otherwise not seen with P F A S exposures in this small sample, the exposure-related changes in lymphocyte subpopulations suggest cellular im mu ni ty effects and dysregulation of T-cell mediated immunity. s. T o further explore the mechanisms, a study w a s carried out in N o r w a y to characterize gene expression in cord blood and its association with P F A S concentrations, antibody concentrations, and infectious disease incidence. Several i m m u n o m o d u l a t o r y genes, especially the C 1 7 gene, we re linked to all three parameters, and these findings therefore supported a PFAS-linked genetic m e c h a n i s m underlying both the lowered antibody response and the increased susceptibility to infectious disease [139]. t. Regarding occupational exposures, a study of 3 M workers found clear associations between increased P F C exposure and decreased leukocyte counts, a sign of adverse 44 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN impact on the h u m a n i m m u n e system. These immunotoxicity appeared to be in agreement with experimental data, especially those in monkeys. T h e results w e re reported b y Frank Gilliland, M D , as an o u tc om e of his thesis w o r k at 3 M about 1990. In his thesis, Dr. Gilliland concluded: "Total serum fluorine w a s negatively associated with all peripheral leukocyte counts except P M N s [PolyMorphonuclear Neutrophils] and M O N O s [Monocytes], wh i c h w e re positively associated." O f note, the basophil count at elevated exposures w a s lower in the adults, while a recent study s h o w e d that they w e re higher in highly-exposed children [138]. Regarding this study, it is worth noting that 3 M w a s aware of Dr. Gilliland's findings of i m m u n e system abnormalities, but, unlike certain other findings in Dr. Gilliland's thesis, n o report of the i m m u n e system findings w a s publicly disclosed. Thus, in articles published after submission of Dr. Gilliland's thesis, and jointly authored with 3 M 's Dr. Geary Olsen, nothing w a s mentioned about leukocyte counts, despite Dr. Gilliland's thesis results that s h o w e d clear suggestions of im munotoxic effects. In June of 1993, Dr. Gilliland forwarded to Dr. Jack Mandel, his thesis adviser w h o w a s part of the 3 M medical service, three manuscripts concerning P F O A - e x p o s e d Chemolite employees; the manuscripts w e re to be submitted for publication that su m m e r . ppp Dr. Gilliland and Dr. M a n d e l co-authored three papers in the public record, one of t h e m on mortality and only t w o on clinical findings,^ rrr,sss but none of t h e m includes the results on P F O A immunotoxicity. Ho we ve r, the files produced b y 3 M contain a manuscript entitled "Peripheral B l o o d L y m p h o c y t e Co u n t in M e n Occupationally E x p o s e d to Perfluorooctanoic Acid."TM It is likely that this manuscript w a s part of the package sent to Dr. M a n d e l in June of 1993. O n e copy of this manuscript includes the following co mm en t: "In addition, the standard immunotoxicologic assessment defined b y the National Toxicology Pr o g r a m needs to be conducted for P F O A , " and a hand-written c o m m e n t says, "bring to attention of L.Zobel, R.Perkins, S. Gordon."uuu A s with 3 M 's m o n k e y study from 1978 that revealed im mu no to xi c effects [140] (see below), the leukocyte count results we re never published, and the r e c o m m e n d e d immunotoxicology assessment w a s apparently not conducted. omAR226-0473. Frank Davis Gilliland, Fluorocarbons and Human health: Studies in an Occupational Cohort (October 1992) (unpublished Ph.D. thesis, University of Minnesota), with Summary. Page 003247. ppp 3MA10017137, # Transmittal Of Revised Versions Of Three Manuscripts Concerning PFOA Exposed Chemolite Employees. page 3MA10017137. qqqGilliland, Frank D., and Jack S. Mandel. "Mortality among employees of a perfluorooctanoic acid production plant." Journal o f Occupational and Environmental Medicine 35.9 (1993): 950-954. rrr Gilliland, Frank D., and Jack S. Mandel. "Serum perfluorooctanoic acid and hepatic enzymes, lipoproteins, and cholesterol: a study of occupationally exposed men." Americanjournal o f industrial medicine 29.5 (1996): 560-568. doi:10.1002/(SICI) 1097-0274(199605)29:5<560::AID-AJIM17>3.0.CO;2-Z. sss Olsen, Geary W., et al. "An epidemiologic investigation of reproductive hormones in men with occupational exposure to perfluorooctanoic acid." Journal of occupational and environmental medicine 40.7 (1998): 614-622. ttt 3MA00323876. Frank D. Gilliland and Jack S. Mandel. Peripheral Blood Lymphocyte Count in Men Occupationally Exposed to Perfluorooctanoic Acid. Pages 3MA00323876 to 3MA00323890. uuu 3MA10016749. Frank D. Gilliland and Jack S. Mandel. Peripheral Blood Lymphocyte Count In Men Occupationally Exposed To Perfluorinated Acid. Page 3MA10016756. 45 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 2. Toxicological evidence Epidemiological evidence showing an association between P F C exposure and adverse effects on h u m a n i m m u n e system functions finds additional support in various toxicological studies of i m m u n e system functions. Experimental studies have provided substantial documentation of im mu no to xi c effects [5, 17, 141]. Immunotoxicity of P F C s has been demonstrated in a wide variety of species and models, as well as in vitro in relation to h u m a n white blood cells. a. Already by 1980, there w a s an interest in the immunotoxicity associated with perfluorinated organic compounds, wh e r e effects observed included decreased antibody formation against sheep red blood cells and spleen enlargement [142]. There seems to have been little or n o follow-up in the published literature on these early findings, as also noted b y DrGilliland. b. A n early 3 M - c o m m i s s i o n e d 90-day study carried out in m o n k e y s in 1978 demonstrated toxicity effects on the gastrointestinal tract and the reticuloendothelial system (i.e. i m m u n e system)vvv In this 90-day study labeled 226-0447,www the doses of F C - 1 4 3 ( P F O A ) given w e re 0, 3, 10, 30 and 100 mg/kg/day. All m o n k e y s at the 100 dosage and three out of four at 30 mg/kg/day died; compound-related microscopic lesions we re seen in adrenals, bone m a r r o w (hypocellularity), spleen and l y m p h nodes (atrophy of lymphoid follicles in both), as also highlighted b y Dr. Gilliland in his thesis from 1993, wh e r e he added: "N o follow-up studies of these observations have been reported."*xxx c. This m o n k e y study and other toxicology studies carried out in or about 1978 were reviewed the following year with colleagues from DuPont, w h o agreed that there we re compound-related effects indicated in both m o n k e y and rat studies, and that additional adverse effects we re revealed in the data (although these additional adverse effects we re not reported b y 3 M in the text of the studies) (226-1456, page 000140).yyy Certain of the findings we re su mmarized in a published review article [44] from 1980. d. Recent interest in experimental immunotoxicity of the P F C s began after year 2000, at first focusing on reductions in lymphoid organ weights, lymphoid cell numbers, and de n o v o antibody synthesis [141]. These studies clearly d o c u m e n t adverse i m m u n e system effects and support the notion of P F C immunotoxicity [5]. Using a standard immunological challenge of injecting sheep erythrocytes into P F O S - e x p o s e d mice, adverse effects w e re seen at serum concentrations similar to levels observed in occupational exposure, and a calculation of safe exposures w a s similar to the highest serum concentrations in people with background P F O S exposure [143], while P F O A immunotoxicity occurred at higher serum concentrations [144]. vvv 3MA02512309. F.D. Griffith and J.E. Long. Animal Toxicity Studies with Ammonium Perfluorooctanoate. *"* AR226-0447. Final Report, Ninety Day Subacute Rhesus Monkey Toxicity Study, International Research and Development Corporation, Study No. 137-090, November 10, 1978. xxxAR226-0473. Frank Davis Gilliland, Fluorocarbons and Human health: Studies in an Occupational Cohort (October 1992) (unpublished Ph.D. thesis, University of Minnesota), with Summary. Page 003214. yyyAR226-1456. March 5, 1979 - DuPont reviewed 3M's new C-8 rat and monkey studies and agreed that there are compound-related effects indicated in both studies, and that additional adverse effects apparently were revealed in the data but not reported by 3M in the text of the studies. (Exhibit P (EID 123133). 46 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN Other studies have s h o w n P F C effects on i m m u n e measures, such as cytokine expression and signaling related to inflammation and T helper cell responses [141]. c. In regard to response to infections, a study in mice demonstrated that P F O S exposure at levels associated with deficient i m m u n e functions s h o w e d elevated P F O S concentrations in blood, but also in thymus, spleen, and lungs as well as reduced survival after influenza A infection [145]. f. Further, studies of mice injected with sheep erythrocytes, as a standard test of i m m u n e system function, demonstrate deficient i m m u n e system responses from PFCs, in parallel to the h u m a n studies of vaccine responses. Several rodent studies have applied this experimental mo d e l to assess any effects on the antibody response. In a study of P F O S , the lowest observed effect level ( L O E L ) for males w a s 0.05 m g / k g total dose and 10-fold higher in females (which excrete P F O S m o r e rapidly). M e a s u r e d s e r u m - P F O S concentrations at these dose levels w e re 91.5 22.2 ng/g and 666 108 ng/g ( m e a n SD), respectively [143]. T h e concentrations w o u l d be almost the s a m e if measure in ng/mL, the unit used for h u m a n blood sdamples. Thus, the serum levels measured in the male mice at the lowest dose applied were similar to the highest conccentrations measured in residents in the exposed area. Yet, these levels we re associated with significant adverse effects. A s n o lower doses w e re applied, the data do not allow consideration to wh i c h extent lower concentrations m a y also be associated with adverse effects in this animal model. g. Available information on i m m u n e system effects from developmental exposure also supports a link between P F C exposure and adverse i m m u n e system effects. In one study of gestational exposure, male pups we re again m o r e sensitive than females to the effects of P F O S and confirmed that the developing i m m u n e system is vulnerable to P F C exposures and that functional deficits in innate and humoral im mu ni ty are detectable at adult age [146]. h. H u m a n white blood cells provide a meaningful in vitro mo de l to assess i m m u n e system effects, and studies have been carried out to determine the in vitro effects of P F C exposure, generally with a focus on cytokine secretion [141]. P F O S w a s found to be the most toxic of several P F C s studied while P F B S w a s m o r e potent than P F O A [147]. Thus, white blood cells from h u m a n volunteers s h o w e d effects at P F O S concentrations of 0.1 p g / m L (or 100 n g / m L ) [147], i.e., similar not only to concentrations seen both in affected male mice in toxicology studies, [143] but also to levels in residents exposed to contaminated drinking water [63]. 3. Perspective a. In connection with the need to identify safe limits for contamination of drinking water with P F O S and P F O A , the E P A surveyed the P F C literature and su mmarized its results in P F O A and P F O S risk assessment reports [148, 149]. T h e E P A draft risk assessment documents finds that P F C s exhibit immunotoxicity in experimental models and that the epidemiological evidence is additive, although m i x e d exposures complicate the attribution of effects to specific PFCs. 47 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN b. A similar conclusion w a s reached in the m o r e recent A T S D R ToxProfile on the perfluoroalkyls [4]. A s I wrote in m y written c o m m e n t s to the Agency, the coverage of h u m a n immunotoxicity is unusually very brief in the ToxProfile, and n o mention of this potential is m a d e in the chapters on public health. In addition, PFAS-associated immunotoxicity in children w a s not even included as a potential critical effect of these substances. M o r e specifically, A T S D R stated (p. 176) that the b e n c h m a r k calculations that w e generated [150] could not be used because of the absence of a "control group." T h e h u m a n evidence w a s therefore not considered in the calculation of a Mi n i m a l Risk Level. I note that the requirement of an unexposed control group is unrealistic and also unnecessary, as described, e.g., b y E F S A [2]. c. M o r e recently, the National Toxicology P r og ra m ( N T P ) in 2 0 16 reviewed the immunotoxicity information on P F O S and P F O A and concluded that both are "p r es um ed" to constitute i m m u n e hazards to h u m a n s [5]. B o t h P F C s suppress the antibody response in animal studies, with a "moderate" level of evidence from studies in humans. T h e evidence indicating that P F O A and P F O S affects multiple aspects of the i m m u n e system supports the overall conclusion that both can be pr es um ed to alter i m m u n e function in humans, even though the m e c h a n i s m s are not clearly understood. T h e reason for considering the h u m a n evidence "moderate" is that all studies are observational (not experimental) and refer to m i x e d exposures, wh e r e the individual and joint roles of P F O S and P F O A are difficult to extract. T h e term "pr es u m e d " is the strongest b e l o w "k n o w n " in the N T P vernacular. d. T h e European F o o d Safety Authority likewise in their initial opinion in 20 08 [1], to wh i c h I contributed, relied on experimental toxicity studies at a time wh e r e little information on immunotoxicity and f e w h u m a n studies w a s available. A n updated version is expected in the fall of 2017. c. According to the recent evaluations, the epidemiological evidence demonstrating an association between (mixed) P F C exposure and adverse effects on the h u m a n i m m u n e system is strong and is supported b y ample toxicological evidence on effects of P F O S and P F O A , while other P F C s have been addressed only in f e w studies. Nonetheless, a recent 3 M financed review [151] has argued against the associations being causal b y highlighting uncertainties and minor discrepancies between studies. f. In a letter to J A M A , three 3 M scientists criticized our article that w a s published in J A M A in 2 0 1 2 [reviewed at A.1.a. above] [83]. T h e y suggested that antibodies toward diphtheria and tetanus were of n o public health interest, as the diseases are rare and not likely to occur in the c o m m u n i t y under study. According to their view, our findings w o u l d reflect the i m m u n e system response only to the specific vaccine antigens, wh i c h in m y m i n d is an inappropriately narrow interpretation. T h e evidence reviewed above s h o w similar tendencies for different antibodies under different circumstances, and our understanding of i m m u n e system does not suggest that each specific vaccination activates functions that are different from those triggered b y other vaccines. In arguing against a link between P F C s and clinically important immunotoxicity, the 3 M authiors referred to the Danish National Birth Cohort study that relied on hospitalizations only and imprecise P F C analyses [129]. In addition, nu me r o u s diseases of different origins we re merged, and the likely impact of multiple social, demographic and other 48 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN factors at different childhood ages we re ignored. Thus, this study should be regarded n o n informative and not as a proof that P F C s are not immunotoxic. T h e 3 M authors also refer to a 3M-supported m o u s e toxicological study that reported n o immunological effects at a high dietary P F O S exposure [152]. This reported conclusion, however, is at odds with the conclusions of other studies of the s a m e strain of mice showing P F O S immunotoxicity [98, 141], as well as with the evidence reviewed above. T o support their v i e w that interactions with cellular receptors vary substantially between animal species and affect P F C toxicity, the authors cite a publication [99] that acknowledges t w o of the 3 M authors for "critical review and helpful suggestions." Still, species differences are of importance. This is certainly true in regard to the capacity to eliminate P F C from the blood, and recent rodent studies have therefore expressed exposures in terms on s e r u m - P F C concentration levels for m o r e appropriate comparison between species [141]. Further, s o m e m o d e s of im munotoxic action likely differ between rodents and h u m a n s due to differences in the expression of certain nuclear receptors. Ho we ve r, experimental evidence suggests that at least s o m e immunotoxicity effects are independent of P P A R a , as demonstrated, e.g., b y studies with P P A R a knockout models exposed to P F O A [153]. In addition, P P A R a is also expressed in humans, although to a lesser degree than in rodents, and potential P P A R a dependence or independence m a y therefore affect species dependence of P F C immunotoxicity [153]. Thus, mechanistic investigations of PFC-induced suppression of cytokine secretion from h u m a n i m m u n e cells have demonstrated that P P A R a activation w a s involved in the P F O A - i n d u c e d immunotoxicity, while other pathways appeared responsible in regard to P F O S immunotoxicity [154]. In agreement with the very detailed N T P review [5], w e find that the species differences do not invalidate our conclusions that P F C exposure likely presents a h u m a n immunotoxicity risk [123, 155]. T h e caveats mentioned above regarding interspecies comparisons indicate that dose-dependence m a y differ between h u m a n s and rodents, but also that such differences do not argue against the i m m u n e system being a major target organ for P F C toxicity in both h u m a n s and laboratory animals. 3 M 's arguments are not weighty enough, and s o m e are clearly flawed, such that they do not change m y opinion that, based on the weight of the epidemiological evidence, and supporting toxicity evidence, P F C s pose a substantial present and potential hazard to the h u m a n i m m u n e system functions with likely severe consequences. Thus, in agreement with the N T P review [5], I conclude that the h u m a n evidence strongly supports the existence of P F C dependent immunotoxicity at background exposure levels, although the relative role of individual P F C s is s o m e w h a t unclear. Ho we ve r, detailed statistical calculation s h o w that P F O A related effects on specific antibody concentrations appear to be independent of other P F C exposures (which are m o r e closely correlated) [120]. Experimental studies d o c u m e n t that all P F C s tested have immunotoxic effects. Regarding P F B S , in vitro studies so far s h o w that it is less potent than P F O S , but m o r e than P F O A [147]. B. Reproductive toxicity It is m y opinion, based on the weight of the epidemiological evidence, and supporting toxicity evidence, that P F C s pose a substantial present and potential hazard to h u m a n reproductive system functions, with risks of adverse effects to the next generation. 49 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN T h e focus in this section is primarily on obstetrical appearances of P F C toxicity, including pregnancy hypertension and preeclampsia, wh i c h conditions w e re determined b y the C 8 Panel to have a probable link to C 8 exposure [156]. I also address pregnancy outcomes, including miscarriage, birth weight, decreased sperm quality and fecundity. These outcomes m a y or m a y not be mediated b y endocrine disruption mechanisms, but are dealt with in this section, as they refer to pregnancy and pregnancy outcomes. T a k e n as a whole, P F C exposure at levels similar to or b e l o w those reported from the East Me t r o area are associated with a range of reproductive toxicity effects. Other outcomes considered m o r e clearly to reflect endocrine disruption are considered in the following section and include changes in serum concentrations of sex hormones, delayed development including delayed puberty, inhibited lactation and shorter breastfeeding durations, early menopause, and changes in reproductive h o r m o n e concentrations in serum. Regarding developmental toxicity affecting the next generation functionally and in regard to subsequent disease risks, i.e., so-called Developmental Origins of Health and Disease ( D O H a D ) , these aspects are properly discussed in regard to the relevant organ systems (such as the i m m u n e system). 1. Epidemiological evidence Data from m y review of available epidemiological studies demonstrate a strong link between P F C exposure and adverse effects on h u m a n reproductive system functions. O n e early and important aspect of reproduction is fecundity, i.e., the capability of achieving pregnancy, and other endpoints are then discussed in proper sequence. F ecu n d ity A s a parameter c o m m o n l y used in epidemiological studies, time-to-pregnancy (TTP) is a measure of couple fecundity. Ho we ve r, both female and male risk factors m u st be taken into regard, and studies in this vary in regard to the validity of data collected [157]. a. T T P w a s obtained in a Danish study of 1240 w o m e n , w h o had achieved pregnancy, thus excluding infertility. T h e subjects with s e r u m - P F O S in the highest quartile had a 2 6 % reduced chance of b e c o m i n g pregnant within the s a m e cycle m o n t h as co mp a r e d to w o m e n in the lowest quartile [158]. b. A recent Canadian study of over 1,700 w o m e n demonstrated that increasing concentrations of P F C s in serum w e re associated with both reduced fecundability, as measured b y increased time to pregnancy, and infertility [159]. Specifically, an increase in one standard deviation in the s e r u m - P F O A concentration w a s associated with a 3 1 % increase in the odds of infertility and an 1 1 % reduction in fecundability. Adverse effects from P F H x S were similar, with a 2 7 % increase in the odds of infertility and a 9 % reduction in fecundability. c. W h e n m y colleagues and I examined P F C exposures in a prospective study of 2 2 2 Danish first-time pregnancy-planners without previous reproductive experience (129 attained pregnancy within 6 months), w e calculated the fecundability ratio (FR) using 50 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN discrete-time survival models [160]. T h e results s h o w e d little, if any, difference associated with s e r u m - P F C concentrations, although the study m a y have been too small to reveal an effect. d. N o n e of these studies involved cohorts with significantly elevated exposures, such as workers in P F C manufacturing plants or residents of contaminated areas. Still, the available evidence suggests that background exposures to P F C s affect T T P to a limited extent, as suggested b y a N o r w e g i a n study [161]. c. N o n e of the recent epidemiological studies used sophisticated technologies that have b e c o m e available in m o r e recent years. In fact, the waiting-time-to-pregnancy (or time required to conceive) measure relies on a simple questionnaire that has been in use since the 1980s [162]. It is therefore unfortunate that n o studies have been located from major P F C producers regarding fecundity of exposed employees. Puberty development, irregular cycles, and m enopause f. A cross-sectional study of P F O A and P F O S regarding indicators of sexual maturation w a s carried out in the M i d - O h i o River Valley. Participants we re 3 0 76 boys and 2931 girls aged u p to 18 years. T h e y were classified as having reached puberty based on either h o r m o n e levels (total > 5 0 ng/dL and free > 5 p g / m L testosterone in boys, and estradiol > 2 0 p g / m L in girls) or onset of menarche. For boys, there w a s a relationship of reduced odds of reaching puberty with increasing P F O S (delay of 190 days between the highest and lowest quartile). For girls, higher concentrations of P F O A or P F O S we re associated with reduced odds of postmenarche (130 and 138 days of delay, respectively) [163]. This study m a y well have underestimated the effects, as it w a s based on current s e r u m - P F C values only g. A m o r e recent study focused on 2,292 children aged 6-9 years w h o had been examined in 2005-2006 in regard to their exposure to P F O A in the U p p e r O h i o River Valley [164]. In boys, a higher s e r u m - P F O A concentration w a s linked to lower testosterone, and P F O S with lower estradiol, testosterone and insulin-like growth factor (IFG-1); in girls, a higher P F O S w a s associated with decreases in both testosterone and IGF-1. h. In regard to puberty development, a British birth cohort at background levels found that P F O A concentrations in stored maternal pregnancy serum w e re slightly higher for 218 daughters w h o had reached me na rc he before age 11.5 years c o m p a r e d to a similar n u m b e r of controls with later onset [165]. T h e results in this study, however, w e re not statistically significant. i. In a Danish study of prenatal exposures judged from maternal serum analysis, 367 daughters' me na rc he w a s significantly delayed at higher prenatal P F O A exposures [166]. A n important strength is that this study focused on prenatal exposure, although with no adjustment for postnatal exposure from breastfeeding and other sources. j. In 950 pre-pregnant w o m e n , higher serum concentrations of P F O A , P F O S , P F N A , and P F H x S s h o w e d increased odds of self-reported history of irregular menstrual cycle and long menstrual cycle [167], i.e., disruptions that m a y be related to an increased risk of subfecundity [168]. 51 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN k. T h e C 8 Health Project ex am in ed 25,957 w o m e n aged 18-65 years regarding serum estradiol concentrations and onset of m e n o p a u s e [169]. T h e odds of having experienced m e n o p a u s e increased significantly at higher exposures to P F O A and P F O S within the subgroup of middle-aged w o m e n . Semen quality l. A joint analysis of data from three countries suggested a substantially lower proportion of morphologically normal sperm cells at increased serum concentrations of P F O S and P F H x S , while a small increase (opposite direction) appeared to be related to P F O A exposure [170]. Effects on reproductive h o r m o n e s w e re also measured and will be dealt with separately below. m. In a study of 2 5 6 m e n examined at a fertility clinic, n o association between the current serum concentrations of P F O A and P F O S and s e m e n parameters w a s found [171]. Ho we ve r, the concomitant concentrations m a y not reflect the exposures at the m o st vulnerable developmental stage or stages wh e r e negative effects on s e m e n formation m a y have happened. n. In 105 y o u n g Danish m e n from the general population, those with elevated co mb i n e d serum concentrations of P F O S and P F O A had a m e di an sperm count that w a s 2.5-fold lower than the m e di an for m e n with l o w P F O S - P F O A exposures [172]. Other associations were not statistically significant, but suggested altered pituitary-gonadal ho rm o n e s at higher P F O S P F O A exposures. o. F r o m a pregnancy cohort established in D e n m a r k in 1988-1989, about one-third of the m e n (169) w a s recruited at age 20 years to obtain a s e m e n sample and a blood sample [173]. P F O A and P F O S were measured in banked maternal pregnancy serum samples. In utero P F O A exposure w a s associated with lower sperm concentrations and sperm counts, while P F O S did not appear to be associated with any of these outcomes. p. According to a recent review, a total of sixteen studies have explored the association between P F C exposure in m e n and s e m e n parameters, reproductive h o r m o n e levels, or TTP. Despite s o m e w h a t inconsistent results, subtle associations between higher P F O S and lower testosterone or abnormal s e m e n morp ho lo gy have been found in s o m e of the studies and cannot be ignored. Also, eleven studies assessed the association between P F A S exposure in w o m e n and T i m e T o Pregnancy (TTP), as a measure of fecundity, or reproductive ho rm on es levels. Four of eight studies found prolonged T T P with higher P F O S or P F O A , while one of the four found an association w h e n restricting to nulliparous w o m e n [174]. Again, a concern is the time of blood collection for exposure assessment, as adverse effects could be due to, say, pre puberty exposures. M iscarriage q. Available evidence suggests that miscarriage and stillbirth are associated with P F C exposure, although the evidence is not yet strong. This is perhaps not surprising as miscarriage and stillbirth, like mortality, are extreme outcomes. A recent study that included m o r e than 300 miscarriages found a tendency towards a positive association with P F O S exposure 52 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN in the M i d - O h i o River Valley, but n o association between P F O A exposure and miscarriage [175]. r. A subsequent Danish case-control study of 51 miscarriages utilized serum collected in first trimester and found a significantly increased risk associated with C 9 and C10, and a tendency in the s a m e direction for P F H x S , but n o clear association for P F O A and P F O S [176]. This study likely had too l o w a statistical p o w e r to reveal minor adverse impacts. Pre-eclam psia and higher blood pressure during pregnancy s. T h e C 8 Science Panel concluded that P F O A exposure is associated with reproductive toxicity, i.e., an increased risk of pre-eclampsia and higher blood pressure during pregnancy [156]. t. This conclusion rests on extensive studies in the contaminated U p p e r Oh io River Valley. Data w e re obtained on 1,845 pregnancies within the 5 years preceding the serumP F O A analysis and on 5,262 pregnancies analyzed for P F O S . Preeclampsia w a s weakly associated with P F O A and P F O S [177]. u. Ho we ve r, a m o r e recent study is less convincing. Relying on the serumP F C analyses from the health examinations in 2005 and 2006, birth records from singleton pregnancies we re obtained to identify the 106 cases of pregnancy-induced hypertension. S e r u m P F O A and P F O S w e re both positively associated with the diagnosis [178]. \. Using data from the N o r w e g i a n M o t h e r and Child Cohort Study, a study w a s conducted of 976 nulliparous pregnant w o m e n , of w h o m 4 6 6 had a validated diagnosis of preeclampsia. N o strongly positive associations between P F A S levels and preeclampsia in this population with l o w background exposures [179]. Thus, the conclusion today is less clear than it w a s w h e n the C 8 Panel based its conclusions solely on the findings in the highly contaminated communities. Preterm birth and low birth w eight w. T h e C 8 Science Panel also evaluated the evidence on preterm birth, birth weight and fetal growth. S o m e studies available b y then suggested small negative shifts at high P F O A exposures [180, 181], but the Panel considered t h e m uncertain and therefore insufficient to conclude the presence of a probable link. E P A in its m o st recent evaluation considered decreased birth weight in rats one of the critical outcomes for P F O S [149]. x. Others have reached the s a m e conclusion that these reproductive outcomes are too uncertain [182, 183]. Ho we ve r, this conclusion m a y be arguable, given that decreased fetal growth and lowered birth weight are crude measures that m a y only weakly reflect m o r e important effects on growth and functional development. O n the other hand, the association can be affected b y distribution factors, rather than toxicity, and dietary intakes of, e.g., n-3 fatty acids play an important role and is difficult to control for [184] in observational population studies. y. Relating to the C 8 studies, w o m e n w h o reported reproductive histories and w h o provided serum for the C 8 study at the examinations we re linked to data on preterm birth 53 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN and birth weight. Elevated serum concentrations of P F O A and P F O S at the health examination in 2005-2006 w e re associated with a greater frequency of lower birth weight at term [178]. z. A study in D e n m a r k demonstrated increased birth weight in girls at higher exposures to P F O S , P F O A , and P F H x S and reduced birth weight in boys, thereby suggesting sex-dimorphic effects [185]. In support of this notion, the s a m e study also measured the anogenital distance in 511 children and observed decreases in girls, though not in boys, at elevated maternal P F C exposures. aa. A s mentioned above, the C 8 Science Panel evaluated the evidence available at that time on preterm birth, birth weight and fetal growth. S o m e studies suggested small negative shifts at high P F C exposures [180, 181], but the Panel considered t h e m too uncertain and therefore insufficient to support a probable link. hh. Others have reached the s a m e conclusion that these outcomes are uncertain [182, 183]. Ho we ve r, this conclusion m a y be arguable, given that decreased fetal growth and lowered birth weight are measures that, although crude, m a y reflect m o r e important effects on functional development. O n the other hand, the association m a y be affected b y a variety of other factors, in addition to fetal toxicity. cc. In a recent study, birth weight in N o r w a y w a s apparently not affected by background levels of P F C exposures [186]. dd. O n the other hand, the m o s t recent report from the Japanese H o k k a i d o cohort sh o w s that l o w background exposures to P F O S and P F O A are associated with decreases in birth weight, and the study also highlights that h o r m o n e s such as leptin and adiponectin m a y play a role [187]. ec. A British study of the A L S P A C birth cohort collected serial data on weight and height up to age 20 m o nt hs and s h o w e d that elevated maternal serum concentrations of P F O S , P F O A and P F H x S we re associated with decreased birth weights in girls but that higher P F O S exposures w e re then associated with increased b o d y weight at 20 mont hs [188]. ff. A similar study from the Faroes revealed that a higher maternal pregnancy s e r u m - P F O S concentration w a s associated with increased weight (and overweight) in the child at age 18 months, while P F O A rather s h o w e d a similar association with weight at 5 years of age [189]. These findings suggest that birth weight as an o u tc om e at a particular point in time m a y need to be seen as part of an intrauterine-postnatal growth profile. gg. N o n e of the recent studies of birth weight and postnatal growth involved cohorts with significantly high parental P F C exposures, such as workers in P F C manufacturing plants or residents of contaminated areas, even though studies might have been easy to carry out. Overall, the available evidence suggests that fertility and pregnancy outcomes m a y not be highly sensitive targets of P F C toxicity, at least not easily detectable in the presence of m a n y other determinants, but the evidence suggests minor adverse effects at elevated levels of background exposure. 54 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN lih. I a m aware of Professor Da v i d Sunding's statistical analysis in connection with this case showing a pattern of increased incidences of premature births, l o w birth weights, and decreased general fertility rates in Minnesota communities that I understand are, and have been, m o s t affected b y PFC-contaminated ground water. His findings are consistent with the evidence summ ar iz ed above showing associations between P F C exposure and adverse effects on h u m a n reproductive functions, in particular risks of adverse effects on fetal growth. D evelopm ental defects and delays Early indications of potential reproductive h a r m from P F C s c o m e s from 3 M 's and D u P o n t 's o w n experience. S o m e incidents we re rather anecdotal and could have been r a n d o m observations, but the evidence is still noteworthy and troubling. It is supported b y toxicological evidence from animal studies, as discussed below. ii. O n e of the first indications of developmental toxicity from P F C s w a s the observation that t w o of seven children born between 1979 and 1981 to female workers at a D u p o n t P F C plant had birth defects. O n e of the D u P o n t employees with heavy exposure to P F O A had a miscarriage followed b y a normal pregnancy. Given the sample size, this m a y have been a chance event, but such occurrences should clearly trigger intensive surveillance and toxicology follow-up. 3 M revealed that they planned to review in detail an experimental teratogenesis study and if a positive finding w a s made, w o m e n of childbearing potential w o u l d be r e m o v e d from jobs with exposure.zzz jj. A September 15, 1981 letter from D u P o n t 's Washington W o r k s to the Shimizu Corporation in Japan proposed an employee blood sampling program. T h e letter had an attachment on pregnancy outcomes a m o n g female production workers, according to wh i c h one child had unconfirmed eye and tear duct defect. A younger child had one nostril and an eye defect,aaaa both exceedingly rare and very unlikely to occur a m o n g as f e w as seven births. kk. In D e c e m b e r of that year, t w o D u P o n t female employees raised questions on the status of the C 8 teratogenesis study; one of the w o m e n had a child with congenital eye defects and wanted to k n o w if the 3 M studies found any malformations affecting the e y e bbbb A s noted below, the studies did in fact reveal developmental effects from P F C s on the eyes of rodents. Apart from malformations, developmental toxicity in a m o r e general sense m a y affect several organ systems and is sometimes reported in connection with reproductive toxicity. M y review of available epidemiological data on developmental toxicity covers these aspects relating to the individual organ systems. T h e overall impression from these data is that only zzzAR226-1375. March 25, 1981 - DuPont's Medical Director, Dr. Bruce Karrh, summarized the birth defect data received from 3M and DuPont's knowledge of the pregnancy outcome status of Washington Works employees exposed to C-8. page 000098. aaaaAR226-1390. September 16, 1981 - A DuPont employee updated by hand DuPont's May 14, 1981 chart summarizing birth and pregnancies among female Washington Works employees to incorporate the C-8 blood results received in July of 1981. (Exhibit R (EID079371-5)). page 000175. bbbbAR226-1393. December 18, 1981 - Questions on C-8 stats report (Ref: C-8 (FC-143) status report, December 15, 1981 (Exhibit U (EID07954)). page 000180. 55 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN suggestive evidence surfaced during the early years, i.e., b y about 1980, and it took 20-25 years before m o r e detailed evidence w a s published from studies of developmental toxicity in the general population with P F C exposure. 2. Toxicological evidence A significant part of the early toxicological evidence concerning reproductive h a r m from P F C s c o m e s from 3 M 's and D u P o n t 's o w n studies. M o r e extensive information on reproductive toxicity in animal models are available in recent reviews [4, 6]. a. A n experimental study conducted by 3 M in 1981 s h o w e d birth defects in eye lens of rats exposed to P F O A (the report w a s filed as a T S C A 8e d o c u m e n t with the U.S. E P A ) [90]. In fact, a total of three teratology studies we re carried out, all of t h e m finding lens abnormalities in exposed animals. In M a r c h 1981, 3 M informed D u P o n t of the rat study, and D u P o n t then r e m o v e d all female employees from C 8 exposed jobs, although without telling t h e m cccc b. Later, Dr. E. G. Lamprecht, a 3 M consultant, argued that the fetal rat lens abnormality w a s incorrectly interpreted as a teratogenic change and that later studies could not repeat the finding of a teratogenic effect.dddd T h e following year, 3 M 's E. Marshall Johnson, in a letter to William C. M c C o r m i c k , III agreed with the position that the teratogenic finding in the teratology studies on P F O S , ethyl F O S E , and P F O A w a s an artifact.eeee A 3 M E m p l o y e e Co mm u n i c a t i o n therefore reported that no teratogenic effect w a s found.ffff M u c h later, in 1998, Dr. M a r v i n Case, a 3 M corporate scientist, reviewed the teratology studies and also agreed that "neither F C 95 nor F C 143 causes teratogenic effects in animals w h e n dosed at levels wh i c h are maternally toxic. ... T h e lens change observed in rat pups in Riker Laboratories studies w a s allegedly a sectioning artifact and w a s not found u p o n repeat studies at independent laboratories."gggg Still, in m y opinion, the very rare eye and facial malformations in children born to highly-exposed female P F C production workers are unusual and highly worrisome, even though experts hired b y 3 M called the eye malformation in the dosed animals an artifact - an unusual critique of a highly-experienced test laboratory. I find this explanation unusual and rather far-fetched. In addition, congenital eye defect concerns also arose in later rodent studies commissioned b y 3 M , and it seems from m y reading of the files that they have not been explained away. H o we ve r, as m y focus is on the epidemiological studies, I have not attempted to locate all documents on this issue. ccccAR226-1378. March 31, 1981 - DuPont notified its employees that all female workers would be removed from jobs "Where there is potential for exposure to C-8" at DuPont's Washington Works. In standby questions and answers for those employees. page 000116. ddddAR226-0287. Memorandum to Riker Study Files: Fetal Rat Lens Artifact -- Summary of Developments to Data. page 004333. eeeeAR226-0530. Letter from E. Marshall Johnson to William C. McCormick, III re Riker teratology studies on PFOS and ethyl FOSE, dated November 12, 1982. page 004139. ffffar226-1395. March 1, 1982 - C-8 (FC-143) Employee Communication (Exhibit W (EID089464)). page 000183. ggggAR226-0468. 3M Memo re FC 95 and 143 Teratology, from Marv Case to Georjean Adams, dated May 26, 1998, enclosing an analysis of the eye lens defect. page 003116. 56 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN c. In regard to P F B A , a recent study on reproductive effects in mice reported full-litter loss only at the very highest exposure and a significant delay in eye-opening in the offspring in all three P F B A groups [190]. In the s a m e study, the age at wh i c h the m o u s e offspring reached puberty w a s determined b y monitoring vaginal opening in females, and significant, dose-dependent delays w e re observed. T h e M D H relied on these effects w h e n determining a limit for P F B A concentrations in ground water [63]. Subsequent sections will focus on endocrine disruption, including se rum-hormone concentrations and breast development that appear to be critical adverse effects in humans, while also docu me nt ed in laboratory animals. 3. Perspective T h e C 8 Science Panel did not conclude that there is a probable link between exposure to P F O A and birth defects [191], and m o r e recent evidence has not added further support for such linkage. A link only to pre-eclampsia w a s considered sufficiently justified by the Panel [156]. A t the time, a link to decreased birth weight w a s not found, but m o r e recent evidence suggests that the evidence on decreased fetal growth should consider time-dependent growth patterns. T h e C 8 Panel did not look as broadly at reproductive and related developmental issues, given that consideration needed to be given to the findings that could be m a d e based on the use of methodologies and population groups available at the time. Recent evidence is highly suggestive of adverse effects on female reproduction, as indicated b y increased occurrence of puberty development, irregular menstrual cycles and decreased fecundity. M a l e toxicity is also m u c h better docu me nt ed now, although exposure misclassification needs to be carefully considered. Therefore, based on the weight of all the evidence, that P F C s pose a substantial present and potential hazard to h u m a n reproductive system functions. C. Endocrine disruption O u t c o m e s usually considered to reflect endocrine disruption will be dealt with in the present section, including changes in reproductive h o r m o n e concentrations in serum and inhibited lactation as indicated b y shorter breastfeeding durations in exposed w o m e n . It is m y opinion, based on the weight of the epidemiological evidence, and supporting toxicity evidence, that P F C exposure at levels similar to or b e l o w those reported from the East Me t r o area pose a substantial present and potential hazard to h u m a n endocrine functions. Endocrine disruption effects are usually defined as adverse effects in an intact organism or its progeny that have an endocrine m o d e of action, i.e., that it alters the function(s) of the endocrine system. D u e to the serious h u m a n health consequences, endocrine disruption has b e c o m e a top priority in chemicals control efforts in the E U and elsewhere [192]. 57 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN Ba s e d on the available evidence, P F C s are convincingly associated with endocrine disrupting effects that m a y have substantial impacts on vulnerable population groups. Wh i l e endocrine disruption is often thought to be related to reproductive toxicity, a wi de variety of h o r m o n e s play a role for various physiological functions, and their disruption can cause a variety of dysfunctions and diseases. H o r m o n e s addressed b y P F C research include sex hormones, thyroxin, and insulin. T h e serum concentrations of s o m e h o r m o n e s vary during the day or with food intake, and variability of sampling times and lack of statistical control for such factors m a y hide or attenuate PFC-associated changes in h o r m o n e concentrations. 1. Epidemiological evidence Serum -horm one concentrations a. Early evidence on endocrine disruption associated with P F C exposure originates from a doctoral thesis project, wh e r e Frank Gilliland, M D , studied clinical pathology parameters in 111 male workers in 3 M 's Chemolite plant in Cottage Grove, M N . There w a s a positive correlation between P F O A exposure measured as serum total organic fluorine and estradiol (an adverse effect), and a negative correlation with free testosterone (also an adverse effect) with this association being stronger in older men. Dr. Gilliland therefore concluded that P F O A m a y affect male reproductive hormones.hhhh This study w a s not reported on its o w n in a scientific journal, but w a s referenced in a subsequent article led b y 3 M authors [88]. b. This subsequent follow-up study further explored serum h o r m o n e abnormalities in exposed workers and likewise s h o w e d a positive correlation between P F O A exposure and serum-estradiol (an adverse effect) [88] in 111 and 80 production workers studied in 1993 and 1995. T h e 1 0 % increase in m e a n estradiol levels observed a m o n g those employees with the highest s e r u m - P F O A concentration w a s argued to be potentially confounded b y b o d y m a s s index (although the risk of obesity m a y be increased at higher P F C exposures, see section VII.E). Despite the fact that t w o sets of data w e re available, and 68 participated in both (and s o m e likely were also ex am in ed b y Dr. Gilliland), the authors chose not to conduct comparisons over time, allegedly due to variability of the h o r m o n e analyses. T h e 3 M authors concluded that the results provided reasonable assurance that, in this production setting, and contrary to the directionality of Dr. Gilliland's findings, there w e re n o significant hormonal changes associated with P F O A at the serum levels measured. A m e m o from Dr. G. Olsen in 1998 proposes that the h o r m o n e concentrations w e re affected b y misclassified b o d y m a s s index as a confounder.*iiii Interestingly, an a n o n y m o u s reviewer w h o assessed the manuscript for a major occupational health journal questioned w h y the authors w o u l d present scatterplots for P F O A and s o m e endpoints that were all non-significant, but not for the estradiol/testosterone results, wh i c h were statistically significant.jjjj c. T h e C 8 Health Project examined 25,957 w o m e n aged 18-65 years regarding serum estradiol concentrations [169]. There w a s a significant inverse association between P F O S and estradiol, though not between P F O A and estradiol, thereby suggesting that hhhhAR226-0473. Frank Davis Gilliland, Fluorocarbons and Human health: Studies in an Occupational Cohort (October 1992) (unpublished Ph.D. thesis, University of Minnesota), with Summary. Page 003246. 11113MA00652081.Memo to file. Geary Olsen. 1/15/98. jjjj 3MA00630994. Reviewer 1, Occupational And Environmental Medicine 1997/187, page 3MA00630995. 58 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN endocrine effects from P F C exposure m a y differ between m e n and w o m e n . A s mentioned in section B, there w a s also an increased odds of having experienced m e n o p a u s e at elevated exposures to P F O A and P F O S a m o n g study participants. d. In a study of nearly 2,300 children living near a P F O A production facility in the M i d - O h i o River Valley, increased P F C exposure w a s found to be correlated with lower levels of sex hormones. Especially in boys, increased P F O S concentrations w e re associated with lower testosterone, estradiol, and IGF-1 levels, and increased P F O A concentrations were correlated with lower testosterone levels. In girls, increased P F O S concentrations were associated with lower testosterone and IFG-1 levels [164]. Again, this study supports the notion that the P F C s are endocrine disruptors and that effects m a y differ also b y age. c. A study of postpubertal w o m e n at age 15 years w h o s e mothers were exposed to P F C s at background levels in the U K found that higher levels of maternal exposure to P F O S , P F O A , and P F H x S w e re correlated with higher testosterone concentrations. Findings from this study suggest that prenatal exposure to P F C s leads to adverse effects that m a y be lasting and m a y be expressed during or after puberty [193]. Again, differential effects m a y be observed in regard to different developmental stages. f. A study of 540-person cohort in T a i w a n found that increased serum concentrations of P F O A and P F O S we re correlated with decreased levels of sex h o r m o n e s in adolescents and y o u n g adults at ages 12-30 years [194]. In particular, P F O S w a s associated with a significant decrease in follicle-stimulating h o r m o n e ( F S H ) levels in y o u n g m e n aged 12-17, and in serum testosterone levels in y o u n g w o m e n of the s a m e age. P F O A w a s associated with a significant decrease in serum levels of sex-hormone binding globulin ( S H B G ) in the y o u n g w o m e n aged 12-17 years, and negative associations between P F C exposures and the ho rm o n e s measured w e re particularly strong in the y o u n g w o m e n g. F r o m the Danish pregnancy cohort established in 1988-1989, the 169 m e n at age 20 s h o w e d higher adjusted levels of luteinizing h o r m o n e ( L H ) and folliclestimulating h o r m o n e (F S H ) associated with higher prenatal P F O A exposures [173]. P F O S did not appear to be associated with h o r m o n e concentrations. h. In 105 y o u n g Danish m e n at background exposures, h o r m o n e profiles suggested poorer function of Leydig cells (which produce testosterone) at higher P F C exposures. Ho we ve r, the associations in this small study w e re not statistically significant [172]. Duration o f breastfeeding i. A study of 1,400 Danish w o m e n reported that the duration of breastfeeding, as recorded b y t w o telephone interviews, decreased at increasing serumconcentrations of P F O A and P F O S , although only in multiparous w o m e n [195]. In multiparous w o m e n , previous breastfeeding might confound the association, and this finding therefore did not provide strong support for a causal association. j. A recent study in the U.S. [196], however, found that increased maternal s e r u m - P F O A concentrations w e re correlated with a decreased duration of breastfeeding, and that 59 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN this association w a s not confined to multiparous w o m e n and is independent of potential confounders, thereby supporting a hypothesis of endocrine disrupting effects. Although the recording of duration of exclusive breastfeeding m a y have been s o m e w h a t imprecise, the fact that the w o m e n we re not aware of their o w n exposure levels excludes any important bias. k. These findings are supported b y a subsequent study of 1,130 n e w mothers in the Faroe Islands [197]. A doubling of maternal serum P F A S concentrations w a s associated with a reduction in duration of both total and exclusive breastfeeding, m o s t pronounced for P F O S , wh e r e a doubling w a s associated with a reduction in total breastfeeding of about six weeks. Similar effects were seen for P F O A , though not for P F H x S . These associations were evident a m o n g both primiparous and multiparous w o m e n , and thus cannot be explained by confounding from previous breastfeeding. 2. Toxicological evidence Endocrine disruption effects in h u m a n s are supported b y a substantial n u m b e r of experimental animal studies [4, 7, 9, 148, 149]. A f e w key studies are highlighted below. a. A n early study of the effects of A P F O (the a m m o n i u m salt of P F O A ) exposure in rats s h o w e d a substantial increase in hepatic aromatase activity [198]. A n increase in aromatase activity is likely to increase the formation of estradiol from testosterone, thus a decrease in serum-testosterone and increased estradiol. b. Accordingly, changes in serum concentrations of testosterone and estradiol are considered likely to be due to PF C-mediated changes in the hepatic aromatase activity [199], but interference with sex h o r m o n e receptors has also been reported [200]. Such m o d e s of action could well mediate the PFC-associated endocrine disruption findings in epidemiology studies. c. A recent study ex am in ed the effect of P F O A and P F O S exposure on proteins and cells related to the male reproductive system and demonstrated that both P F O A and P F O S inhibit important drug transporting proteins present in the blood-testis barrier, thereby potentially contributing to male infertility [201]. d. Endocrine disruption effects appear to be independent of P P A R activation and therefore are likely relevant to h u m a n P F O A toxicity [7]. A m o n g reported mechanisms, P F O A can activate nuclear receptors other than P P A R , i.e., the constitutive androstane receptor ( C A S ) and the pregnane X receptor (PXR), and activation of the estrogen receptor (ER) m a y also be involved [202, 203]. Local testicular effects are indicated b y induction of Leydig cell hyperplasia and a d e n o m a in experimental studies, apparently independent of P P A R activation [17]. c. A toxicological study designed to evaluate the impact of P F C exposure on androgen secretion in the testis examined the effects of P F C exposure on inhibition of h u m a n and rat microsomal enzymes. T h e study demonstrated that P F O S is a potent inhibitor of h u m a n 17P-hydroxysteroid dehydrogenase 3 e n z y m e in testicular cells. T h e results support the conclusion that P F O S exposure adversely affects reproductive hormones, as the e n z y m e 60 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN inhibition demonstrated in this study m a y well contribute to decreased androgen secretion in the testicles [204]. f. Experimental studies s h o w that m a m m a r y gland development in mice is inhibited b y P F O A exposure during early development at serum concentrations similar to those occurring in the East Me t r o residents [7, 9, 205]. T h e State of N e w Jersey regards this as one of the mo st sensitive non-carcinogenic endpoints [206] (see section VIII). 3. Perspective a. Regarding endocrine disruption, substantial research activity has emerged during the last 20 years or so, and m u c h improved toxicologic understanding of the m e c h a n i s m s involved has resulted. In addition, a wealth of epidemiological studies has docu me nt ed the adverse h u m a n health consequences [192]. Although the studies of serum-hormone concentrations at 3 M coincided with early discoveries on endocrine disruption, this issue w a s apparently not treated as a priority for further studies at the time. b. A s mentioned under C.1.b, the paper co-authored b y 3 M scientists and Dr. Gilliland in 1998 reported on clinical pathology results from serum analyses, including reproductive h o r m o n e s and concluded that there w e re n o significant hormonal changes associated with P F O A at the serum levels measured [207].kkkk Wh i l e this conclusion w a s counter to Dr. Gilliland's findings in his thesis project, the data analysis in the published paper seems inadequate. O n e difference between the 1998 and Dr. Gilliland's thesis is that Dr. Gilliland relied on total organofluorine concentrations, and the subsequent study referred to P F O A , but this issue w a s not explored in the published paper. Ho we ve r, in September of 1998, Dr. Olsen wrote a proposal for a study titled "fluorochemicals exposure assessment of Decatur Chemical and Film plant employees." According to this document, "approximately 1/3 of these film plant employees had past w o r k history experience in the chemical plant." ml Thus, truly low-level exposures were difficult to locate. T h e article lists the lowest s e r u m - P F O A level as "0.00 p p m , " i.e., b e l o w 10 ng/mL, thus reflecting a high detection limit that w o u l d not allow analysis of current-day exposures of the general population. c. It is worth noting that all or most of the studies above used only concurrent blood samples for P F C analysis, and less or n o information therefore is available on the impact of prenatal exposures. In one of our recent prospective studies focused on P C B s , another pollutant also suspected of causing endocrine disruption including Leydig cell dysfunction, the results s h o w e d that changes in h o r m o n e concentrations we re correlated with P C B concentrations in cord blood. Ho we ve r, only scattered relationships we re observed in relation to P C B concentrations in concurrent blood samples [208]. There is reason to believe, therefore, that P F C levels in cord blood could s h o w even stronger correlations than those s h o w n kkkkAR226-0474. Geary W. Olsen, et al., An Epidemiological Investigation of Reproductive Hormones in Men with Occupational Exposure to Perfluorooctanoic Acid, 40 JOEM 614-619 (July 1, 1998), with Summary of study. page 003468. llll AR226-0950. Fluorochemical Exposure Assessment of Decatur Chemical and Film Plant Employees (Data Summary, Protocol and Final Report). page 001249. 61 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN in the studies above using concurrent blood samples, showing increased adverse effects from P F C s on h u m a n health during vulnerable developmental time windows. d. S e r u m - h o r m o n e changes observed in relation to s e m e n quality suggest that they are both related to endocrine disruption mechanisms, as also indicated b y experimental studies in laboratory animals. A concern is again that mo st studies have relied u p o n P F C concentrations measured in serum obtained at the w r o n g time, i.e., not regarding the mo st vulnerable developmental stage. Such exposure misclassification will likely result in an underestimation of the true P F C effects [101]. Thus, in m y opinion, results showing adverse effects could be even stronger during vulnerable developmental time w i n d o w s is also supported b y our studies on prenatal exposure to pesticides [208-210]. Further, the findings m a y be of greater public health relevance, given the k n o w n excess mortality associated with decreased serum-testosterone concentrations in m e n [211]. D. Thyroid hormones and related diseases Although thyroid function is usually dealt with as part of the endocrine spectrum of functions, P F C effects on the thyroid gland appear to be independent of other forms of endocrine disruption and are therefore dealt with separately. T h e s a m e is the case regarding insulin and metabolic abnormalities (see section VII.E). For the sake of clarity, these specific gland-related effects are dealt with in separate sections. It is m y opinion, based on the weight of the epidemiological evidence and supporting toxicity evidence, that P F C s pose a substantial present and potential hazard to h u m a n thyroid functions, with related adverse effects. Ba s e d on the available evidence, the thyroid gland appears to be a target organ for P F C toxicity, as supported b y laboratory experimental data [212] and recognized b y the C 8 Panel [213]. T h e Panel report is quite comprehensive, and the coverage of thyroid dysfunction will therefore emphasize major and recent studies. T a k e n as a whole, P F C exposure at levels similar to or b e l o w those reported from the East M e t r o area are associated with thyroid disruption effects. E v e n subclinical hypothyroidism is a health concern, especially during pregnancy, as fetal brain development is highly vulnerable to deficiency in maternal thyroid h o r m o n e supplies [214, 215]. Moreover, as the thyroid gland is the target for a substantial n u m b e r of environmental chemicals, P F C s are likely to contribute to joint effects in combination with exposures to other thyroid toxicants [216]. 1. Epidemiological evidence a. Dr. Gilliland analyzed in his doctoral thesis data from his cross-sectional study of 3 M production plant workers regarding thyroid effects associated with organofluorine concentrations in serum. A positive correlation w a s seen between organic fluorine and the thyroid stimulating h o r m o n e ( T S H ) in serum.m m m m Elevated T S H is often seen w h e n thyroid mmmmAR226-0473. Frank Davis Gilliland, Fluorocarbons and Human health: Studies in an Occupational Cohort (October 1992) (unpublished Ph.D. theses, University of Minnesota), with Summary. Page 003247. 62 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN functions are deficient. In a later paper on thyroid function measurements in about 500 workers from 3 M production plants in A l a b a m a and Belgium, 3 M scientists argued that variable associations with thyroid ho rm o n e s (and all other clinical pathology parameters measured in this study) w e re of limited, if any, clinical relevance [217]. Attention is d r a w n to lifestyle factors, while sex-dependent associations, m o r e sophisticated exposure classification, duration of exposure, and additional data available from Minnesota we re clear weaknesses. b. A cross-sectional data set from the C 8 Health Project on 52,296 adults with a year or m o r e of exposure to contaminated drinking water s h o w e d that both P F O A and P F O S in serum w e re associated with significant elevations in serum thyroxine (T4) and a significant reduction in T 3 uptake in all participants, thus showing disruption of thyroid functions [218]. c. A later study of 33,254 exposed c o m m u n i t y m e m b e r s and production workers applied calculated temporal trends in s e r u m - P F O A concentrations [219]. T h e occurrence of 2,109 cases of functional thyroid disease, i.e., hyperthyroidism and hypothyroidism, w a s associated with P F O A exposure in w o m e n , while exposed m e n s h o w e d a tendency of hypothyroidism at elevated exposures. d. In 10,725 children and adolescents aged 1-17 years ex am in ed within the C 8 c o m m u n i t y study, a tendency w a s seen toward an increased risk of an increased serum concentration of total T 4 concentrations w a s found for P F O S , but w a s not significant for P F O A . Further, an increased odds ratio for hypothyroidism (observed in 39 cases) w a s found at the highest quartile of P F O A exposure [220]. c. In a study based on N H A N E S data on 3,974 adults, serum concentrations of P F O A and P F O S we re co mp a r e d between subjects with and without self-reported thyroid disease [221]. W o m e n with a s e r u m - P F O A concentration in the highest quartile w e re m o r e than twice as likely to report current treated thyroid disease co mp a r e d to w o m e n with l o w P F O A levels. T h e s a m e tendency w a s seen in men, although it w a s of borderline significance. For P F O S , the trend w a s significant in men, but not in w o m e n . f. In a similar N H A N E S - b a s e study of 1,181 adults, higher serum concentrations of P F O A w e re associated with increased serum concentrations of T3, while P F H x S w a s linked to increases in both T 3 and T4, but to lower T 4 in m e n [222]. These findings suggest sex-dimorphic effects of P F C s on thyroid functions. 2. Toxicological evidence a. Thyroid dysfunction in h u m a n s exposed to P F C s is supported by a wide range of studies in laboratory animals [4], and the sensitivity of thyroid functions to environmental chemicals is well d o cu me nt ed [223]. M o s t of the evidence regards P F O S and P F O S , with s o m e studies also covering P F H x S and longer-chain PFCs. b. Regarding P F B A , an experimental study reported morphological changes in the thyroid gland and decreased serum concentrations of total T 4 [224]. T h e M D H relied on 63 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN this study and additional information from first author Dr. J.L.Butenhoff from 3 M (see section VIII) [63]. 3. Perspective a. Although the very detailed report from the independent C 8 Panel found a probable link between P F C s and adverse thyroid function effects [213], even the fairly recent reports from 3 M purported to s h o w n o such link [217, 225-227]. b. Dr. Gilliland reported a positive correlation between P F O A (measured as serum total organic fluorine) and the thyroid stimulating h o r m o n e ( T S H ) in serum.nnnn A s mentioned above (section D.1.a), 3 M scientists argued in a later paper that variable associations caused any thyroid effects to be of limited, if any, clinical relevance [217]. Ho we ve r, w h e n m o r e recent findings pointed again to adverse effects on the thyroid gland, the s u m m a r y from an internal 3 M w o r k s h o p on research prioritization said: "U n a n s w e r e d questions, h o w to explain a w a y the findings: m e as ur em en t issue, thyroid e c o n o m y - long term effects/ consequence?"TM TM Unfortunately, the d o c u m e n t does not describe h o w findings could be or we re explained away, although the intention is obvious. c. Recent evidence suggests that thyroid toxicity is of particular relevance in pregnant w o m e n with a pre-existing thyroid dysfunction [228]. This issue has not yet been explored regarding P F C effects, and the evidence so far does not allow any conclusion on impact of P F C exposure in the presence of borderline iodine deficiency. In addition, sex-dimorphic differences in thyroid vulnerability to P F C exposures need to be characterized. In general, the lack of assessment of P F C exposures at the mo st vulnerable time, lack of assessment of contributing factors, and other determinants will tend to w e a k e n any true association between P F C exposure and adverse effects on the thyroid gland. E. Insulin and diabetes Ba s e d on the weight of the epidemiological evidence and supporting toxicity evidence, I conclude that P F C s pose a substantial present and potential hazard to h u m a n health regarding increased risks of developing diabetes and metabolic disease or dysfunction. These outcomes include an increased risk of developing overweight or obesity. A s these conditions are c o m m o n and increasing in prevalence, even a small increase in diabetes risk and obesity w o u l d be of major public health impact [229, 230]. Although the C 8 Science Panel concluded that the evidence available to t h e m at the time w a s insufficient to conclude that P F O A is linked to the development of type 2 diabetes (T 2 D ) [231], recent evidence suggests that P F C exposure has a potential of causing adverse metabolic effects, including the development of type 2 diabetes (T 2 D ) and obesity, as part of 226-0473.pdf, Frank Davis Gilliland, Fluorocarbons and Human health: Studies in an Occupational Cohort (October 1992) (unpublished Ph.D. theses, University of Minnesota), with Summary. Page 003173. mm 3MA00432615. Results of Research Prioritization Workshop / PFOS Research Prioritization Workshop Outcomes June 2-3 2005. Page 3MA00432616. 64 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN current exposures to so-called obesogenic chemicals [232], especially in regard to developmental exposures [233]. 1. Epidemiological evidence I have not been able to identify early studies on metabolic changes associated with P F C exposures. T h e global distribution of P F C s and the emerging obesity epidemic and increasing T 2 D occurrence might have inspired m o r e studies in this field. Glucose metabolism parameters are likely m o r e sensitive and precise than diagnostic records and death certificates. Bu t m a n y of the existing studies rely on the latter. Thus, PFC-induced insulin insensitivity needs attention, in particular in prospective studies. a. In regard to diabetes, an increased mortality due to this diagnosis has been reported in PF C - e x p o s e d workers [73, 76], and the increase w a s significant in comparison with unexposed c o m p a n y controls. This increased risk is noteworthy, as diabetes itself is usually not a primary cause of death. Ho we ve r, functional studies of exposed workers, e.g., utilizing oral glucose tolerance testing, have not been identified. b. In a c o m m u n i t y study, P F O A exposure in 54,000 adults w a s not found to associated with development of T 2 D , and neither w a s the fasting serum glucose concentration associated with the exposure [234]. c. In an extension of this study, the C 8 Science Panel did not find any indication that P F O A exposure w a s related to diabetes [219] [231]. Ho we ve r, as diabetes as a cause of death on death certificates is not a reliable w a y of obtaining information on the occurrence of this disease, the absence of clear associations should not be regarded as proof that diabetes is not associated with P F C exposure. In addition, data from the study suggest that fasting serum insulin decreases at higher P F O A exposures [235], and these results w o u l d need to be explored further in regard to the possible effects of cofactors, such as age and age-dependent exposure profiles. d. Additional support for background P F A S exposure being a risk factor for T 2 D c o m e s from certain cross-sectional studies such as the N H A N E S data [236, 237]. Given that the data are cross-sectional, the validity of these results is s o m e w h a t uncertain, and they m a y likely underestimate the true diabetogenic impact of P F C exposure. c. In a study in Taiwan, P F O A in adults w a s positively correlated (adverse effect) with their beta cell function (possibly as a sign of compensation for insulin resistance), and P F O S w a s positively correlated (adverse effect) with blood insulin, insulin resistance (homeostasis mo d e l assessment), and beta cell function [236]. f. Data on 4 9 9 prepubertal children [238] s h o w e d that current exposures to P F C s are linked to increased risk of overweight and deficient glucose homeostasis. In 811 children from the Danish national birth cohort, prenatal exposure to P F O A and P F O S did not s e e m to be associated with height and weight at age 7 [239], but the validity of the P F C measurements has later been called into doubt [100]. 65 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN g. Additional evidence exists that P F C exposure m a y be related to overweight, wh i c h is also a risk factor for T 2 D diabetes. Thus, P F O A in stored pregnancy serum from a birth cohort of 42 2 subjects examined again at age 20 years w a s positively correlated (adverse effect) with b o d y m a s s index ( B M I ) and other indicators of obesity [240]. h. Similar findings we re reported from a birth cohort of 1,006 children in Boston [241], wh e r e higher prenatal exposure to P F O A , P F O S and P F H x S w a s associated with higher B M I , skinfold thicknesses and D X A assessment of total b o d y fat, although only in girls at 7-8 years of age. Additional support derives from a birth cohort study that relied on joint data from Greenland and the Ukraine [242]. i. In Faroese children, born in 2007-2009, maternal pregnancy serum concentrations of P F O S and P F O A we re associated with increased B M I and/or overweight risk at age 5 years [189]. j. In a recent review of the epidemiological evidence, the joint impact of P F C s and other endocrine disruptors on overweight and obesity is highlighted, with the recommendation that future studies should examine multiple exposures, rather than one at the time [243]. This is indeed a concern, as exposures usually include m o r e than one P F C , and often m a n y other toxicants. 2. Toxicological evidence a. Given the epidemiological and experimental evidence (described below), w e have used computer modeling to ascertain links between T 2 D and environmental chemicals, as recently described [244]. W e first identified genes and proteins k n o w n to be involved in the pathogenesis of T 2 D and obesity development and then analyzed interactions with proteins and protein-protein interactions to establish links between environmental chemicals and T 2 D . Both P F O S and P F O A appeared to trigger m e c h a n i s m s associated with T 2 D risk (unpublished results), but whether the mo d e l e d interactions with the diabetes-related genes and proteins result in disease requires confirmation. Nonetheless, this finding supports the notion that P F C s should be considered as likely diabetogens. b. Re vi ew s of experimental evidence on obesogenic and diabetogenic chemicals cover the evidence in greater depth [232, 245], also regarding developmental exposures [233]. c. Ro de nt studies have demonstrated that glucose homeostasis is adversely affected b y P F C exposure [246]. d. Wh i l e P P A R activation m a y play a role, the adverse o u tc om e pathway for development of T 2 D m a y also involve PP AR -i nd ep en den t pathways, such as activation and disruption of the hepatocyte nuclear factor 4 a [247]. c. That the pancreas m a y be a target organ for P F C toxicity is also suggested b y an increased occurrence of pancreatic cancer in P F C - e x p o s e d rodents [17]. 66 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN f. In mice, prenatal exposure to P F O A induced elevated serum concentrations of leptin and insulin, and overweight in mid-life [78]. g. A separate study demonstrated that prenatal exposure to P F O A affected expression of genes involved in the control of lipid and glucose metabolism [235]. 3. Perspective A s diabetes develops over m a n y years and m a y even be p r o g r a m m e d prenatally [248], the evidence available needs to be considered in regard to the validity of exposure parameters that m a y not reflect the P F C exposure at the relevant age or developmental stage, wh e r e disease initiation m a y have begun. Likewise in regard to obesity, approximately one-third of adults and 1 7 % of youth in the United States are obese [249]. Recent prospective studies of birth cohorts suggest that early-life exposures impact anthropometric measures in childhood and adolescence. M u c h attention is currently being paid to these issues regarding P F C exposures, but in birth cohorts and in prospective studies of adult populations. I a m aware of a study carried out in A m e r i c a n nurses, currently under publication, that s e r u m - P F C concentrations measured in serum samples collected in the 1990s w e re associated with an increased risk of subsequently developing T 2 D (Qi S u n et al.). Another study based on the P O U N D S Lost intervention study, s h o w e d that participants with higher s e r u m - P F C concentrations had greater difficulty maintaining the weight loss induced b y calorie restriction than those with lower exposures (Liu et al.). Studies like that are likely to substantially extend the data base to evaluate the adverse impacts of P F C exposures. Recent information suggests that obese persons are healthier and live longer n o w than in previous decades, likely because of better care and risk-factor m a n a g e m e n t [250]. Ho we ve r, the joint public health effect of increased prevalence and decreased mortality leads to m o r e years spent with obesity and m o r e time for the d a m a g i n g coexisting illnesses, such as type 2 diabetes and chronic kidney disease, to develop. F. Neurobehavioral functions M y above conclusions in regard to adverse P F C effects on thyroid functions is relevant also in regard to possible developmental neurotoxicity, as ho rm o n e s are crucial to brain development [251], in particular the thyroid h o r m o n e [215]. A recent neurotoxicology review suggested that developmental effects due to P F A S s m a y be mediated b y thyroid toxicity, influence on calcium homeostasis, protein kinase C, synaptic plasticity and cellular differentiation, perhaps as part of a cocktail of substances that in combination reach harmful concentrations [252]. Gi ve n the PFC-associated hormonal disruptions that m a y occur during fetal development, neurodevelopmental toxicity is likely. T a k e n as a whole, P F C exposure at levels similar to or b e l o w those reported from the East Me t r o area are associated with a range of adverse neurobehavioral effects. Concerns regarding adverse effects on brain development we re apparently not explored during the early years, but evidence is n o w emerging. T h e developing brain is a highly sensitive target for environmental chemicals [253], and the available evidence sh o w s that the 67 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN h u m a n brain is also likely to be vulnerable to P F C toxicity, although assessment of the magnitude of PFC-associated neurotoxic risks is not yet possible. 1. Epidemiological evidence a. Neurobehavioral outcomes were assessed regarding maternal pregnancy serum concentrations of P F O A and P F O S in 1400 m e m b e r s of the Danish National Birth Cohort. Maternal s e r u m - P F O S w a s found to be associated with delayed gross motor development in the first t w o years of life [254], but n o other significant associations w e re found with crude developmental milestones during the first t w o years of life or children's subsequent behavioral or coordination problems at age 7 (mailed questionnaire) [96]. All results w e re adjusted for duration of breast-feeding, although lactational P F C exposure w a s not modeled. Despite the size of the study that should provide s o m e statistical power, the o u tc om e measures used m a y not be sufficiently sensitive and also too non-specific. b. T w o studies have considered children's serum P F C concentrations in relation to parental reports of A D H D . First, cross-sectional N H A N E S data from 1999-2000 and 2003-2004 for adolescents aged 12-15 years ( N = 571) found significantly increased adjusted odds ratios (ORs) for A D H D [255]. Ho we ve r, cross-sectional data on subjects aged 5-18 years ( N = 10,546) from the C 8 Health Project s h o w e d an O R that w a s significantly l o w for P F O A , significantly high for P F H x S , and not significant for P F O S [177]. These cross-sectional data are difficult to evaluate, as a causative exposure w o u l d have to precede the diagnosis, and because A D H D behaviors could conceivably affect P F C exposure levels in children. In addition, a m o r e recent follow-up of 321 of these children s h o w e d disagreements between maternal and teacher reports on children's behavior, thus m a k i n g it difficult to d r a w a conclusion from these data [256]. c. In a small cross-sectional study of 63 children aged 9-11 years in Oswe go , N e w York, s e r u m - P F C concentrations w e re measured and co mp a r e d to a neuropsychological measure of differential reinforcement of l o w rates of responding. Higher concentrations of P F O S and other P F C s we re associated with significantly shorter inter-stimulus response times as a measure of response inhibition or impulsivity [257], i.e., functions that are impaired in A D H D . This study had the advantage of using a standardized and specific neurobehavioral task, but w a s limited b y the cross-sectional design and the small size of the population examined. d. M o r e recently, increased maternal s e r u m - P F O S concentrations in a U.S. birth cohort w e re associated with poorer behavior regulation, metacognition, and executive function [258]. Wh i l e P F O A did not s h o w any adverse associations in the 25 6 mother-child pairs, increases in maternal s e r u m - P F O S w e re associated with poorer behavior regulation, metacognition, and global executive functioning in the child. c. O u r o w n birth cohort study in the Faroes assessed behavioral difficulties b y the parent-reported Strengths and Difficulties Questionnaire ( S D Q ) at age 7 years [259]. Wh i l e n o associations we re observed with prenatal exposures in the 539 children, overall S D Q scores w e re associated with postnatal P F O A exposure, and significant adverse associations were found regarding hyperactivity, peer relationship, and conduct problems, as well as internalizing and externalizing problems and autism screening scores. 68 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN f. Neurological abnormalities have also been linked to P F C exposures, such as cerebral palsy [260]. Although the m e c h a n i s m is u n k n o w n , it is possible that P F C s contribute to a multicausal etiology so that elevated exposure is associated with an increased risk of this outcome. T h e blood samples in this study were analyzed b y a Danish laboratory, not b y 3 M as in previous reports on this cohort. 2. Toxicological evidence a. Developmental neurotoxicity has been demonstrated in various experimental models, mainly in rodents and in vitro [4]. A f e w studies will be highlighted due to their potential significance for effects in humans. b. Neonatal exposure to P F O S and P F O A seems to affect the cholinergic system, as indicated b y a hypoactive response to nicotine, diminished habituation to a novel environment and lack of activity [261]. Similar experimental outcomes have been described for well-established h u m a n neurotoxicants, such as PC B s . c. A recent neurotoxicology review suggested that developmental effects due to P F C s m a y be mediated b y thyroid toxicity, influence on calcium homeostasis, protein kinase C, synaptic plasticity and cellular differentiation, perhaps as part of a cocktail of substances that in combination reach harmful concentrations [252]. In particular, the PFC-associated effects on thyroid function (see section D ) w o u l d appear to be highly relevant. 3. Perspective O n ly recently has scientific attention focused on developmental neurotoxicity as a highly sensitive o u tc om e that can have serious consequences, also in regard to economic costs to society [253]. Wh i l e it could have been relevant and appropriate to examine neurotoxicity from prenatal exposures, and while evidence of such effects due to lead contamination w e re well k n o w n already in the 1980s [262], this issue w a s not explored at the time, and the C 8 Panel did not m a k e this a priority, presumably because little evidence at the time suggested that the developing brain could be an important target organ in regard to P F C exposures. M y ju dg me nt therefore relies on the coverage and directionality of very recent research and the likelihood that established neurotoxic m e c h a n i s m s are triggered b y P F C exposure. G. Liver toxicity It is m y opinion, based on the weight of the epidemiological evidence, and supporting toxicity evidence, that P F C s pose a substantial present and potential hazard to h u m a n liver functions, with related adverse health effects. T h e liver is an important target organ for P F C toxicity in h u m a n s as in animals, and such effects have been referred to b y regulatory agencies in regard to determining safe exposure limits [1, 4, 148, 149]. T a k e n as a whole, P F C exposure at levels similar to or b e l o w those reported from the East Me t r o area are associated with a range of liver toxicity effects. P F C 's adverse effects on liver functions are reflected b y elevations of serumcholesterol concentrations and other important serum lipid parameters. E v e n small increases are 69 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN likely to have negative implications regarding cardiovascular disease and mortality. In this regard, the C 8 Panel concluded that P F O A is linked to an increased risk of elevated serumcholesterol, and but not (yet) hypertension and coronary artery disease [263]. A s hypercholesterolemia and cardiovascular disease are of major public health concern, these issues are discussed under a separate heading (Section VII.G). 1. Epidemiological evidence a. Increased serum-cholesterol concentrations at elevated P F C exposures likely relate to toxic effects on liver functions, and increased concentrations of liver e n z y m e s in serum at higher P F O A exposures support this notion in about a dozen epidemiological studies. A n early mention of adverse liver effects from P F C s in workers is from 1978, w h e n D u P o n t medical officers found elevation in liver function tests a m o n g a group of workers leading to the conclusion that "it is possible that C-8 m a y be causing very minimal, and certainly not clinically apparent, toxic effects to the liver. Because the total n u m b e r of records reviewed is small (31), I do not believe any findings of this study are statistically valid."pppp It appears that these results we re discussed with the 3 M medical colleagues. External consultant, Professor Harold H o d g e in 1978 reported to 3 M 's medical director, Dr. F.A.Ubel: "There appears to be indications of liver change from the physical examination results. In terms of indicators of liver disorder, there are [sic] a higher percentage at Chemolite than at Decatur and the organically b o u n d fluorine level at Chemolite is correspondingly higher."qqqq b. In early 1980, D u P o n t shared the results of a pilot study called "liver e n z y m e study of workers exposed to C-8 at Parkersburg," wh e r e they found elevated m e a n serum concentrations of aspartate aminotransferase (AST, previously referred to as S G O T ) and alkaline phosphatase (AP) a m o n g workers operating the T F E process. rrrrD u P o n t provided m e m o r a n d a summarizing their findings. O n June 9, 1980, the assistant medical director V a n n A. Brewster wrote to L. F. Percival that "I a m concerned that the `Draft' implies that the Medical Division will not continue the study of liver tests on those employees potentially exposed to C-8. E v e n though w e have found n o `conclusive evidence of an occupationally related health problem,' w e still cannot explain w h y the m e a n S G O T [same as A S T ] w a s significantly higher a m o n g T F E process workers and that the m e a n A P w a s significantly higher a m o n g F E P process and service workers."ssss I have been unable to identify evidence that liver tests were of continued. pppp AR226226-1453.pdf, September 20, 1978 - DuPont Washington Work's Medical Director, Dr. Younger pared a memo summarizing his review of the medical records of eleven operators and eighteen laboratorians [at the Washington Works Plant] who have had long-term exposure to C-8. (Exhibit L (EID080236-40)). Page 000135. qqqq3MA00967742. (1978.08.23). Minutes of Meeting with H.C. Hodge, p. 3MA00967744. rrrr AR226226-1465.pdf, January 28 1980, Liver Enzyme Study of Workers exposed to C-8 at Parkersburg, Exhibit CC (EID099433-34). Page 000186. ssss AR226-1469. June 9, 1980 - DuPont prepared a memo expressing his concern that a draft communication to DuPont's Washington Works employees regarding the outcome of DuPont's liver study of employees "implies that the Medical Division will not continue the study of liver tests on those employees potentially exposed to C-8." (Exhibit GG (EID102477)). Page 000192. 70 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN c. In his thesis, Dr. Gilliland found increases in serum concentrations of S G O T and S G P T ( n o w referred to as A S T and A L T ) , as well as a tendency toward lower H D L cholesterol, as markers of adverse effects on liver functionTM Other cross-sectional studies [191, 248] w e re not informative in this regard, perhaps because a variety of other factors can impact on liver functions. T h e studies w e re reviewed in greater detail in regard to i m m u n e function parameters in Section VII.A. d. T h e fact that the liver w a s an important target organ recognized b y 3 M 's Dr. Richard Purdy, and in 1999 he wrote to 3 M colleagues, Drs. John Butenhoff and A n d r e w Seacat that his calculations s h o w e d that a "general population m e m b e r with 70 ppb (in o n e 's blood) could have 36 times m o r e in his liver"uuuu due to life-time accumulation. Ho we ve r, I have been unable to determine h o w this conclusion w a s dealt with at 3 M . c. Further analyses of medical surveillance data on P F O A - e x p o s e d workers in Minnesota led to a 3 M paper that relied on cross-sectional analyses of P F O A and liver function data collected in 1993, 1995 and 1997 [66]. Wh i l e the wording differs from a previous report [207], the authors concluded that employees' serum P F O A levels we re not positively associated with either clinical hepatic toxicity nor hepatic responses to obesity and alcohol.vvvv f. Drs. Olsen and M a n d e l reported in 1998 results on P F O S - e x p o s e d A n t w e r p and Decatur male fluorochemicals production workers, in wh i c h they concluded that hematological, clinical chemistry and hormonal abnormalities we re not associated with serum P F O S levels u p to 6 p p m (6,000 ng/mL).wwww Although deviations occurred at higher exposures, the authors disregarded these findings, basing this decision on a determination there we re too f e w subjects to allow a firm conclusion [264]. A later analysis of medical surveillance data from a fairly small n u m b e r of employees again h o w e d a positive association between the s e r u m - P F O A concentration and both cholesterol and triglycerides. These findings we re considered implausible, as they are not in accordance with animal data at m u c h higher exposures.*xxxx Another limitation that the 3 M authors emphasized w a s the possible non-adherence b y s o m e workers to the fasting requirement, although blood-glucose w a s not affected.yyyy tttt AR226-0473. Frank Davis Gilliland, Fluorocarbons and Human health: Studies in an Occupational Cohort (October 1992) (unpublished Ph.D. thesis, University of Minnesota), with Summary. Page 003247. uuuu 3MA01403075. Thoughts on human safety factors. Page 3MA01403075. vvvv AR226-0477. Geary W. Olsen, et al., An Epidemiologic Investigation of Plasma Cholecystokinin and Hepatic Function in Perfluorooctanoic Acid Production Workers, 3M Final Report EPI-0003 (1997), with Summary of study, Protocol, and Manuscript accepted for publication in 2000, Drug & Chemical Toxicology. Page 003511. wwwwAR226-0030. An Epidemiologic Investigation of Clinical Chemistries, Hematology and Hormones in Relation to Serum Levels of Perfluorooctane Sulfonate in Male Fluorochemical Production Employees (List of Section Attachments is first page of this File). Page 001074. xxxx 3M_MN02334964. Final report, A Longitudinal Analysis of Serum Perfluorooctanesulfonate (PFOS) and Perfluorooctanoate (PFOA) Levels in Relation to Lipid and Hepatic Clinical Chemistry Test Results from Male Employee Participants of the 1994/95, 1997 and 2000 Fluorochemical Medical Surveillance Program. Page 3M_MN02334966. yyyy 3MA01784788. An AnalYsis of the 2000 Pluorochernical (Perfluorooctanoate, PFOA) Medical Surveillance Program at 3M Company's Antwerp (Belgium), Cottage Grove (Minnesota), and Decatur (Alabama) Facilities. Page 3MA01784817. 71 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN g. Although these findings w o u l d justify, at a m i n i m u m , further follow-up, it is not clear if that happened and if the findings w e re ever analyzed and published. T h e C 8 Health Project ex am in ed 47,092 adults for effects of P F O A and P F O S on alanine transaminase (ALT), gamma-glutamyltransferase, and bilirubin as markers of liver function. These results s h o w e d a positive association between serum P F O A and P F O S concentrations and the serum A L T concentration [265], usually interpreted as a sign of hepatocellular damage. 1. W h e n s e r u m - P F O A concentrations we re mo d e l e d as cumulated concentrations, the adverse effect on serum A L T concentrations w a s replicated in a m i x e d O h io River Valley population [266]. j. Likewise, in a general population sample from the N H A N E S study, liver e n z y m e s s h o w e d significant, though small, increases at higher s e r u m - P F O A concentrations [267]. k. Several occupational studies, both cross-sectional and prospective, have assessed liver function parameters in serum, the m o s t recent ones [66, 268, 269] showing that, in general, liver e n z y m e s tend to increase, while bilirubin decreases at higher P F C exposure levels. 2. Toxicological evidence a. Although liver dysfunction in exposed workers w a s discovered fairly early, it w a s not taken seriously for m a n y years. A study conducted in 1976 b y Dr. Taves from the University of Rochester reported P F O A stimulated lipid peroxidation (LP) in an in vitro experiment. T h e author noted that, at the time, in vivo effects of P F O A were u n k n o w n . zzzz b. T h e liver w a s early identified as a m a i n target organ in rodents [44]. Although toxic m e c h a n i s m s m a y differ between rodents and h u m a n s [7, 17], as I discussed above, the PPAR-related m e c h a n i s m is n o longer believed to be the differentiator 3 M once m a d e it out to be [10]. c. Detailed discussion of liver toxicity in experimental models is included in recent evaluations b y regulatory agencies [4, 148, 149], to wh i c h little recent evidence adds only little. d. O n e aspect deserves consideration, i.e., the intrahepatic lipid metabolism. S o m e P F A S s have the potential to induce hepatic lipid accumulation in cyno mo lg us m o n k e y [270] and induce lipid synthesis gene expression in h u m a n hepatocytes [271]. c. In mice, P F O S administration induced hepatic steatosis in time-and dosedependent m a n n e r along with corresponding C D 3 6 and Lpl expression induction and decreased mitochondrial P-oxidation in mice [272]. Also, in exposed animals, accumulation of lipid zzzz 3MA02512169. Comparison of Lipid Peroxidation by Perfluoro-Octanoic Acid or CCL4. Page 3MA02512169. 72 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN droplets in hepatocytes w a s observed. These findings suggest that steatosis and fatty liver degeneration m a y be relevant outcomes of elevated P F C exposure. f. In regard to P F B A , the M D H [63] noted liver weight changes and morphological changes in liver in experimental studies carried out b y a private laboratory in Japan [273]. T h e article provides n o information on funding. 3. Perspective E v e n though the liver and lipid metabolism w e re identified early on as likely targets of P F C exposure, it appears that the understanding of the impact on workers' health and on exposed communities in general developed very slowly and that great hesitation w a s repeatedly voiced in regard to accepting a hypothesis of P F C hepatotoxicity. a. During a meeting on M a r c h 16, 1983, a handout prepared by 3 M 's D o n R o a c h mentioned preliminary data from a study of "abnormal" liver e n z y m e s from film vs. chemical plant workers. N o details w e re given in the document.aaaaa Ho we ve r, a report previously prepared b y D.E. R o a c h reported results from certain health evaluations beginning in late 1976. In this paper, he refers to serum gamma-glutamyl-transferase ( S - G G T ) being elevated, but he considered it, without a fully supported explanation, as "traceable to alcohol consumption." During the three-year monitoring program, there w a s a decline in the n u m b e r of individuals with "abnormal" laboratory findings, but Dr. R o a c h attributed that, again without clearly supported explanation, to health care counseling and improved lifestyles.bbbbb b. T h e first systematic studies w e re carried out b y Dr. Frank Gilliland in connection with his thesis project. ccccc c. Additional health studies appear to have been carried out at 3 M subsidiaries. In 1995, 3 M issued an internal report based on a 3 M Be l g i u m blood testing program in Antwerp. T h e report stated, "There w a s w e a k relationships found between P F O A and triglycerides (corr.coef=.48, p=0.0005)." T h e authors claimed, "the relationship has never appeared in the medical or toxicology literature and is likely to be a chance finding."ddddd Considering the high correlation coefficient and the extremely l o w p value, that conclusion seems inappropriate and defensive. d. Dr. M a n d e l later asked Dr. Schmickler from 3 M B e l g i u m about the reporting requirements for fluorocarbons, i.e., whether they need to report findings from toxicology studies. T h e reply w a s that it w a s required to report to the safety committee of the aaaaa3MA00047974. Handouts and Transparencies for your Files from March 16, 1983 Fluorochemical Study Committee Meeting. Page 3MA00047975. bbbbb3MA00047519. Health Status of Plant Workers Exposed to Fluorochemicals - a Preliminary Report. Page 3MA00047524. cccccAR226-0473. Frank Davis Gilliland, Fluorocarbons and Human health: Studies in an Occupational Cohort (October 1992) (unpublished Ph.D. thesis, University of Minnesota), with Summary. dddddAR226-0175. Antwerp Blood Testing Results from June 1995. Page 004265. 73 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN company but not governm ental institutions.66666 I did not find any such reports, and perhaps this judgm ent explains the apparent secrecy about any findings o f adverse effects. c. As late as 2003, 3M authors argued that a positive association betw een PFO A exposure and cholesterol in 3M w orkers "is contrary to the substantial body of toxicological literature that suggests a negative association in laboratory anim als" [217]. However, in a m ore recent article [274], the 3M authors relied on a species difference in liver metabolism (associated with the PPA R receptor) and for this reason concluded that hepatocellular tumors in rats are "not likely to be relevant to hum ans." However, these positions are inconsistent. It is not appropriate in one connection to require sim ilar hepatotoxic effects in different species and in another to raise doubt about such similarity. f. In regard to liver steatosis, up to 10% o f adolescents have non-alcoholic fatty liver disease (NAFLD) [275, 276]. As a considerable and apparently grow ing public health problem o f partially unknow n origin, this outcom e requires attention in future studies o f PFCassociated adverse hum an health effects. H. Risk factors for cardiovascular disease It is my opinion, based on the w eight o f the epidem iological evidence, and supporting toxicity evidence, that PFCs pose a substantial present and potential hazard to human health due to elevated cholesterol and increased risk o f cardiovascular disease. Based on the available evidence, the kidney may also be a likely target organ for PFC toxicity in hum ans as in animals. How ever, this evidence is yet uncertain, as decreased kidney function of other causation may im pair the elimination of PFCs and thereby indirectly cause elevated serum-PFC concentrations. The evidence is nevertheless strong that PFCs cause adverse cardiovascular effects, in part associated w ith elevated cholesterol and w hether or not kidney disease is a contributing factor. An autopsy study showed that PFBA in particular lodges in the hum an kidney [52], but virtually no inform ation is available on nephrotoxicity and cardiovascular toxicity related to this PFC. As discussed above, serum concentrations o f total cholesterol and other important serum lipid param eters increase at higher PFC exposures. Even a small increase w ould likely have negative im plications regarding cardiovascular disease and possibly mortality. The C 8 Panel concluded that PFO A is linked to an increased risk o f hypertension in pregnancy, elevated serum-cholesterol, and potentially also coronary artery disease, although the latter was not considered sufficiently supported by the evidence available at the time. eeeee 3MA00000688. Email between Jeffrey H. Mandel and Dokter Schmickler with the subject "reporting." Page 3MA00000688. 74 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN Due to the high incidence o f cardiovascular disease, even a small increase in life tim e risk is o f serious public health importance. An article recom m ends im m ediate action to prevent even `b a c k g ro u n d ' exposures to P F O A [277]. 1. Epidemiological evidence a. The early 3M occupational study by Dr. G illiland addressed serum chemistry abnorm alities in exposed w orkers [207] and showed an inverse correlation (adverse effect) betw een organic fluorine compounds (assum ed to be m ainly PFO A) and HDL cholesterol. b. In other 3M production plant w orkers, serum PFO A and total organic fluorine (TOF) were positively and significantly associated with cholesterol and triglycerides in a longitudinal study. fffff A positive correlation (adverse effect) w ith total cholesterol (PFO S and PFO A) and triglycerides (PFOA) was later found in regard to PFC concentrations in serum in one study, though not in another [217, 278]. c. A later study o f current and form er Cottage G rove em ployees showed both PFO A and PFOS were positively and significantly associated w ith total cholesterol, LDL, and triglyceride levels above healthy reference ranges.ggggg PFO S w as significantly associated w ith m etabolic syndrome, as well. Oddly, 3M discontinued this study showing significant adverse effects in its w orkers.hhhhh d. Similar evidence from cross-sectional and, in particular, prospective studies o f workers at other plants also suggested that increased PFO A exposure is associated w ith higher serum -cholesterol concentrations [268, 269, 279]. Cross-sectional data on 1216 subjects from the 1999-2003 NH AN ES showed that increasing serum-PFOA concentrations were positively associated with self-reported cardiovascular disease, including coronary heart disease and stroke, and objectively m easured peripheral arterial disease (an ankle-brachial blood pressure index o f less than 0.9). The highest PFO A quartile showed a doubling o f cardiovascular disease after confounder adjustm ent [280]. f. C om m unity and general population groups w ith low er levels o f P F O A exposure have also revealed positive correlations (adverse effects) betw een PFO A and cholesterol concentrations in serum [65, 281-283]. g. In som e populations, other P F C s w ere also m easured, and positive associations were found in regard to PFOS exposure [281-283], a finding that we have replicated fffff 3M_MN02334964. Final report, A Longitudinal Analysis of Serum Perfluorooctanesulfonate (PFOS) and Perfluorooctanoate (PFOA) Levels in Relation to Lipid and Hepatic Clinical Chemistry Test Results from Male Employee Participants of the 1994/95, 1997 and 2000 Fluorochemical Medical Surveillance Program, page 3M_MN02334966. ggggg 3MA02543911. Untitled draft. See also Deposition Testimony of Dr. Geary W. Olsen (Sept. 8 , 2017), 142:13 150:21. hhhhhDeposition Testimony of Dr. Geary W. Olsen (Sept. 8 , 2017), 158:24-160:06. 75 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN in elderly subjects from the Faroe Islands, where exposure levels are sim ilar to background levels in the US (unpublished results). h. D ata from the C 8 project show ed that total cholesterol and LD Lcholesterol increased at higher PFO A exposures (and HD L increases in w om en) [235]. The variability in P F C -associated cholesterol changes is quite large [8 ], although a variety o f factors, such as age, sex, and body mass index could affect the degrees o f the relationship [284]. i. Indirect evidence suggests th at PFC m etabolism is n ot linked to changes in lipid m etabolism (which w ould suggest a reverse causation), thereby rejecting a hypothesis that both PFCs and cholesterol could be affected by a comm on cause that would produce apparent positive associations betw een PFCs and cholesterols in serum. Thus, subjects w ho are taking statins to decrease their serum -cholesterol do not show any low er serum -PFC concentrations [282]. This report agrees w ith our findings in the Faroes (unpublished). j. W hile early data from C ottage G rove suggested no risk, an increased risk o f cerebrovascular disease was indicated by a m ortality study that relied on com parisons w ith the general population [285]. The subsequent 3M -supported follow-up [76] again showed strongly elevated risk of cerebrovascular death in workers with high exposure, especially when compared to an internal control group. The draft report by Drs Lundin and A lexander from the University o f M in nesotaiiiii provides a b alanced presentation, b u t the published article th at w as co-authored by 3M 's Dr. Olsen calls the association "inconsistent" because the m ortality was not clearly elevated when compared to the general M innesota population. k. C ross-sectional N H A N E S data suggest that serum -PFO A concentration is associated w ith systolic blood pressure and the risk o f hypertension [286]. H ypertension may relate to an increased risk o f cerebrovascular mortality. l. N H A N E S data suggest th at increased serum concentrations o f P F O A and PFOS are associated with an increased risk of chronic kidney disease, as defined by a low g lom erular filtration rate [287].JJJJJ H ere, reverse causation cannot be ruled out, i.e., th at kidney disease prevents PFC excretion via the urine[59]. A m ajor caveat, however, is that PFB A was not considered, as it may not have been detectable in the serum, while the m ajor accum ulation in the body is in the kidneys [52]. m. Regarding uric acid, the C 8 Project exam ined its association w ith serum PFO A levels after adjustm ent for potential confounders. An increased risk o f elevated uric acid was found in adults, including clinically defined hyperuricem ia [288]. Again, this evidence is yet somewhat uncertain, as reverse causation may be present. 11111 3MA02557439.pdf. Final report, Mortality of Employees of an Ammonium Perfluorooctanoate Production Facility, August 22, 2007. jjjjj I note that the first author of this article, and co-author of five other publications on cardiovascular outcomes in PFOS-exposed populations, has provided erroneous information to the West Virginia University regarding his educational background. The articles in question were co-authored be established colleagues, and none of them has been retracted. 76 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 2. Toxicological evidence a. Detailed discussion of elevated cholesterol and other risk factors in experimental models is included in recent evaluations b y regulatory agencies [4, 148, 149]. b. In regard to P F B A , experimental findings in laboratory models included exposure-related decreased in red blood cell counts, hematocrit and hemoglobin, effects that we re highlighted b y the M D H in regard to setting a safe limit for P F B A concentrations in water [63]. Further information is given in Section VIII. 3. Perspective a. D u e to the impact of covariates and the fact that cholesterol concentrations in the C 8 study w e re lower than elsewhere in the U.S., selection bias in this study has been suggested [284]. Ho we ve r, although s o m e selection forces likely played a role, there is little evidence to suggest that it caused bias a w a y from the null hypothesis regarding P F O A exposure and cholesterol. E v e n though PFC-associated increases in cholesterol in mo st studies s e e m to be fairly small, the overall impact at population level could be substantial. T h e strongest associations in general refer to total cholesterol, but s o m e studies have also examined lipoprotein fractions and found that especially low-density (LDL-cholesterol) increases at higher s e r u m - P F C concentrations. b. E v e n ifP F C exposure explains only a small part of the variation in serum- cholesterol concentrations, still small increases in total and L D L cholesterol are associated with increased risks of cardiovascular disease. S o m e have noted that an increased mortality attributed to this cause has not been d o cu me nt ed so far [8]. Thus, previous studies have suggested that cardiovascular mortality in P F C workers is b e l o w expectation. H o we ve r, this could arise from a healthy worker effect. S o m e evidence of increased risk 10 years after first e m p l o y m e n t w a s noted [74], as w o u l d be expected w h e n the healthy-worker effects wears out. c. Other serum parameters that m a y reflect kidney dysfunction, such as creatinine and blood urea nitrogen ( B U N ) , w e re assessed in occupational studies. Ho we ve r, no clear associations with P F C exposure biomarkers w e re found [279]. I. Carcinogenicity It is m y opinion, based on the weight of the epidemiological evidence, and supporting toxicity evidence, that P F C s pose a substantial present and potential hazard to h u m a n health as carcinogens. A s detailed evaluations have been recently published[6], I shall briefly summarize the early reports along with the m o st recent evaluations. Early studies in PF C - e x p o s e d workers suggested a risk of prostate cancer, and support for this association c o m e s from m o r e recent studies that also include populations exposed at background levels. Wh i l e early studies usually referred to cancer mortality, wh i c h is appropriate for sites like liver and pancreas due to the high fatality of these diseases, other sites 77 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN are better explored using incidence data, while considering the impact of screening efforts, e.g., for prostate cancer. A variety of subsequent studies demonstrate that P F C exposure is associated with development of cancer at several sites. P F O A has been found to satisfy the E P A 's criteria to be classified as "likely to be carcinogenic to h u m a n s " [289]. T h e I A R C concluded last year that P F O S is a possible (Group 2B ) h u m a n carcinogen [6]. T h e C 8 Science Panel concluded that there is a probable link from P F O A exposure to testicular cancer and kidney cancer. All of these effects have been reported at background levels or at elevated exposures overlapping with those docu me nt ed in East Me t r o residents. T h e m a i n organs affected are the kidneys, testicles, prostate, and perhaps bladder and breast. Ba s e d on the available evidence, exposure to P F C s has a substantial potential to cause cancer, mo st clearly in regard to cancers of the kidneys and the testicles, and highly likely also in regard to prostate cancer and bladder cancer. A possible risk of breast cancer is also of concern. 1. Epidemiological evidence Cancer risk assessments a. In the most recent evaluation of cancer risk association with P F O A exposure, the I A R C classified this substance as a possible h u m a n carcinogen (Group 2B ) and concluded that there w a s relevant, though limited, evidence in h u m a n s that P F O A causes testicular and kidney cancer, while evidence for h u m a n carcinogenicity at other sites, such as bladder, prostate, thyroid, liver and pancreas w a s inadequate at the time of the evaluation; relevant evidence in experimental animals w a s also considered limited [6]. b. T h e E P A 's Science Advisory Boar d in 2 0 06 reviewed the information available on P F O A at the time and suggested that the cancer data were consistent with the E P A Guidelines for Carcinogen Risk Assessment descriptor "likely to be carcinogenic to h u m a n s " [289]. T h e evidence available for P F O S and/or other P F C s m a y also be sufficient for this classification, but n o conclusion w a s reached. c. A s I describe in m o r e detail below, the C 8 Science concluded that there is a probable link from P F O A to testicular cancer and kidney cancer. T h e C 8 Science Panel conducted further studies, including an update of a previous occupational study [73]. Data on prostate cancer did not provide further support for the associations already referred to, although caveats described b e l o w m u s t be kept in mind. d. I a m aware of Professor Da v i d Sunding's statistical analyses in connection with this case showing a pattern of increased occurrences of total cancer as well as cancer at specific sites in areas in Minnesota that I understand have been m o s t affected b y P F C contaminated ground water. His findings are consistent with the evidence su mmarized be l o w showing correlations between P F C exposure and cancer rates at individual sites. O ccupational studies 78 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN c. As described below , a dose-associated increased risk o f kidney cancer was observed in workers in a fluoropolym er production plant in W est Virginia, USA, and in the local community exposed to releases from the plant. Likewise, an increased risk of testicular cancer occurred in highly exposed local residents. Relevant evidence in hum ans also referred to other cancers [6 ]. f. O ccupational m ortality studies have b een carried out in P F O S -exposed w orker populations from the 3M production plant in A labam a [285, 290] and PFO A-exposed w orkers from W est V irginia [291] and M innesota [93]. In evaluating this evidence, account m ust be taken o f duration o f exposure, exposure assessm ent m ethods, age at entry, and duration of follow -up, as discussed by IA R C [6 ]. In addition, follow -up and case-control studies have focused on exposed communities. Some reports are not considered here, as small num bers of cases or other weaknesses m ake them less relevant to the evaluation. W hile the recent evaluation report relied on published evidence, some internal studies have been conducted in the past and provide some supplem entary evidence. A key concern in these studies is the choice of com parison population, cf. the com m ents m ade above in regard to the "healthy w orker effect" (see section V.A.1). g. In A pril 1989, D u P o n t issued an Internal R eport, "A n investigation into the occurrence o f leukem ia at W ashington W orks." The standardized m ortality ratio was 2.1, but w as determ ined by D u P o n t n ot to be statistically significant.kkkkk L eu k em ia has n ot been considered in recent reports that m ainly relied on m ortality data, that are not a reliable source for hematopoietic cancer incidence. h. In a subsequent report from 1992, D uP ont exam ined the cancer surveillance data for 1956-1989 and m ortality data for 1957-1991. For cancer, DuPont found a null result overall, but significant findings existed for specific sites, such as buccal cavity, pharynx, kidney and leukemia, among m ale employees (too few fem ale employees). For mortality, m ostly seen were deficit deaths (healthy w orker effect) among m ales; among females, there was a significant excess o f residual causes o f deathsTM Insufficient inform ation is available to judge this report, but it illustrates that attention was paid to cancer risks early on. i. In response to the concerns about P F C -associated cancer risks, the M innesota Cancer Surveillance System on cancer incidence includes D akota and W ashington counties and related zip codes, and data from the period 1988-2004 have been published [292]. A recent update does not appear to be available. The results relate to the patient residence at the tim e of diagnosis, and this report therefore has lim ited power to detect excess cancer rates associated w ith past PFC exposure from contam inated drinking w ater in parts o f these counties. Unsurprisingly, none was found. In interpreting such studies, one m ust also note that the period kkkkkAR226226-1308-1.pdf, DuPont Internal Report, "An Investigation Into The Occurrence of Leukemia At Washington Works" (April 1989) (EID584220-30) DuPont Internal Final Report, "A Case-Control Study of Leukemia At the Washington Works Site" (12/3/91) (EID151953-65). Page EID584221. llmAR226-1546. Washington Works Cancer Surveillance Data Mortality and Cancer Incidence. Pages EID521396 and EID521399. 79 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN from first exposure to the developm ent o f clinical cancer m ay be longer than 2 0 years, while the residence is likely to change at least once during this period. Kidney cancer j. R egarding kidney cancer, the C 8 Panel concluded: "F o r kidney cancer, the w orker m ortality study conducted by the Science Panel showed a higher risk in the m ost highly exposed group compared to lower exposure groups among the workforce, but the risks were not elevated com pared to the US population. In the cohort study, there was a gradient o f increasing risk with increasing exposure but m ost strongly in the analyses that included exposure up to the tim e o f diagnosis. W hen the 10 years o f exposure prior to diagnosis w as excluded, the association was less evident. N o association w as seen in the prospective analysis of cohort data, although the latter is lim ited by small numbers. In the geographic study, some results suggested an increasing risk of kidney cancer w ith increasing exposure and others did not. The science panel considers that the excesses observed indicate a probable link betw een PFO A and kidney cancer." k. The C 8 Panel review ed and relied on several studies, as did the IA RC w orking group. Increased risk of kidney cancer with a statistically significant exposure-response trend was reported in workers in a fluoropolym er production plant in W est Virginia and in an exposed comm unity near the plant [291, 293]. l. In furth er detail, elevated m ortality from m alig n an t kidney disease w as docum ented am ong 5,791 w orkers exposed to PFO A in W est V irginia [291]. N o clear risk was seen in the PFO S-exposed w orkers from M innesota [93]. H ow ever, com m unity-based evidence [293, 294] showed an elevated incidence o f kidney cancer associated w ith PFO A exposure. Testicular cancer m. The C 8 report concludes: "For testicular cancer, there is evidence o f a positive trend in risk across exposure groups, in some analyses, w ith the highest exposure group in both the internal analyses of the cohort study and the geographical cancer study showing estimated relative risks ranging from 3 to over 6 comparing the highest to lowest exposure groups. [...] The high exposure group, w here the higher risk w as observed, com prises only six cases therefore there rem ains some uncertainty. The Science Panel notes that there is experimental evidence o f testis cancer being increased in exposed animals. The Science Panel considers observed excesses to indicate a probable link betw een PFO A and testicular cancer."mmmmm The conclusions from IA R C [6 ] are sim ilar in regard to testicu lar cancer. n. M ortality studies are unlikely to identify all cases o f testicular cancer, and better evidence m ust rely on incidence data. The C 8 Panel and IARC em phasized the results from the comm unity study that docum ented an elevated incidence of testicular cancer at higher PFO A exposures in the M id-O hio R iver Valley near the production plant [293, 294]. mmmmm3 M'SDr. Butenhoff agrees and takes no issue with the C8 Science Panel's conclusion of a statistical link between PFCs and testicular cancer in humans. Butenhoff Dep. Tr. at 139. 80 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN Prostate cancer Bo th prostate cancer and bladder cancer are diagnoses that m a y not necessarily appear on a death certificate, as the patient m a y die from s o m e other disease, rather than the cancer, because the cancer is often curable or m a y not be fatal for several years. In addition, cancer risks in these types of studies often are calculated on the basis of small nu m b e r s of cases and m a y therefore not deviate with statistical significance from expectation. Thus, although the risk m a y not be significantly elevated, the upper confidence limit could be 5 or higher, suggesting that, at the s a m e time, a 5-fold increased risk or greater cannot be ruled out. However, based on published evidence, neither the C 8 Panel nor the I A R C considered prostate cancer a probable risk in regard to P F C exposure. o. In a follow-up to 3 M 's 1989 mortality study carried out by Dr. Mandel, involving almost 3,000 male 3 M workers at Cottage G r o v e from 1947 to 1983, ten years of e m p l o y m e n t in exposed jobs w a s associated with a statistically significant increase in prostate cancer mortality (more than three-fold). Still, this calculation w a s based on four cases a m o n g the exposed workers [72]. Unfortunately, comparisons with the general population of Minnesota probably biased the results of this study toward underestimated risks. p . In 1993, Dr. Gilliland and Dr. M a n d e l of 3 M published a paper based on the n e w mortality study included in Gilliland's thesis.nnnnn There were mostly null findings, except, as before, for prostate cancer (a 3.3-fold increase in mortality). Again, the n u m b e r of cases w a s small (n = 6). T h e paper sought to explain the observed prostate cancer deaths as due to a purported higher prevalence of prostate cancer in Minnesota than the U.S. (control group) and/or as a chance finding.oooo Ho we ve r, subsequent analyses provided to 3 M b y the author indicated that prostate cancer w a s actually less prevalent in Minnesota than in the U.S., suggesting an even greater excess a m o n g 3 M workers.ppppp q. In a mortality study of almost 4,000 employees exposed to P F O A , n o clear tendencies w e re found for liver, pancreatic or testicular cancer. A n increased standardized mortality ratio, however, w a s found for prostate cancer, and a 6.6-fold increased risk w a s found in workers with definite exposure. I note that the statistical significance relied u p o n t w o cases a m o n g the workers with k n o w n high exposure [76]. r. T h e m o s t recent update is from a thesis completed in 2013.qqqqq Using air monitoring results (and ignoring non-respiratory intakes), this study of 9,000 workers hired after 1947 co mp a r e d deaths at Cottage G r o v e with those at the unexposed St. Paul plant through to 2002. W h e n dividing the workers into six different exposure groups, a dose-dependent risk appeared for prostate cancer, although not statistically significant. T h e author concluded that the results supported previous findings of a prostate cancer risk. Ho we ve r, in the published report mmnnAR226226-0472.pdf, Frank S. Gilliland & Jack S. Mandel, Mortality Among Employees of a Perfluorooctanoic Acid Production plant, 35 JOM 950-954 (September 1993), with Summary of study. Page 003166. 00000AR226-0471. Jack S. Mandel & Leonard M. Schuman, "Mortality Study at the 3M Chemolite Plant" (January 1989), with Summary of study. Pages 003148-003149. ppppp 3MA00632313. Letter from Jack S. Mandel to Larry Zobel (Apr. 6 , 1989); 3MA00632314. Table 5 (attachment to Mandel letter to Zobel) (Apr. 6 , 1989). qqqqq3 M MN03059185. Cancer mortality in 3M chemical workers (PhD thesis by Katherine Koehler Raleigh). 81 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN [93] that w as co-authored by 3M 's Dr. Olsen, emphasis w as on com parisons w ith the general population, no clear trend was found in quartile exposure groups, and the lack of association of prostate cancer w ith the exposure estimate was said to be in agreem ent w ith the findings in other studies. In discussing the possible risk factors, the D iscussion section o f the published article notes that fam ily history o f prostate cancer m ay play a role (see section t. below ). The difference from the thesis becomes clear when considering the preference expressed by Drs. Olsen and Zobel in a 14-page letter to Dr. A lexander, w ho w as supervisor o f Dr. R aleigh's thesis project. In the letter,rrrrr w h ich related to an earlier C ottage G rove m ortality study, the 3M doctors advocated using a control group from the general population rather than conducting the w ithin-cohort analysis which revealed a statistically significant excess risk of prostate cancer m ortality among highly exposed employees. s. In support o f prostate cancer as a potential outcom e o f PFC exposures, a nested case-control study o f cancer incidence (based on diagnosis reporting, rather than m ortality) in a D anish general population group focused on 713, 332, 128, and 67 cases of prostate, bladder, pancreatic, and liver cancers found during a follow-up of approximately 1 0 years after a baseline examination w ith blood sampling. A t the background exposure w ith a small variance, m odest positive associations were found betw een serum concentrations o f both PFO A and PFO S in regard only to prostate cancer m orbidity [97]. t. A recent case-control study from Sw eden show ed sim ilar serum -P FC concentrations in 201 cases and 186 population-based controls. Heredity, i.e., a first-degree relative w ith the disease, was a risk factor, as has been docum ented before, and am ong those with a positive family history, elevated serum concentrations of both PFO A and PFOS were associated w ith a significantly increased risk o f prostate cancer [94]. A ccordingly, PFC exposure may contribute to the etiology o f this cancer type, although this m ay not be evident, unless fam ily history is taken into account. Bladder cancer The current evidence is less strong as to bladder cancer, in part because m ortality studies are unlikely to reflect this diagnosis, in part because incidence studies carried out relied on self-reports o f past diagnoses. In addition, some studies included too few cases to support statistical analyses, and the tw o evaluations by the C 8 Panel and IARC did not consider the evidence sufficient to draw a conclusion. u. A study o f P F O S -exposed w orkers show ed th at b lad d er cancer m ortality was elevated among individuals with at least one year of exposure. I note that this finding was based on three deaths only, all o f w hich occurred in w orkers deem ed to have been highly exposed [285]. In addition, the results were obtained in com parison w ith the Alabam a general population, and the low m ortality ratio for lung cancer did not suggest that smoking w as an im portant confounder. \. In a subsequent reevaluation of the same cohort, mail questionnaires were used to include incident cases of bladder cancer. The incidence was not found to differ much rrrrr 3MA02557490.pdf. Letter of July 14, 2006 from Drs. Olsen and Zobel to Dr. Alexander. 82 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN from expectation, although an increased risk among the m ost highly exposed workers could not be ruled out [290]. In an unpublished version o f this study, Dr. Alexander, the principal investigator, rem arked that the study's statistical pow er to detect excess risks was "inherently low and prohibited a convincing exposure response analysis."sssss T he m ost recen t follow -up by these authors used a job-exposure m atrix to com plem ent exposure estimates for comparison with cancer registry data; the findings do not support elevated bladder cancer risks otherwise observed in com parison w ith state averages, although the study cannot rule out the possibility o f a risk [93]. A lthough serum -PFAS m easurem ents w ere available for m any employees, these data w ere not included. In addition, as smoking is a risk factor for bladder cancer, a com parison betw een rates for lung cancer and bladder cancer would have elucidated w hether smoking-related cancer risks were similar in the occupational groups and in the com parison population. For example, the plausibility of a bladder cancer risk in fluoride-exposed workers [295]is supported by the observation of a greater increase in bladder cancer risk than in lung cancer risk am ong the workers, although the latter risk is m uch greater in smokers. The pattern was the same in the PFO A-exposed workers, w ith a greater excess in bladder cancer than in lung cancer. Considering uncertainties in mortality data for bladder cancer and the wide confidence intervals asssociated w ith small num bers of cases, these studies showed non significant associations, but could not rule out effects of a m agnitude that w ould be of substantial public health concern. Other sites w. Thyroid cancer seem ed elevated in one analysis, and the same w as true for pancreatic cancer [296].ttttt R isk o f liver cancer w as apparently not elevated in any study [6 ]. In rodents, PFO A acts as a PP A R a agonist, w hich is linked to the developm ent o f liver tumors, pancreas acinar cell adenomas, and Leydig cell tumors. Thus, in regard to tum or site, animal studies are not predictive o f the m ost relevant sites in humans. x. A part from the focused studies referred to above, a recent review by the Institute o f M edicine suggested that PFO A exposure m ay lead to breast cancer [297]. Likewise, the E PA 's Science Advisory Board called attention to this potential, given the elevated occurrence of fibroadenom as and adenocarcinom as of the breast in tw o feeding studies in rats [298]. A lthough breast cancer may also be plausible from the evidence o f endocrine disruption, only lim ited epidem iological support is at hand. y. F o r exam ple, a study o f 31 breast cancer cases in G reenland found elevated current serum PFC concentrations as com pared to controls [299]. A n extended study of 77 cases and 84 controls [300] again showed higher serum -PFC concentrations in cases sssss 3MA00755705. Bruce Alexander, Bladder Cancer in Perfluorooctanesulfonyl Fluoride Manufacturing Workers (Nov. 21, 2004). Page 3MA00755720. ttttt 3M's Dr. Butenhoff agrees, and recognized as early as 2001, that a study of the effects of PFOS on rats showed "significant increases in liver, pancreatic, and thyroid tumors," noting that "[h]epatocellular combined adenoma and carcinoma is clearly significant in the femals at the high dose," and that the "responses in the males [are] just over the border on the positive side" (3MA02608788). 83 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN co mp a r e d to controls, but similar differences also occurred in lipophilic contaminants (such as PCBs), and it is impossible to determine the possible contribution b y P F C s alone. 2. Toxicological evidence a. Cancer effects in h u m a n s from P F C s are supported by experimental animal studies. T h e I A R C evaluation considered the published evidence regarding m e c h a n i s m s of PFOA-associated carcinogenesis to be moderate, wh i c h did not lead to a change in the overall classification of P F O A as a G r o u p 2 B carcinogen [6]. I understand that P F O A is currently being tested in two-year bioassays b y the National Toxicology Program. b. P F O A w a s examined for carcinogenicity b y the oral route of exposure in t w o studies in rats, with s o m e initiation-promotion studies, as reviewed b y the I A R C in the evaluation discussed above [6]. c. T h e results from a rat bioassay sponsored b y 3 M w e re submitted to the E P A in 1983 and almost 30 years later released in a journal publication in 2 0 1 2 [225]. T h e results of this 2-year study docu me nt ed dose-related P F O A - i n d u c e d liver tumors and Leydig cell tumors of the testicles [224], and subsequent review suggested effects on pancreatic acinar cell a d e n o m a and carcinoma, while m a m m a r y gland lesions were found not to reflect possible breast cancer development [6]. d. Results from a second rat study s h o w e d elevated incidence of hepatocellular adenoma, testicular Leydig cell adenomas, and pancreatic acinar cell a d e n o m a and carcinoma [301]. S o far, n o bioassay has been conducted in another m a m m a l species. c. In mice, P F O A exposure induced stromal hyperplasia in m a m m a r y glands at 18 months, an effect that is hypothesized to increase susceptibility for tu mo r growth in rodents and h u m a n s [9]. f. Certain scientists, particularly those at 3 M , argue that a PPAR-related m e c h a n i s m m a y explain liver carcinogenicity in s o m e animal models. Ho we ve r, as discussed above, and as concluded in a recent risk assessment, both h u m a n and m o u s e P P A R - a l p h a are activated by P F O A in vitro [10]. W h e n 3 M discussed D u p o n t 's results on cancer in male rats with colleagues from the U K c o m p a n y ICI in 1995, the latter strongly espoused that A P F O should be considered an animal carcinogen, as the benign tomours observed are simply early lesions that ultimately lead to malignant tumors (a v i e w that is in agreement with interpretations adopted b y the I A R C , PG), but the 3 M representatives diagreed.uuuuu g. Also, certain studies have used specially-bred rodents to control for any such possible affect of the P P A R . In any event, E P A guidelines suggest that the n o n - P P A R dependent tumors, wh er e available data do not justify establishing a rodent-specific m o d e of action, should be pr es um ed to be relevant to h u m a n s [7, 17]. Still, the liver does not appear to be a primary target for cancer in humans. uuuuu 3MA10024469.pdf. 84 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 3. Perspective Evidence o f carcinogenicity in hum ans can be somewhat equivocal, m ainly when occupational populations are small and relatively young, follow-up durations are short, historical exposure levels are uncertain, individuals have had simultaneous exposures to other compounds, when cancer cases may be incom pletely ascertained, and when preexisting conditions may complicate the evaluation. The imprecise or incom plete data often lim it the inform ation that can be extracted from these studies [8 ]. Still, the w eig h t o f the evidence show s a likely PFC carcinogenicity, as the C 8 Panel for exam ple found for testicular cancer and kidney cancer [296]. W hen the num bers suggest a lack o f statistical significance, that is only one side o f the coin. The other side regards the m agnitude o f a possible adverse effect that could have been overlooked, given the available data. Some researchers m ight conclude that the absence o f a statistically significant excess risk suggests that the risk is absent, while others, including myself, representing w hat is becom ing the leading view, believe that such results m ust be interpreted in light o f the total inform ation at hand [79]. H ow large an effect could be overlooked or ignored as `n on -sig n ifican t'? O ften, available studies o f young w orking populations w ith lim ited follow -up, possible incom plete ascertainm ent o f cases, w ith healthy w orker bias, and other limitations, cannot provide confidence th at a risk is absent (see T able 1). a. Regarding occupational m ortality studies, a retrospective cohort study was completed in 1980 by Dr. M andel in 1980 at the 3M Chem olite plant (Cottage Grove, M innesota) site, as sum m arized in D r. G illilan d 's th e s is vvvvv N o significant findings w ere observed, but the standardized m ortality ratio for some cancers exceeded one (e.g., for cancer of the prostate and testis), a finding that would seem worrisom e, given the anticipated healthyw orker effect and short follow-up time. Continued surveillance for m ortality was said to be w arranted.TM TM ' b. In the m ortality study o f PFO A production w orkers carried out as part of Dr. G illiland's thesis project, a 3.3-fold increase in prostate cancer m ortality com pared to no em ploym ent in PFO A production, although based on only six prostate cancer deaths [72]. W hen 3M sent their custom ers (Eastech Chemical Inc., Tim berland Com pany, and W .L. Gore & Assoc.) a list of published articles regarding w orker safety and fluorochem icals, they inexplicably did n ot include D r. G illilan d 's m ortality p aper.xxxxx c. In a retrospective cohort m ortality study from 1995 on 3M em ployees in D ecatur Alabama, data were collected through 1991. N o significant excess in m ortality was seen. It w as recom m ended to follow up in 1998 w hen the 1992-1996 data w ould be available.yyyyy Presumably, this may have been the origin of the observations on bladder cancer in workers employed through the end o f 1997, published in 2003 [285], while a later follow-up report in vvvvv AR226-0473.pdf, Frank Davis Gilliland, Fluorocarbons and Human health: Studies in an Occupational Cohort (October 1992) (unpublished Ph.D. theses, University of Minnesota), with Summary. Pages 003173, 003217. wwwwwAR226-0470.pdf, Leonard M. Schuman & Jack S. Mandel, an Epidemiologic Mortality Study of Employees at the Chemolite Plant (February 1980), with Summary of Study. Page 003126. xxxxx 3M_MN01238940. Published Articles Re Worker Safety And Fluorochemicals. Page 3M_MN01238940.. yyyyyAR226-0032. Mortality Study of Employees at 3M Plant in Decatur, Alabama. Page 001643. 85 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 2007, also b y 3 M scientists, challenged those findings of excess bladder cancer, as ascertained b y postal questionnaire only [290]. d. A t the request of the C 8 Panel, a similar retrospective follow-up study in W e s t Virginia included over 6,000 m e n and w o m e n empl oy ed during 1948-2002 and followed u p through 2002. T h e results found little deviation that w o u l d suggest an excess cancer risk [73], although the findings as well could also not exclude the presence of an importantly elevated risk. c. In regard to experimental toxicology studies, early in vitro mutagenicity assays commissioned b y 3 M concluded that P F O A w a s not mutagenic under the test conditions,zzzzz and this finding has been replicated in several studies since then, thus suggesting that P F O A genotoxicity is not a relevant m e c h a n i s m [6]. f. A m o n g other potential mechanisms, immunotoxicity m a y be involved (see Section A). T h e genotoxicity question gave rise to s o m e controversy at 3 M . In connection with a 1999 13-week dietary toxicity study on N - M e F O S E exposed rats, Dr. A n d r e w Seacat from 3 M asked Dr. Peter J. T h o m f o r d from the contractor, C o v a n c e Laboratories Inc., to increase the calculated N O A E L , and 3 M refused to accept the study report.aaaaaa A mutagenicity study w a s not approved, the m a i n concern being that a c o m p o u n d with lower P F O S concentrations (T6316) had been used instead of in terms of T-6906, and the reported data w e re insufficient to reach a valid conclusion regarding the mutagenic activity of the substance in this assaybbbbbb Although differences in opinion m a y well occur, a variety of validated genotoxicity tests are available, including s o m e that rely on h u m a n white blood cells. Still, although insufficient testing w a s carried out at the time, negative results have been obtained in the vast majority of genotoxicity tests carried out since then [6]. g. A n im al toxicity studies carried out b y 3 M in 1978 caused s o m e discussions with D u P o n t on w h a t to conclude. O n M a r c h 5, 1979, N a n c y C h r e w r y from D u P o n t reviewed 3 M 's n e w C-8 rat and m o n k e y studies and wrote in a letter to Bill Krauss at 3 M that there were compound-related effects indicated in the animal studies, and that additional adverse effects apparently w e re revealed in the data though not reported b y 3 M in the description of the studies. cccccc A t a meeting of 3 M scientists in 1979, recent results from the Fluorochemicals in Bl o o d Pr o g r a m and the data from the 90-day subacute toxicity studies on P F O S and P F O A , J.R. Mitchell c o m m e n t e d " S o m e of the s y m p t o m s in animals from these 90 day studies are similar to zzzzz3MA02512309. F.D. Griffith and J.E. Long. Animal Toxicity Studies with Ammonium Perfluorooctanoate. aaaaaaAR226-0301. Letter from 3M to Covence regarding errors in Audited Final Draft Report [226-0300]. Page 005100. bbbbbbAR226-0254. Letter to 3M from Covence Laboratories reviewing the Study (226-0253) and concluding that it was technically inadequate. Page 001393. ccccccAR226-1456. March 5, 1979 - DuPont reviewed 3M's new C- 8 rat and monkey studies and agreed that there are compound-related effects indicated in both studies, and that additional adverse effects apparently were revealed in the data but not reported by 3M in the text of the studies. (Exhibit P (EID 123133)). Page 000140. 86 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN those observed w ith carcinogens. "dddddd H ow ever, such concerns about carcinogenicity apparently did not lead to any further 3M studies in the short term. h. R egarding a study conducted at R iker Laboratories, Inc. in 1983, 3M and an independent research institute again had different interpretation on the tum origenesis data obtained from a study to determine the chronic toxicity and carcinogenic potential o f the FM 3924 (N-EtFOSE) in rats. 3M reported "the overall incidence o f hepatocellular adenomas and carcinom as w as low in both control and FM -3924-treated groups w ith only the high-dose female rats possibly having a tum or incidence outside historical control lim its . . . . Based on tum or incidence, types of tumors, onset tim e of tum or appearance, malignancy patterns of tumors and the final m ortality values at tw o years, FM -3924 was not considered to be carcinogenic in the rat."eeeeee L ater in 1988, 3M am ended the report to em phasize th at the results w ere not statistically significant. The text revision was "the overall incidence o f hepatocellular adenomas and carcinom as was low in both control and FM -3924-treated groups w ith only the high-dose fem ale rats possibly having a tum or incidence that, while not statistically significant, w as outside historical control lim its." ffHff In 1998, however, the tum origenesis data were sent to Pathology Associates International (PAI, W est Chester, OH) for review. 3M regarded the original pathology interpretations as adequate, and exam ination o f m icroscopic tissue sections was not included in the review process. On Novem ber 25, 1998, PA I concluded that liver cancer in fact was correlated w ith the PFC exposures, stating "It is my opinion that dietary FM -3924 for 2 years resulted in chronic liver changes (m egalocytosis) in m ales at all dose levels (10, 30, 100 ppm ) and for fem ales at the high dose concentration (100 ppm). . . Incidence values for liver proliferative lesions indicated that FM -3924 should be regarded as a liver carcinogen for SD rats." gggggg VIII. PFHXS, PFBA, AND OTHER SHORT-CHAIN PFCs To date, PFH xS, PFB A and other short-chain PFCs have been less studied than longer-chain PFCs such as PFO A and PFOS and their precursors that were commercially dom inant for many years, and as a result they understandably garnered m ost of the attention from the scientific community. The scientific comm unity lags, and often cannot keep up with, changing industry strategies and technologies, particularly as to PFCs as to which com panies like 3M can develop and synthesize a large num ber o f PFC variants, including precursors. Shortchain PFCs have been less studied despite the fact that hum an exposures to short-chain PFCs, including PFBA, are apparently increasing [302]. For example, w hen 3M analyzed serum samples from the D anish national birth cohort, only PFO A and PFOS w ere quantified [254]. The dddddd3MA10034826. Meeting Minutes Meeting With Jr Mitchell # Re Review Of Recent Results Relevant To The Fluorochemicals In Blood Program (1979.04.26). Page 3MA10034828. eeeeee 226-0257. Two Year Oral (Diet) Toxicity/Carcinogenicity Study of Fluorochemical FM-3924 in Rats, Volume 1, age 001488. ffffffAR226-0262.pdf, Report Amendment No. 1: Two Year Oral (Diet) Toxicity/Carcinogenicity Study of Fluorochemical FM-3924 in Rats. Page 003085. ggggggAR226-0264. Pathology Review of Reported Tumorigenesis in a Two Year Study of FM-3924 in Rats. Page 003104. 87 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN M D H has determined that insufficient evidence is available for a proper risk assessment of both P F H x S and PFBA. Certain scientists, particularly industry-affiliated scientists, have a s su me d that short-chain P F C s such as P F B A have lower or n o toxicity based on P F B A 's short apparent elimination half-life ( T / ) in blood serum. Ho we ve r, a recent study sh o w s that P F B A is retained significantly in h u m a n kidneys and lungs, rather than blood. In this autopsy study, P F C s were measured in different organs. Short-chain P F C s are barely detectable b y standard L C - M S techniques in blood serum, but in h u m a n b o d y organs P F B A s h o w e d the highest concentrations in h u m a n lung tissue, and P F B A also w a s the predominant P F C in h u m a n kidneys [52]. Thus, the fact that P F B A appears in lower concentrations in blood than the longer-chain P F C s is probably misleading as to its toxicity. Rather, given that P F B A is k n o w n to accumulate in the kidney and the lung, it is reasonable to assu me that P F B A exposures likely contribute to target organ effects, especially in kidneys. P F H x S is retained in several organs, including the brain [52].Animal studies already have be g u n to bear this out, as I discuss below. T h e short-chain PFCs, moreover, are equally as bio-persistent as the long-chain PFCs, in that they are synthetically made, are not found in nature, and do not degrade. T h e molecular structure of P F B A and other short-chain P F C s also are highly similar to the structure of long-chain PFCs, particularly at the end group of the chain. Given the similar structure of P F B A and other short-chain P F C s to more-studied PFCs, particularly at the end group of the chain, and their equal bio-persistence, and given the adverse h u m a n health effects that have been s h o w n as to P F O A and P F O S , m a n y in the scientific c o m m u n i t y have called for short-chain P F C s to be as s u m e d to be, and treated as, equally risky. In their M a d r i d Statement, this large group of scientists in the field concluded: "Wh i l e s o m e shorter-chain fluorinated alternatives s e e m to be less bioaccumulative, they are still as environmentally persistent as long-chain substances or have persistent degradation products. Thus, a switch to short-chain and other fluorinated alternatives m a y not reduce the amounts of P F C s in the environment. In addition, because s o m e of the shorter-chain P F C s are less effective, larger quantities m a y be needed to provide the s a m e performance." These colleagues also noted that, "Wh i l e m a n y fluorinated alternatives are being marketed, little information is publicly available on their chemical structures, properties, uses, and toxicological profiles" [81]. In an ac co mp an yi ng editorial to the M a d r i d Statement w h i c h I co-authored with Dr. Linda Birnbaum, Director of the National Institute of Environmental Health Sciences, w e emphasize concern about the "potential risks of the short-chain" PFCs, and call for studies especially regarding low-dose endocrine disruption and immunotoxicity [303]. In response to these statements b y academic researchers, the industry FluoroCouncil agreed to m u c h of the content, but stated that "the short-chain P F A S substances studied to date are not expected to h a r m h u m a n health or the environment," as they "are eliminated m o r e rapidly from the b o d y and are less toxic than long-chain substances" [304]. Wh i l e that m a y be true for blood serum, short-chain substances like P F B A clearly linger in certain organs [52], as already mentioned, and the potential risks of these c o m p o u n d s are in no w a y clear. T h e y therefore clearly cannot be considered " safe" at this point [303]. 88 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN E v e n with f e w studies concerning P F B A , and the s o m e w h a t greater n u m b e r addressing P F H x S , m a n y of those have s h o w n adverse effects. Indeed, even early 3M-sponsored studies s h o w e d adverse effects from P F B A in laboratory animals in the liver, thyroid system, cholesterol levels, and negative developmental effects on the next generation. I will briefly describe these studies, w h i c h are supportive of m y opinion. For P F H x S , several epidemiology studies have already been mentioned, and a brief s u m m a r y is given here. a. T h e autopsy study already mentioned analyzed a total of 21 P F C s in 99 sets of tissue samples from autopsies in Spain (brain, liver, lung, bone, and kidney) [52]. In kidney and lung, P F B A w a s the mo st frequently determined P F C , and at the highest concentrations (with medians of 68 and 141 ng/g, respectively, i.e., greatly in excess of concentrations measured on blood, including those of major P F C s like P F O S and P F O A ) . L u n g tissues accumulated the highest total P F C concentrations. P F H x S is retained particularly in liver, kidney, lungs and brain. T h e authors suggest that the accumulation of different P F C s in h u m a n tissues should be of high importance for the validation of toxicokinetic models and should lead to further studies on the distribution of P F C s in the h u m a n b o d y to help interpreting n o n detectable concentrations in serum. b. Experimental animal studies on P F B S likewise s h o w a distribution in the b o d y that differs from P F O S [305], supporting the v i e w that one should not assu me that serum concentrations necessarily reflect organ retention levels. c. In its review of P F B A , the M D H referred to the following adverse effects seen in experimental studies: delayed eye opening and delayed vaginal opening, as mention in section A, changes in cholesterol, thyroid function, and liver function [63], as referred to above in the relevant sections. For P F H x S , the M D H noted adverse effects on thyroid function in a toxicity test study and certain epidemiology findings [306], but considered the evidence too incomplete to derive an exposure limit. d. P F B A has been s h o w n to modulate gene expression and cause enlarged liver size and toxicity through a PPAR-related m e c h a n i s m that reflects similarities with P F O A toxicity [307]. M o s t of the effects also occurred in P P A R - a l p h a hu ma ni ze d mice, thus suggesting that species differences w e re of limited consequence. c. A m o n g k n o w n effects of the related C4, P F B S , is inhibition of the e n z y m e aromatase, an effect that w o u l d suggest potential endocrine disruption, and changes in lipid metabolism, in placental cells [199]. f. Epidemiological evidence cited above shows that elevated s e r u m - P F H x S concentrations are associated with lowered antibody responses to vaccinations [35, 126, 128], lowered fecundity [159], abnormalities of menstrual cycles [167], lowered sperm count [170], greater risk of miscarriage [176], changes in birth weight [188], hormonal changes [193], including thyroid h o r m o n e changes [222], and metabolic changes [242]. Ho we ve r, the toxicology evidence is far from sufficient, and proper animal bioassays are missing. A 3 M commissioned rat study of developmental P F H x S exposure noted that the pups were "potentially" exposed via milk [308], although an article published about the s a m e time [309] 89 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN s h o w e d detectable concentrations of P F H x S in h u m a n milk collected in Massachusetts [309]. Bo th articles w e re authored b y 3 M 's Dr. Butenhoff. A s major adverse effects have only recently been addressed in scientific publications, n e w evidence in this area will likely emer ge in the future. Thus, the true risk of PFC-associated adverse health effects is probably even greater than it appears so far, especially for P F B A , k n o w n to accumulate in h u m a n tissues, though not in blood. That said, even based on the evidence to date, given the similar structure of P F B A and P F H x S to other more-studied P F C s particularly at the end group of the chain, given the adverse h u m a n health effects that have been s h o w n as to P F O A and P F O S , and given the adverse effects already s h o w n as to P F B A and P F H x S , it is m y opinion that both P F B A and P F H x S pose a substantial present and potential hazard to h u m a n health. IX. RISK ASSESSMENTS AND CURRENT LIMITS FOR PFC EXPOSURE W h e n translating scientific insight and documentation into public rule making, the evidence is scrutinized in a process usually referred to as risk assessment. If the evidence is substantial, it is often easier to reach a conclusion, but oftentimes, there are major reservations or limitations that m a k e it difficult to conduct a formal evaluation. In Section V, I described s o m e of the major caveats in regard to drawing conclusions on incomplete evidence. T h e weaknesses of risk assessment as currently practiced we re highlighted b y a committee of the U.S. National Research Council ( N R C ) [111]. A particular concern w a s the default assumptions (or lack of same), in particular the so-called "untested chemical assumption," i.e., that a chemical is innocuous, unless testing sh o w s otherwise. A s noted by prominent scientists from the U . S E P A earlier this year, risk assessment has failed w h e n adverse health effects are demonstrated at exposure levels predicted from animal studies to be safe for h u m a n s [310]. That seems to be the case regarding the PFCs. T o further illustrate this issue, I recall h o w I participated in the drafting of E F S A 's opinion on P F O S and P F O A in 2007-2008 [1]. T o begin with, w e received a tentative list of topics to cover in a logical sequence, i.e., a draft table of contents. O n the list, there w e re a variety of issues for consideration. If there w a s n o evidence, w e should just leave out the topic. A t the time, there w a s little evidence on immunotoxicity, so it w a s not discussed. T h e deletions did not have any impact on the conclusions, wh i c h we re based solely on the evidence available. A s noted b y the N R C , such default decisions are inappropriate and could be invalidated b y subsequent discoveries, i.e., in regard to immunotoxicity of P F C s in the present case. In the present context, t w o m a i n concerns will be highlighted, i.e., the identity of the population(s) at greatest risk, and the likely target organs that m a y suffer adverse effects at the lowest doses. This section will also include a brief review of exposure limits. Pregnant w o m e n and y o u n g children mu st be considered populations at increased risk due to the vulnerability during early development. Other population groups at increased risk include adults with pre-existing diseases of the kidneys, the liver, or the cardiovascular system, in particular those with a high intake of water. 90 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN Regarding target organs, m o st risk assessments have focused on the liver, as rodent studies have clearly d o cu me nt ed increased liver size and s o m e functional changes as being strongly related to elevated P F C exposures. Ho we ve r, species differences, particularly in regard to P P A R expression m a y complicate the translation of rodent data to the h u m a n situation. Other potential target organs, such as the i m m u n e system and the endocrine system, have been considered in animal studies, but only in part covering the subtler effects that are of concern in regard to h u m a n health. Thus, reliance on animal studies of liver toxicity n o w appears not to protect adequately against adverse effects on the latter organ systems. Adverse effects on i m m u n e functions and on breast development have been recently docu me nt ed at background exposure levels. Consequently, m a n y U.S. regulatory agencies have set current limits for P F C s in drinking water that appear insufficient, to protect against adverse health risks, especially in vulnerable subgroups. B e n c h m a r k calculations based on decreased response to vaccine suggest that existing limits m a y be 100- to 1000-fold too high [150]. A t this time, the U N Stockholm Convention included in 2 0 0 9 P F O S and its precursors on the list of substances to be phased out, and it seems that P F O A will likewise be included in the very near future. T h e European Chemicals A g e n c y ( E C h A ) has already listed P F O A and recently added P F H x S to the Candidate List of substances of very high concern ( S V H C s ) for mandatory authorization procedures. A s these t w o groupings require substantial evidence to justify the listing, the fact that other P F C s are not included only m e a n s that sufficient evidence is not yet available, not that they are not persistent, bioaccumulative, highly toxic, and therefore undesirable. A s noted b y the N R C [111], risk assessments can take a very long time and can be difficult to revise, once a consensus has been obtained. Nonetheless, as scientific evidence accumulates, revision of the health limits w o u l d appear necessary. In the case of PFCs, the substances accumulate in the body, and present-day exposures will contribute to the b o d y burden for m a n y years in the future. Proposed calculations from 2013 [150], w h i c h relied on epidemiology findings, have not yet been applied in formal risk assessments. A. Drinking water limits T h e limits set b y the E P A , the states, and various foreign governments vary from each other (see Table 3), based on values and assumptions used, and based on h o w recently the limits have been set or revised [311]. Differences between the acceptable limits are often due to differences in default values, e.g., for uncertainty factors used in the calculations, although estimates of water intakes and lifetime accumulation also differ. In addition, there is a clear tendency that limits decrease over time, as m o r e evidence b e c o m e s available. Table 3. E x a m p l e s of current limits for P F C concentrations (ng/L) in drinking water in different jurisdictions. PFC U.S.EPA MDH PFOS 70 27 PFOA 70 35 PFHxS - 300 PFBA - 7,000 Total P F C s - 91 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN Canada Sweden 600 200 -- ... - - 90 M ost lim its address individual PFCs, but the Swedish limit, for example, accounts for the sum of m ajor PFC contaminants. Likewise, some agencies have decided for a joint lim it for PFOS and PFHxS. Recent toxicology evidence suggests that adverse effects from some PFA Ss may not be additive, and synergistic effects have been identified [312, 313]. The epidemiology offers little guidance, and toxicology studies suggest that at least PFO A has a mode o f action that differs from the one o f PFOS [147, 247]. The present subsection will first summarize the m ost relevant limits, m ainly for concentrations in drinking water. The great variability in the num bers illustrated by the above table m ust be considered in light o f changing needs for exposure lim its in different settings, emerging scientific insight, and the latency within regulatory agencies in regard to developing new guidelines. Also, the limits have been developed based on different definitions and assumptions. In the subsequent subsection, I shall discuss the m ajor approaches to setting exposure limits. 1. Federal and state drinking water limits EPA In 2009, the EPA issued provisional health advisories o f 0.4 pg/L (400 ng/L) for PFOA, and 0.2 pg/L (200 ng/L) for PFO S [314]. A t the tim e, EPA concluded that " [e]pidemiological studies o f exposure to PFO A and adverse health outcom es in hum ans are inconclusive at present." Also, the evidence for the carcinogenicity o f PFOS is considered "suggestive of carcinogenicity." Similar conclusions were drawn in 2015, w hen EPA updated their previously proposed lim its for PFO A and PFOS in w ater to 0.07 pg/L (70 ng/L) for both, as based on calculations relying on the m ost recent toxicological and supporting data [148, 149]. The U.S. EPA has selected 0.00002 m g/kg/day (or 0.02 p g /k g d ) as the Reference D ose (RfD) for PFO A and 0.00003 m g/kg/day (or 0.03 p g /k g d ) as the RfD for PFOS. Incidentally, the RfD fo r P FO S is 6 -fold h ig h er than the safe reference level calculated by 3 M 's D r. B u te n h o ffhhhhhh A gency fo r Toxic Substances and D isease Registry The Agency for Toxic Substances and Disease Registry (ATSDR) first issued a draft toxicological profile in 2009, but concluded that there was insufficient evidence at the tim e to develop a m inim al risk level [17]. A n updated version from 2015 [4] again focused on the experimental animal studies to develop a M inim al Risk Level o f 0.02 p g /k g d ay for PFO A and 0.03 p g /k g d a y for PFOS, as the only PFC s that had sufficient evidence to allow this calculation. The M RLs are the same as the E PA 's RfDs. A TSD R stated that changes in birth w eight and serum liver enzymes in humans are small and not likely biologically relevant; the same applied to imm une function changes, and the findings in children were not even m entioned in the section discussing exposure limits. In fact, A TSD R rejected a calculated benchm ark dose for im m unotoxicity as the study did not include an unexposed group o f children (which is clearly hhhhhh 3MA01366219.pdf. 92 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN wrong). Still, if calculated for a w om an w eighing 50 kg (110 pounds), the M RL w ould result in intakes o f 1 and 1.5 pg per day, and allow ing w ater consum ption o f at least one liter per day to contribute up to 10% o f the daily PFC intake, these M RLs are in agreem ent w ith the EPA w ater lim its. M innesota D epartm ent o f H ealth The M H D in 2008 issued H ealth R isk Lim its (HRLs) for PFO S and PFO A, i.e., upper lim its in concentrations in w ater that are believed at the tim e to pose little or no appreciable risk to a person drinking the w ater [315]. The HRLs are proposed and adopted as rules by the State follow ing a public rule-m aking process. These tw o H R L values o f 0.3 pg/L (300 ng/L) w ere based on PFO S effects on the liver and thyroid, and PFO A effects on the liver, fetal developm ent, reduction in red blood cell numbers, and im m une system changes in experimental studies [292]. These lim its have recently been revised. In M ay o f 2017, M D H released its updated lim its o f 0.027 pg/L and 0.035 pg/L for PFO S and PFOA, respectively. The lowered lim its were considered to better protect the fetus and the breastfed infant. The M D H relied on the toxicology evaluation carried out by the EPA [148, 149], but then applied toxicokinetic calculations to take into account increased serum-PFC concentrations during infancy in connection w ith breastfeeding [37], w hile still em phasizing life-tim e exposures. This resulted in a decreased o f 50% or more, as compared to the EPA limits, and close to a 10-fold decrease com pared to the 2008 values. Notably, in regard to the RfD for PFOS o f 0.0051 pg/kg/day and the corresponding serum concentration o f 63 ng/ml, the M D H provides the follow ing caveat: "Note: this serum concentration is inappropriate to use for individual assessment. Serum concentration is useful for inform ing public health policy and interpreting population-based exposures. This value is based on population-based param eters and should not be used for clinical assessm ent or for interpreting serum levels in individuals." This statem ent w ould generally apply to such calculations, although in this case, where a m uch low er lim it is suggested by epidem iological evidence, it m akes less sense. Also, 3M 's Dr. B utenhoff calculated that a reference level in plasm a fo r chronic PF O S exposure from all sources w ould be 1.5 ppb, or 1.5 n g/m L .iiiiii Previously, M D H had established HRLs for PFBA and PFBS, with chronic exposure lim its o f 7 pg/L. These tw o HRLs w ere based on PFB A effects on liver, thyroid, blood, cholesterol, and developm ental changes, and PFBS effects on blood, kidney and liver changes in experimental studies. The RfD s established based on animal studies are 0.0042 m g /k g d ay (or 4.2 p g /k g d ay ) for PFBS and 0.0038 m g /k g d ay (or 3.8 p g /k g d ay ) for PFB A [316]. The corresponding w ater lim its are 9 and 7 pg/L. N either o f these tw o C4 com pounds have been exam ined w ith a view to im m unotoxicity, and likewise, no carcinogenicity data exist. M D H also released guidance regarding PFHxS that identified effects of PFHxS on body weight, blood, and cholesterol in experimental studies that occurred at a sim ilar human linn Op.cit. 93 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN equivalent dose as was determ ined for the PFOS HRL. M D H has indicated that while it awaits further information, the H B V for PFO S o f 0.027 qg/L will apply to PFH xS [306]. O ther U.S. states U sing slightly different assum ptions, the state o f N ew Jersey at first decided on a lim it o f 0.04 qg/L (40 ng/L) (and recently lowered that, see below ) for PFOA. N ew Jersey relies on an adult drinking w ater intake rate o f 2 liters per day (for a BW o f 70 kilogram s), w hich is about h a lf o f the w ater intake rate (at the 95th percentile) used fo r calculations in M innesota. N ew Jersey's revised lim its are among the m ost recent revisions [10], and it took into account a cancer risk assessment. Based on evidence on testicular cancer in rats, a cancer slope factor w as calculated, and a lifetime risk o f 1 x 1 0 " 6 w as found to correspond to a w ater concentration o f 14 qg/L. A lm ost the sam e level w as found w hen using liver w eight as a sensitive non-cancer outcom e and taking into regard uncertainty factors. Previously, scientists from N ew Jersey used data on breast developm ent in a rodent study to calculate BM D L for an endocrine disruption outcom e [7]. In this case, the B M D L w as translated to a safe serum -PFO A concentration in hum ans o f 0.8 ng/mL, thus suggesting that endocrine disruption m ay occur at low exposure levels w here im m unotoxicity is otherw ise the only adverse effect deocum ented so far. However, delayed m am mary gland developm ent w as not recom m ended as a critical effect because of lack of precedent for use of this endpoint as the primary basis for risk assessment [10]. Three independent epidem iological studies have now shown decreased duration of breastfeeding in lactating wom en at elevated PFC exposures (see section VII.C.1). Several other states have decided on PFC lim its for drinking water. Lim its published several years ago tend to be higher. For example, the State o f N orth Carolina used assum ptions sim ilar to those previously used by N ew Jersey, but N orth Carolina had Previously decided on a lim it o f 0.63 qg/L (630 ng/L) for PFOA. Although my review o f existing lim its for PFA S in drinking w ater may not be up to date, I have noted a range o f PFO A lim its from 0.014 qg/L (New Jersey) to 24 qg/L (Oregon). Likewise, PFOS limits for drinking w ater developed by state authorities vary from 0.011 qg/L (M ichigan) to 1.3 qg/L (A laska). A lthough som e o f the very high limits may have been rescinded, the range illustrates the im pact o f emerging evidence over tim e and o f state-level default practices. I have not found lim its for PFBS and PFBA, while A ustralia has a PFH xS lim it o f 0.5 qg/L. As far as I am inform ed, calculation o f w ater lim its are generally based on adverse outcom e m easures in animal studies (e.g., liver weight), and not driven by epidem iological evidence. In addition, cancer and delayed breast developm ent in anim als have been considered as critical effects by the state o f N ew Jersey only, at least so far. 2. Foreign exposure limits European F ood Safety Authority The European Food Safety Authority (EFSA) in 2008 released its recom mended T olerable D aily Intake (T D I) at 1.5 q g /k g d a y fo r P F O A and 150 ng/kg (0.15 q g /k g d a y ) for PFO S [1]. These lim its rely on the sam e anim al toxicology data as the assessm ents published by the EPA and the A TSD R about the same time, and the TDI values correspond to the initial RfD and M R L values arrived at. Som e E U m em ber states have later published national lim its for PFC 94 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN concentrations in drinking water, and s o m e of the m e m b e r states (such as Sweden, see above table) have decided for total-PFC concentration limits, i.e., including all of the (C8, C 6 and C 4 c o m p o u n d s considered here. These limits are therefore stricter than m o st current limits. E F S A is expected to release an updated opinion with revised T D I levels in the fall of 2017. 3. Fish consumption advice In the aquatic environment, s o m e P F C s accumulate in fish. In Mississippi River Pool 2, for example, P F O S w a s generally detectable in surface water samples u p to a level of 136 ng/L [23]. A s reviewed above under exposure sources, P F C contamination of waterways results in biomagnification in aqueous food chains, with P F O S causing m u c h higher accumulation than P F O A or P F H x S [20-23]. Av er ag e P F O S concentrations have been reported to be u p to 275 n g/g. Using similar risk assessment methodologies and calculations as described in regard to the water limit, below, with a meal size of 22 7 g fish for a 70-kg adult, the M D H has issued a health-based fish consumption advice of > 40 ng/g P F O S (1 meal/week) and > 2 0 0 ng/g P F O S (1 meal/month) for the several lakes in the T w i n Cities area and portions of the Mississippi River with elevated concentrations of P F O S in fish tissue. In total, specific consumption advice has been issued for 34 lakes and the Mississippi River Pool 2 [24]. B. Setting drinking water health limits Generally speaking, a health limit is calculated from a "point of departure," wh ic h in turn is based on a calculation starting from a b e n c h m a r k dose, or from a L O A E L or N O A E L . T h e point of departure is usually derived from animal toxicity studies reflecting a point of critical effect. A s cancer risk has not been formally applied so far, the discussion here focuses on n o n cancer risks, and a brief section b e l o w will summarize the approach to cancer risks. T h e point of departure is then adjusted using kn ow le dg e of the half-life of the relevant P F C in humans, and of the differences between h u m a n s and animals in the toxicity study in terms of water intake and retention, to arrive at a "h u m a n equivalent dose." Then, because P F C s might not affect h u m a n s in exactly the s a m e w a y that they affect, say, cyno mo lg us m o n k e y s or laboratory rats tested in toxicity studies, w e adjust the h u m a n equivalent dose b y applying "uncertainty factors." There are multiple types of uncertainty factors that m a y be appropriate, depending on the nature of the animal study, w h a t is k n o w n about the differences between the animal in the study and humans, and the strength of the "database" of know le dg e about the health effects of P F O A . O n e uncertainty factor that might be applied concerns potential differences in toxicodynamics (differences in the flow of P F C s within the bodies of animals versus humans). Another concerns intraspecies variability (recognizing that P F C s might affect h u m a n subpopulations differently, including m o r e vulnerable populations such as children). Other uncertainty factors that m a y be appropriate include a "subchronic to chronic extrapolation" (for w h e n w e have a subchronic exposure value, but there is a risk that a chronic exposure could lead to h a r m at lower levels), a " L O A E L - t o - N O A E L " extrapolation (for w h e n w e have evidence from studies of a l o w level at wh i c h adverse effects occur, but n o g o o d data on an exposure level at w h i c h n o effects occur), and a "database uncertainty" factor (for 95 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN when there have not been enough animal studies o f certain types to have some degree of confidence that the results are a reasonable measure of risk and harm). The various uncertainty factors are then m ultiplied to arrive at a total uncertainty factor. The hum an equivalent dose is divided by the total uncertainty factor leading to a "reference dose." The reference dose is then plugged into a form ula to arrive at a health limit, for example a form ula like this: (Rfd) x (RSC) x (Conversion factor) (Chronic intake rate, L/Kg/d) W here RfD is the reference dose, RSC is the "relative source contribution," the conversion factor changes m illigram s into m icrograms, and the "chronic intake rate" is the rate chosen to represent the tim e-w eighted average intake rate over the num ber o f years estim ated to achieve a steadystate serum concentration based on the half-life o f the PFC. The RSC is often a default based on EPA guidance such as 0.2 (or 20% ), representing an assum ption that 80% o f a person's exposure to the PFC com es from non drinking w ater sources. A ppropriate values for an RSC typically range from 0.2 (com m on) to 0.8 (more appropriate for a pharm aceutical drug w here exposure from other sources is less likely). If there is additional inform ation about exposure sources for a chemical, there m ay also be justification for departing from the 0.2 default. The chronic intake rate depends on the half-life of the chemical and reasonable estim ations o f how m uch w ater people drink. There is a different rate for calculating an acute dose level, or a sub-chronic level. As cancer is a likely outcom e o f PFC exposure, I shall briefly refer to a recent calculation o f cancer risk, although this has not been form ally applied so far. I shall then review the methods for non-cancer risks used in risk assessments. 1. Cancer slope factor The basic assum ption generally applied is that the cancer risk increases w ith the dose. A lthough this may not be correct for non-genotoxic substances, such as the PFASs, the calculation will m ost likely err to the conservative side and provide im proved protection. The goal is then to calculate the exposure level that will cause a life-tim e cancer risk o f one in a m illion. N e w Jersey arrived at a cancer slope factor o f 0.021 (m g/kg- day)-1 fo r PFO A , as based on increased incidence o f testicular tum ors in a chronic rat study. N ew Jersey used this slope factor to develop a health-based lim it protective for cancer effects at the 1 - 10"6 (one in one m illion) lifetim e cancer risk level, arriving at a w ater concentration lim it o f 0.014 pg/L (14 ng/L) [10]. This lim it is identical to N ew Jersey's health-based lim it based on non-cancer endpoints, and is one-fifth o f the m ost recent w ater lim it value proposed by the EPA [148]. M y conclusion is that, on the basis o f current evidence, cancer risk alone is unlikely to drive the risk assessm ent for PFASs, because specific target organ effects are likely to 96 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN be at least as important. This conclusion is supported b y the observation that the P F A S s appear not to be genotoxic. A n y carcinogenicity is likely due to other mechanisms, such as immunotoxicity and endocrine disruption, i.e., non-cancer endpoints that are already considered highly vulnerable to P F A S exposures. 2. Benchmark dose T h e various regulatory agencies appear to be in overall agreement in using a b e n c h m a r k dose mo d e l to calculate non-cancer health limits for drinking water [2]. In this method, a dose-response function is fitted to the data or adopted from a default assumption. T h e b e n c h m a r k dose ( B M D ) is defined as the dose wh i c h leads to a specific loss (or degree of abnormality) k n o w n as the b e n c h m a r k response ( B M R ) in the ou tc om e variable. T h e B M R mu st be specified before the analysis. In epidemiological studies, a 5 % change is often used for the B M R . A larger B M R will lead to a higher B M D . T h e statistical uncertainty in the B M D estimation is taken into account b y calculating its lower one-sided 9 5 % confidence limit, the b e n c h m a r k dose level ( B M D L ) . T h e B M D L is then used as the point of departure for calculation of the exposure limit. A n advantage of this approach is that reliance on smaller and less certain studies, everything else being equal, can result in m o r e protective standards, wh i c h w o u l d be in accordance with the precautionary principle [317]. C. Applications of benchmark dose calculations A c o m m o n endpoint that has been used to set a b e n c h m a r k dose is an increase in liver weight observed in rodent studies. Similar nu m b e r s have been used b y the E P A , E F S A and M D H for their standards setting efforts. For P F O A , the M D H relied on a B M D L of 23 p g / m L serum, wh i c h w a s estimated to correspond to a h u m a n equivalent dose of 0.0023 mg/kg- d based on the uptake, distribution within the body, and elimination of P F O A b y h u m a n s (first order kinetics) [292]. Similar calculations s h o w e d a B M D L at 3 5 p g / m L for P F O S in serum, with a h u m a n equivalent dose of 0.0025 mg/kg- d. Uncertainty factors we re included, while also taking into regard that the data base regards subchronic exposure effects only. T h e M D H averaged drinking water exposure over each period (starting from birth) using data from national studies of large n u mb er s of people. Intake (using the 95th percentile of intake) over the first 19 years of life is 0.053 L/kg - d and intake over the first 27 years of life is 0.049 L/kg- d. Drinking water standards generally are calculated to allow for other exposure sources, so that the dose from drinking water contributes a portion (usually 20 percent) of the reference dose. B e n c h m a r k dose calculations for P F C s have been carried out using data from other toxicological studies, including a study in pregnant mice [318] and a study on breast development in pups [7]. T h e N e w Jersey committee [10] s h o w e d that these findings w o u l d result an R f D as l o w as 0.11 ng/kg - day - m u c h b e l o w the concentration levels relied on b y the 97 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN M D H - after consideration of the uncertainty factors. W h e n modeling the results in terms of s e r u m - P F O A concentrations, they s h o w e d that the Target H u m a n S e r u m Level (analogous to the R f D expressed on a serum level basis) w o u l d be 0.8 ng/mL, i.e., b e l o w the m e di an serum P F O A level in the U.S. general population. Again, the N e w Jersey committee refrained from using this information for calculating a safe drinking water limit. D. Comparison of existing limits with recent research information Recent evidence have identified adverse effects of P F C s on sensitive outcomes in laboratory animals following developmental exposure [9] and on i m m u n e system functions in children [35]. It is then appropriate to take into account the existence of likely effects that target the p r o g r a m m i n g of organ system function and future disease risks [248]. W h e n developmental toxicity is likely, a National Research Council committee 20 years ago proposed to include an extra 10-fold uncertainty factor to protect children against food contaminants [319]. Using existing data and b e n c h m a r k dose calculations, m o r e precise results can be obtained that reflect the importance of developmental vulnerability, rather than average sensitivity during decades of exposure. M o r e specifically, as noted b y the N e w Jersey committee [10], data n o w available on m a m m a r y gland development in mice suggest that clear delays result from m u c h lower developmental exposures at s e r u m - P F O A concentrations of about 25 n g / m L [9]. Immunotoxicity has been demonstrated in mice at concentrations of the s a m e order of magnitude [141]. Using these data, and m a k i n g b e n c h m a r k dose calculations using a 1 0 % B M R and an interspecies 10 fold uncertainty factor, results in a B M D L s e r u m - P F O A concentration of 23-25 n g / m L [7]. O n e could argue that this m o u s e experiment w a s not based on chronic exposures, so that additional default 3-fold uncertainty factor w o u l d be appropriate, thus reducing the B M D L to 8 ng/mL. Formal limits for water or total intake will likely differ between jurisdictions, as default values used b y different agencies will differ and will therefore result in slight deviations. Ho we ve r, m y opinion is that, as science improves, P F C exposure limits in the various jurisdictions will continue to decline and w e will later v i e w m a n y of the current exposure limits for P F O A and P F O S as too high, likely b y a factor of 100 to 1,000. Moreover, as science improves and drinking water limits are lowered to account for increased concern about the adverse effects of PFCs, the n e w limits typically refer to long term exposures, perhaps for a lifetime, not to exceed the limit. T h e y do not take into account past exposures above that level. Ho we ve r, as the water limits in Minnesota have been applied and subsequently lowered, m a n y people, particularly in the East Me t r o area, we re in the past exposed to P F C s in excess of the drinking water limits, especially the mo st recent visions - in s o m e cases, for decades. T h e P F C b o d y burdens of those individuals w o u l d reflect a history of higher exposure than w o u l d be contemplated under a newly-adopted limit, and, without an entirely PFC-free source of drinking water, the P F C b o d y burdens of these individuals w o u l d exceed that contemplated b y the revised drinking water limits for m a n y years, if not decades. 98 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN E. Proposed limits based on recent epidemiological evidence B e n c h m a r k doses derived from animal studies and the decisions on applicable uncertainty factors needed for extrapolation to h u m a n s always include s o m e degree of uncertainty, and reliance on h u m a n data w o u l d be preferable. O n the other hand, while b e n c h m a r k dose calculations from toxicology studies are fairly straightforward, using epidemiological studies is m o r e complicated due to the need for covariate adjustments, the absence of a non-exposed control group, and m o r e co m p l e x decisions on the shape of the doseresponse curve [317]. Solutions to the statistical problems can be obtained both b y the so-called hybrid m e t h o d and relative methods [2], but decisions on dose-response models m a y have important consequences for the B M D L results. In our recent study of immunotoxicity [35], five P F C s w e re measured to assess prenatal and postnatal exposure in regard to associations with concentrations of t w o specific antibodies against childhood vaccines. T h e antibody responses could be categorized either in terms of the concentration as such or whether it w a s b e l o w the clinically protective concentration of 0.1 IU/mL. W i t h a b e n c h m a r k response of 5% , w e calculated that the B M D L for the diphtheria antibody concentration at age 7 years is 0.7 ng/mL, 0.6 ng/mL, and 1.1 n g / m L for the child's serum concentrations of P F O S , P F O A , and P F H x S , respectively. Results for tetanus were higher. For prenatal exposure in regard to diphtheria antibody concentration at age 5, the results we re 0.6 ng/mL, 1.0 ng/mL, and 3.4 n g / m L for the s a m e PFCs. For prenatal exposure in regard to the limit of 0.1 I U / m L for the diphtheria antibody concentration at age 5 years, the results we re 0.8 ng/mL, 0.4 ng/mL, and 1.0 ng/mL. Linear dose-response curves generally resulted in higher B M D L results, likewise also with a B M R value of 1 0 % [150]. These results are fairly similar and m u c h b e l o w the B M D L values from animal studies. Given the correlation of the major PFCs, it is difficult to include mutual adjustment in the models. Ho we ve r, in agreement with the regression results, the b e n c h m a r k dose levels suggested that P F H x S is less im munotoxic than P F O S and P F O A . O u r recent calculations s h o w that P F O A , after adjustment for other PFCs, sh o w s clear negative impacts on i m m u n e system responses to vaccinations, but it is not possible to separate the effects of P F O S and P F H x S [120]. O u r calculations rely on the association between early-life exposure to P F C s in regard to deficient i m m u n e system responses to vaccinations later on in childhood, as well as adverse developmental effects already considered b y the M D H and the E P A . Wh i l e the i m m u n e system-related associations are statistically significant, the exact curve shape for the associations id u n k n o w n , and assumptions m u st be m a d e to judge the approximate magnitude of the point of deviation, from wh i c h an exposure limit can be derived. In addition, the i m m u n e system-related calculations w e re based on serum concentrations measured at age 5 years, assuming that this level is representative for the exposure at the m o s t vulnerable age. This assumption m a y not be true [113], and the results m a y therefore be biased toward higher and less protective levels. Nonetheless, w e applied several different models and then applied standard default factors that correspond to routine practice within the E P A and other regulatory agencies. Thus, w h e n calculated based on both i m m u n e system related effects and developmental endpoints, current exposure limits in the United States remain too high. 99 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN F. Underestimation of hazards posed by PFCs A s discussed in section V.A, there are important tendencies that will result in underestimations of P F C toxicity. T h e previous section sh ow s that reliance on animal toxicity data with a focus on enlarged livers and similar routine outcomes from rodent toxicity studies can greatly underestimate the risk to h u m a n health. M o r e targeted studies on immunotoxicity and endocrine disruption have recently been carried out in mice and revealed adverse effects at m u c h lower exposures than those that lead to liver damage, especially w h e n exposures were determined on the basis of blood concentrstions and not on the a m o u n t in the feed. Subsequently, h u m a n studies demonstrated that deficient antibody responses to routine vaccinations occur at elevated background P F C exposures. Supporting studies showing m o r e frequent infectious disease in children at higher P F C exposure emphasize that the i m m u n e system is a highly vulnerable target organ. A s a result, recent scientific insight suggests that, as it relates to children's i m m u n e systems and endocrine disruption, further consideration of protective levels m a y be necessary. A s an indication h o w n e w data can reveal adverse effects at exposure levels previously thought to be safe, the A T S D R Toxicology Profile in 2 0 0 9 concluded that n o data we re available on immunotoxicity in h u m a n s [17], but the authors did not k n o w that our study in the Faroes w a s under w a y and w a s to be published a couple of years later [35]. T h e recent report from the N T P considers P F O S and P F O A "p r es um ed" immunotoxicants (the level just b e l o w "k n o w n ") [5], but this conclusion has not yet impacted risk assessments carried out b y regulatory agencies at state level or otherwise. Given the fact that m u c h of the leading scientific literature is fairly recent, the conclusions that can be d r a w n at this point mu st be regarded tentative to s o m e extent, and they m a y even represent underestimations. Current understanding of the P F C s is very different from 3 M 's assertion in 2000, that "the presence of these materials at these very l o w levels do not pose a h u m a n health or environmental risk" (see Section IV.D). Current understanding is also very different from 20 08 w h e n the first formal risk assessments w e re published [1, 315]. A t the present time, the m a i n uncertainty regards effects on the m o st vulnerable target organs and critical exposure conditions, such as prenatal exposures, that have yet to be studied in greater depth. T h e m o s t appropriate conclusion that can be d r a w n is that adverse effects on breast development and on adaptive i m m u n e system development likely represent critical effects and that B M D L s should focus on these outcomes. In drawing conclusions, w e m u s t still consider w h a t w e could possibly k n o w today from the types of evidence available to us, and w h a t w e do not yet know. In this light, m y above conclusions m a y well be underestimated. 100 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN X. AFFIRMATION I affirm under penalty of perjury that the foregoing is a true and correct statement of m y opinions in this matter and the grounds for those opinions. djean 22 September, 2017 101 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN EXHIBIT A ABBREVIATIONS ADHD attention deficit hyperactivity disorder ALT alanine transaminase APFO a m m o n i u m perfluorooctanoate; A P alkaline phosphatase, liver e n z y m e AST aspartate aminotransferase, liver e n z y m e ATSDR A g e n c y for Toxic Substances and Disease Registry BMD benchmark dose BMDL lower limit b e n c h m a r k dose BMR be nc hm ar k response BUN blood urea nitrogen B W body weight C 8 or C-8 perfluorinated octanol c o m p o u n d s CDC Centers for Disease Control and Prevention CI confidence interval EPA Environmental Protection A g e n c y EtFOSE see N - E t F O S E E t - P F O S A - A c O H 2-(N-ethyl-perfluorooctane sulfonamide) acetic acid FC-143 P F O A (with u p to 3 . 5 % C6, C7, and C 9 co mp o u n d s ) FC-807 N-ethyl perfluorooctane sulfonamido ethanol-based phosphate esters (metabolized into P F O S ) FC-95 see P F O S FDA F o o d and Dr ug Administration FEP fluorinated ethylene propylene FM-3924 see N - E t F O S E GGT gamma-glutamyl-transferase, liver e n z y m e GOT glutamic-oxaloacetic transaminase, s a m e as A S T HDL high density lipoprotein HRL Health Risk Limit IARC International A g e n c y for Research on Cancer L C / M S / M S liquid chromatography/tandem m a s s spectrometry LD50 lethal dose, 5 0 % kill LDL l o w density lipoprotein LOAEL lowest-observed-adverse-effect level MDH Minnesota Department of Health MeFOSE see N - M e F O S E NHANES National Health and Nutrition Examination Survey N - E t F O S E 2-(N-ethylperfluoro-1-octanesulfonamido)-ethanol N - M e F O S E 2-(N-methylperfluoro-1-octanesulfonamido)-ethanol NOAEL n o observed adverse effect level NRC National Research Council NTP National Toxicology Program 102 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 OR PFAS PFBA PFBS PFC PFDA PFHxA PFHxS PFNA PFOA PFOS PFOSA POSF PPAR T/ T2D T3 T4 T-6316 TFE TOF TSCA TSH odds ratio see P F C perfluorobutyric acid perfluorobutane sulfonic acid perfluorinated c o m p o u n d perfluorodecanoic acid perfluorohexanoic acid perfluorohexane sulfonic acid perfluorononanoic acid perfluorooctanoic acid; perfluorooctane sulfonic acid perfluorooctane sulfonamide perfluorooctanesulfonyl fluoride peroxisome proliferator activated receptor biological half-life type 2 diabetes triiodothyronine thyroxine N-EtFOSE-containing product tetrafluoroethylene total organic fluorine Toxic Substances Control Act thyroid-stimulating h o r m o n e Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 103 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN EXHIBIT B PHILIPPE GRANDJEAN, M.D. (CV) Office address Institute of Public Health University of Southern D e n m a r k Winsl0wp ar ke n 17 D K - 5 0 0 0 O d e n s e C, D e n m a r k Tel. (+45) 6550.3769 Fax (+45) 6591.1458 Email: pgrand@health.sdu.dk http://www.sdu.dk/staff/PGrandjean.aspx Home Nabol0s 4 DK-1206 Copenhagen Denmark Tel: (+45) 33 133 933 Harvard School of Public Health 10 D a n a Street Department of Environmental Health apt 315 L a n d m a r k Center, 3E-110 Cambridge, M A 02138 401 Park Drive Mailing address: P.O. B o x 15697 P.O. B o x 390589 Boston, M A 02215 Cambridge, M A 02139 Tel: 617-384-8907 Tel: 617-331-3317 Fax: 617-384-8994 Email: Pgrand@hsph.harvard.edu http://www.hsph.harvard.edu/faculty/philippe-grandiean/ A c a d e m i c degrees 1974, M.D., University of C o p e n h a g e n 1975, D i p l o m a in basic medical research, University of C o p e n h a g e n 1979, D.M.Sc. (dr.med.), University of C o p e n h a g e n Chronology of employment 1974- 1975 Postgraduate training fellowship, University of C o p e n h a g e n 1975- 1978 Research fellow, Institute of Hygiene, Univ. C o p e n h a g e n 1978-1980 Senior research fellow, University of C o p e n h a g e n Visiting fellow, Department of C o m m u n i t y Medicine, M o u n t Sinai School of Medicine, N e w Y o r k 1980-1982 Director, Department of Occupational Medicine, Danish National Institute of Occupational Health 1982 Professor of Environmental Medicine, O d en se University 1983-2017 Consultant in Toxicology, Danish Health Authority 1994-2002 Adjunct Professor of Public Health (Environmental Health) and Neurology, Boston University School of Medicine 2003- Adjunct Professor of Environmental Health, Harvard School of Public Health Awards and honors 104 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN Prize essay in medicine, University of C o p e n h a g e n (1972) Fulbright senior research scholarship (1978) Ke yn ot e speaker, O d e n s e University anniversary (1983) Gitlitz M e m o r i a l Lecture, Association of Clinical Scientists, U S A (1985) Knight of the Dannebrog, awarded b y the Q u e e n of D e n m a r k (1990) T h e D a n n i n prize for medical research (1991) Fellow, A m e r i c a n Association for the A d v a n c e m e n t of Science (1994) Irish Congress Lecturer, Royal College of Physicians of Ireland (1996) Knight of the Dannebrog, First Degree, awarded b y the Q u e e n of D e n m a r k (2003) `Me r c u r y ma d n e s s a w a r d ' for excellence in science in the public interest from eight U S environmental organizations (2004) Emeritus Fellow, International U n i o n of Pure and Applied Chemistry, I U P A C (2009) Honorary Research Awar d, International Order of O d d Fellows (2010) Science Co mm u n i c a t i o n Award, University of Southern D e n m a r k (2012) Bernardino Ramazzini A w a r d (2015) Basic & Clinical Pharmacology & Toxicology Nordic A w a r d (2015) Margrethegaarden honorary prize (2016) John R. Goldsmith Awar d, International Society for Environmental Epidemiology (2016) Editorial boards A m e r i c a n Journal of Industrial Medicine (1987-2017) Applied Organometal Chemistry (1985-1991) Arbejdsmilj0 (Occupational Environment, in Danish, 1983-1990) Archives of Environmental Health (European Editor, 1986-1992) Archives of Toxicology (1987-) Biomarkers (1996-2001) Central European Journal of Occupational and Environmental Medicine (2015-) Critical Re v i e w s in Toxicology (1985-2012) Danish Medical Bulletin (1994-2003) Environmental Health (Editor-in-Chief, 2002-) Environmental Health Perspectives (2003-) Environmental Research (1981-1994 and 2014-, Associate Editor, 1995-2014) Industrial Health (2000-2005) International Journal of Hy gi en e and Environmental Health (2001-) International Journal of Occupational and Environmental Health (1994-2011) International Journal of Occupational Medicine & Environ Health (1991Journal of Clean Technology, Environmental Toxicology, and Occupational Medicine (1992 1998) Journal of Environmental Medicine (1998-1999) Naturens V e rd en (Natural Science, in Danish) (1987-1991) Ugeskrift for L ^ g e r (Danish Medical Journal, in Danish) (1991-2007) Scientific societies A m e r i c a n Association for the A d v a n c e m e n t of Science (Fellow, 1994) American Public Health Association Collegium Ramazzini (Fellow, 1987; M e m b e r of the Council, 2005-2013) 105 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN Danish M edical Association D anish Societies o f Clinical Chemistry, Epidem iology, Occupational and Environm ental M edicine, and Public Health Faroese Society of Science and Letters International Com m ission on Occupational Health International Society for Environm ental Epidem iology Research support as Principal Investigator since 2000 2000- 2006 NIEHS M ercury associated neurobehavioral deficit in children 2001- 2003 Nordic Arctic Research Program m e (NARP) Changing patterns of biom agnified pollutants in the northern m arine environm ent 2001- 2004 D anish M edical Research Council Exposure assessm ent for endocrine disruptors 2002- 2004 D anish M edical Research Council Environm ental epidemiology research 2003- 2004 European Commission Assessm ent of Neurobehavioral Endpoints and M arkers of Neurotoxicant Exposures (ANEM ONE) 2003-2005 D anish M edical Research Council Research in hormone related substances 2003-2006 NIEHS ES11687 Effects of perinatal disruptors in children 2003- 2007 EPA STAR RD-83075801-0 C hildren's vulnerability to environmental im m unotoxicant 2004- 2011 NIEHS ES12199 Epidem iology of im m unotoxicant exposure in children 2006-2011 NIEHS ES13692 Health effects of lifetime exposure to food contam inants 2006- 2012 NIEHS ES14460 Three-generation hum an study o f reproductive effects o f m arine food contam inants 2008-2012 D anish Council for Strategic Research Environm ental pollutant impact on antibody production against current and new childhood vaccines 2007- 2013 NIEHS ES009797 M ercury associated neurobehavioral deficit in children M ajor Current Funding as Principal Investigator 2011- 2017 NIEHS ES012199 Epidemiology of im m unotoxicant exposure in children 2012- 2018 NIEHS ES021993 and NSF OCE-1321612 Imm unotoxicity in Hum ans with Lifetim e Exposure to Ocean Pollutants 2013- 2018 NIEHS ES021477 Glucose M etabolism in Adults Prenatally Exposed to D iabetogenic Pollutants 2013-2018 NIEHS ES021372 Pollutant-related diabetes in the N urses' H ealth Study II 106 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 2014-2017 NIEHS ES023376 Gut M icrobiom e in Adults with Early Life Exposures to Environm ental Chemicals 2017-2022 NIEHS P42ES027706 Sources, Transport, Exposure and Effects of PFASs (STEEP) M ajor committees, boards and elective offices Danish: Danish M edical Association: M ember, Prevention Council (2011-2014) D anish M edical Research Council: Consultant on environm ental m edicine (1985-1990); M ember, Joint Research Council Committee on Environm ental Research (1986-1991); M em ber of DM RC (1992-1998) Danish Society of Community M edicine: Secretary (1977-1978) Danish Society of Industrial M edicine: Board M em ber (1974-1983) M inistry of Education: M ember, Comm ittee on Toxicology (1984-1986); M ember, Committee on Environmental Education (1986-1987) M inistry of the Environment: M ember, Council on Environm ental Chem icals (1983-1989); M ember, Environm ental Appeal Board (1986-2010); M ember, Environmental Research Council (1990-1992); M ember, Advisory Committee on Pesticide Research (1995-2004); M ember, Advisory Committee on Arctic Research (1996-2004) M inistry o f Health: num erous comm ittee appointments; Chair, Com m ittee on Risk Perception (2000-2001) M inistry of Labour: Consultant on Occupational Health, Council on Occupational Safety and H ealth (1983-1993); M ember, Occupational Health Council Research Committee (on behalf of the D anish M edical Research Council) (1984-1990 and 1999-2003) M inistry of Research: Chair, Committee on Research at the Faroe Islands (1995-1996); M ember, Com m ittee on Scientific Dishonesty (2004-2006); Chair, Com m ittee on Non-Ionizing Radiation (2004-2009) Odense U niversity (from 2000 U niversity o f Southern Denmark), elected offices: Chairman, Institute o f Com m unity H ealth (1982-1985; 1996-1999); M em ber o f Executive Committee, Institute of Community Health (from 2000 Institute of Public Health) (1986-1995; 2000 2005); M ember, Faculty Research Committee (1983-1985); M ember, Curriculum Committee (1984-1986); M ember, Faculty Council (1985-1993); Vice-Dean (1991-1993); M em ber, Scientific Integrity Com m ittee (2003-) United States and international: Academy of Finland: m em ber of panel evaluating the National Institute of Public H ealth (1995), site visit o f center o f excellence (2001) Agency for Toxic Substances and D isease Registry: W orkshop Rapporteur, Neurobehavioral Test Batteries for Use in Environm ental H ealth Field Studies (1992); M ember, Expert Panel of M ercury (1998) Association of Schools of Public Health in the European Region: Treasurer (1975-1977) BioM edCentral: M ember, Editors Advisory Group (2011-2013) Boston Environm ental Hazards Center: Consultant (1994-1999) Collegium Ramazzini: President, International Conference, The precautionary principle: Implications for research and prevention in environmental and occupational health (2002); M ember, Executive Council (2005-2013) 107 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN Com m ission of the European Communities: National Expert, W orking Party on Environm ental and Lifestyle-Related Diseases (1988-1990); ad hoc Consultant for evaluation of research applications; ad hoc Scientific A dvisor on R isk Assessm ent (2009-); M em ber, SCHER W orking group on Dental Amalgam (Hum an Health) (2012-2013) European Environm ent Agency: M ember, Scientific Committee (2012-2018) European Food Safety Authority: M ember, Panel on Contam inants in the Food Chain responsible for 85 opinions (2003-2009); M em ber o f W orking G roups on m ercury, polychlorinated biphenyls, cadmium, lead, and benchm ark dose Food Advisory Committee, U.S.FDA, M ethylmercury: invited expert (2002) IN M A (Infancia y M edio Am biente), Spain: M ember, Project Steering Com m ittee (2010-) Institut de Recherche Sante, Environnem ent et Travail, France: M ember, Board o f Advisers (2015-) International Agency for Research on Cancer: M em ber o f Task Group, M onographs on the Evaluation o f Carcinogenic Risks to Humans, Vol. 47 (1988), Vol. 49 (1989), as chairman, Vol. 58 (1993), and as Subgroup chair, Vol. 100C (2009) International Com m ission on Occupational Health: D anish Delegation Secretary (1982-90); M em ber, Scientific Com m ittee on the Toxicology o f M etals (1987-); M em ber o f the Board (1990-1996) International Program m e on Chem ical Safety: M em ber o f Task Group, Environm ental Health Criteria, Vol. 36 (1984) and 72 (1986) International Society for Environm ental Epidemiology: Councillor (1991-1994) International U nion o f Pure and Applied Chemistry: M ember, Subcom mittee on the Toxicology of Nickel (1979-1989); Titular M em ber (1985-1991) and Chairm an (1987-1991), Com m ission on Toxicology; Chairman, Subcom mittee on Risk A ssessm ent (1985-1989) K arolinska Institute (Stockholm, Sweden): M em ber of international evaluation panel on environmental m edicine (1993) M inistry for Scientific Policy (Belgium): Consultant on national research program on health hazards (1990 and 1994) National Institutes o f Health (USA): M em ber o f Special emphasis panels (2009-) N A TO Priority Area Panel on Environm ental Security: M em ber (1996-1997) N orw egian Research Council: ad hoc review er (2001-2008); Chairm an o f Environm ent and H ealth Review Group (2009-2010); m em ber of steering committee (2011-2015) Prenatal program m ing and Toxicity (PPTOX) conferences: Organizer/Chair/ Co-chair, Torshavn (2007), M iami (2009), Paris (2012), Boston (2014), Kita-Kyushu (2016) Society of Occupational and Environm ental Health: M ember, Governing Council (1990-1993) Swedish Council for W ork Life Research: M em ber, Priority Com m ittee on Chemical Health R isk s(1997-1998) U.N. Environm ent Program m e: M em ber, Global M ercury Assessm ent W orking Group (2002) U.S. Environm ental Protection Agency: M em ber, SAB/SAP Endocrine D isruptor Screening Program Subcommittee (1998-1999); M ember, Food Quality Protection Act (FQPA) Science Review Board (SRB)(1999-2003) W hite House Office o f Science and Technology Policy: Team leader and presenter, W orkshop on Scientific Issues Relevant to Assessm ent o f Health Effects from Exposure to M ethylmercury (1998) W orld H ealth Organization: Temporary A dviser or Consultant on several occasions, five tim es elected Rapporteur; M ember, European Advisory Com m ittee on Health Research (2011-) 108 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN EXHIBIT C LIST OF GRANDJEAN PUBLICATIONS FROM RECENT 10 YEARS Publications in international peer-reviewed journals 167. Grandjean P, B udtz-J0 rgensen E. Total im precision o f exposure biomarkers: Im plications for calculating exposure limits. Am J Industr M ed 2007; 50: 712-9. 168. Grandjean P. M ethylm ercury toxicity and functional program m ing. R eproduct Toxicol 2007; 23: 414-20. 169. Grandjean P, M urata K. Developm ental arsenic neurotoxicity in retrospect (editorial). E pidem iology 2007; 18: 25-6. 170. W erm uth L, B ech S, P etersen M S, Joensen P, W eihe P, G randjean P. H igh prevalence and incidence o f Parkinson's disease in the Faroe Islands. A cta N eurol Scand 2008; 118: 126-31. 171. M urata K, Grandjean P, Dakeishi M. N europhysiological evidence o f m ethylm ercury neurotoxicity. Am J Industr M ed 2007; 50: 765-71. 172. B udtz-J0 rgensen E, Grandjean P, W eihe P. Separation o f risks and benefits o f seafood intake. Environ H ealth Perspect 2007; 115: 323-7. 173. A ndersen HR, N ielsen F, N ielsen JB, K jaerstad M B, Baelum J, G randjean P. X enooestrogenic activity in serum as m arker o f occupational pesticide exposure. Occup Environ M ed 2007; 64: 708-714. 174. A ndersen HR, Schm idt IM, Grandjean P, Jensen TK, B udtz-J0 rgensen E, K jaerstad M B, Baelum J, N ielsen JB, Skakkebaek NE, M ain KM . Im paired reproductive developm ent in sons of wom en occupationally exposed to pesticides during pregnancy. Environ Health Perspect 2008; 116: 566-72. 175. Petersen M S, H alling J, D am kier P, N ielsen F, G randjean P, W eihe P, B rasen K. Polychlorinated biphenyl (PCB) induction o f the CYP3A4 enzyme activity in Healthy Faroese adults. Toxicol Appl Pharmacol 2007; 224: 202-6.176. Choi AL, Budtz-J0 rgensen E, J0 rgensen PJ, Steuerwald U, Debes F, W eihe P, Grandjean P. Selenium as a potential protective factor against m ercury developm ental neurotoxicity. Environ Res 2008; 107: 45-52. 177. Grandjean P. Seven deadly sins o f environm ental epidem iology and the virtues of precaution. E pidem iology 2008; 19: 158-62. 178. Grandjean P. Late insights into early origins o f disease. Basic Clin Pharm acol Toxicol 2008; 102: 94-9. 179. Petersen M S, W eihe P, Choi A, Grandjean P. Increased prenatal exposure to m ethylm ercury does not affect the risk of Parkinson's disease. N eurotoxicology 2008; 29: 591-5. 180. P etersen M S, H alling J, B ech S, W erm uth L, W eihe P, N ielsen F J0 rgensen PJ, B udtzJ0 rgensen E, Grandjean P. Im pact of dietary exposure to food contaminants on the risk of Parkinson's disease. N eurotoxicology 2008; 29: 584-90. 181. H alling J, Petersen M S, B rosen K, W eihe P, G randjean P. G enetic predisposition to Parkinson's disease: CYP2D6 and H FE in the Faroe Islands. Pharm acogenet Genom ics 2008; 18: 209-12. 182. Choi A, C ordier S, W eihe P, G randjean P. N eg ativ e confounding in the evaluation o f toxicity: The case o f m ethylm ercury in fish and seafood. Crit Rev Toxicol 2008; 38: 877-93. 183. Grandjean P, O zonoff D. Environm ental Health: the first five years. Environ H ealth 2007; 6: 27. 109 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 184. Grandjean P, Choi A. The delayed appearance o f a m ercurial warning. Epidem iology 2008; 19: 10-1. 185. P ouzaud F, Ibbou A, B lanchem anche S, G randjean P, K re m p f M , P h ilippe H -J, V erger P. Use of advanced cluster analysis to characterize seafood consumption patterns and m ethylm ercury exposures among pregnant women. J Exp Anal Environ Epidemiol 2010; 20: 54 68. 186. Grandjean P, Perez M. Developm ental neurotoxicity: Im plications o f m ethylm ercury research. International Journal o f Environm ent and H ealth 2008; 2: 417-28. 187. Choi AL, G randjean P. M ethylm ercury exposure and health effects in hum ans. Environ C hem 2008; 5: 112-20. 188. W eihe P, K ato K, Calafat AM, N ielsen F, W anigatunga AA, N eedham LL, Grandjean P. Serum concentrations of polyfluoroalkyl compounds in Faroese w hale m eat consumers. Environ Sci Technol 2008; 42: 6291-5. 189. Grandjean P, B udtz-J0 rgensen E, B arr DB, N eedham LL, W eihe P, H einzow B. Elim ination half-lives o f polychlorinated biphenyl congeners in children. Environ Sci Technol 2008; 42: 6991-6. 190. Coccini T, M anzo L, Debes F, W eihe P, Grandjean P. A pplication o f lym phocyte muscarinic receptors and platelet m onoamine oxidase-B as biomarkers of CNS function in a F aroese children cohort prenatally exposed to m ethylm ercury and PC B s. B iom arkers 2009; 14: 67-76. 191. B udtz-J0 rgensen E, D ebes F, W eihe P, Grandjean P. Structural equation m odels for m eta analysis in environmental risk assessment. Environm etrics 2010; 21: 510-27. 192. Choi AL, W eihe P, B udtz-J0 rgensen E, J0 rgensen PJ, Salonen JT, Tuom ainen T-P, M urata K, N ielsen HP, Petersen M S, A skham J, G randjean P. M ethylm ercury exposure and adverse cardiovascular effects in Faroese w halingm en. Environ H ealth Perspect 2009; 117: 369-72. 193. B j0 rling-Poulsen M , Andersen HR, G randjean P. Potential developm ental neurotoxicity of pesticides used in Europe. E nviron H ealth 2008; 7: 50. 194. Julvez J, Grandjean P. N eurodevelopm ental toxicity risks due to occupational exposure to industrial chemicals during pregnancy. Industr H ealth 2009; 47: 459-68. 195. Grandjean P, B udtz-J0 rgensen E. An ignored risk factor in toxicology: The total im precision o f exposure assessment. Pure Appl Chem 2010; 82: 383-91. 196. K irkegaard M , Sonne C, D ietz R, Letcher RJ, Jensen AL, H ansen SS, Jenssen BM, Grandjean P. Alterations in thyroid hormone status in Greenland sledge dogs exposed to w hale blubber contam inated w ith organohalogen compounds. Environ Qual Saf 2011; 74: 157-63. 197. B lair A, Saracci R, Vineis P, Cocco P, Forastiere F, Grandjean P, K ogevinas M, Kriebel D, M cM ichael A, Pearce N, Porta M , Sam et J, Sandler DP, Costantini RS, V ainio H. Epidem iology, public health and the rhetoric o f false positives. Environ H ealth Perspect 2009; 117: 1809-13. 198. Schlezinger JJ, Bernard PL, H aas A, Grandjean P, W eihe P, Sherr DH. D irect assessm ent of cum ulative aryl hydrocarbon receptor agonist activity in sera from experim entally exposed mice and environm entally exposed hum ans. Environ H ealth Perspect 2010; 118: 693-8. 199. W hite RF, Palum bo CL, Yugelun-Todd DA, H eaton KJ, W eihe P, D ebes F, Grandjean P. Functional M R I approach to developmental m ethylm ercury and polychlorinated biphenyl neurotoxicity. N eurotoxicology 2011; 32: 975-80. 200. Lincoln RA, Vorhees DJ, Chesney EJ, Shine JP, Grandjean P, Senn DB. Fish consumption and m ercury exposure among Louisiana recreational anglers. Environ H ealth Perspect 2011; 119: 245-51. 110 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 201. Yorifuji T, Tsuda T, G randjean P. Unusual cancer excess after neonatal arsenic exposure from contam inated m ilk powder. J N atl C ancer Inst 2010; 102: 360-1. 202. H arari R, Julvez J, M urata K, B arr D, B ellinger DC, D ebes F, G randjean P. Neurobehavioral deficits and increased blood pressure in school-age children prenatally exposed to pesticides. Environ H ealth Perspect 2010; 118: 890-6. 203. Grandjean P, Satoh H, M urata K, Eto K. Adverse effects of m ethylm ercury: Environm ental health research im plications. Environ H ealth Perspect 2010; 118: 1137-45. 204. M ahaffey KR, Sunderland EM, Chan HM, Choi AL, Grandjean P, M arien K, Oken E, Sakamoto M, Schoeny R, W eihe P, Yan C-H, Y asutake A. B alancing the benefits o f n-3 polyunsaturated fatty acids and the risks o f m ethylm ercury exposure from fish consumption. N utrit Rev 2011; 69: 493-508. 205. Julvez J, Debes F, W eihe P, Choi A, Grandjean P. Sensitivity o f continuous perform ance test (CPT) to m ercury exposure at age 14 years. N eurotoxicol Teratol 2010; 32: 627-32. 206. D algard C, Petersen M S, Schm edes A V , B randslund I, W eihe P, G randjean P. H igh latitude and m arine diet: Vitam in D status in elderly Faroese. B r J N utr 2010; 104: 914-8. 207. Heilm ann C, Budtz-J0 rgensen E, Nielsen F, Heinzow B, W eihe P, Grandjean P. Serum concentrations of antibodies against vaccine toxoids in children exposed perinatally to im m unotoxicants. Environ H ealth Perspect 2010; 118: 1434-8. 208. Grandjean P, Poulsen LK, Heilm ann C, Steuerwald U, W eihe P. Allergy and sensitization during childhood associated with prenatal and lactational exposure to m arine pollutants. Environ H ealth Perspect 2010; 118: 1429-33. 209. Grandjean P, Henriksen JE, Choi AL, Petersen MS, Dalgard C, Nielsen F, W eihe P. M arine food pollutants as a risk factor for hypoinsulinem ia and type 2 diabetes. Epidem iology 2011; 22: 410-7. 210. Yorifuji T, D ebes F, W eihe P, G randjean P. Prenatal exposure to lead and cognitive deficit in 7- and 14-year-old children in the presence of concom itant exposure to similar m olar concentration o f m ethylm ercury. Neurotoxicol Teratol 2011; 33: 205-11. 211. Grandjean P. Even low-dose lead exposure is hazardous. The Lancet 2010; 375: 855-6. 212. Spulber S, R antam aki T, N ik k ila O, C astren E, W eihe P, G randjean P, C eccatelli S. E ffects of maternal smoking and exposure to methylm ercury on Brain-Derived Neurotrophic Factor (BDN F) concentrations in cord serum. Toxicol Sci 2010; 117: 263-9. 213. M ozaffarian D, Shi P, M orris JS, Spiegelman D, Grandjean P, Siscovick, W illett WC, Rim m EB. M ercury exposure and risk o f cardiovascular disease in tw o U.S. cohorts. N Engl J M ed 2011; 364: 1116-25. 214. O zonoff DM , Grandjean P. M ilestones and im pact factors (editorial). Environ Health 2010; 9: 35. 215. N eedham LL, G randjean P, H einzow B, J0 rgensen PJ, N ielsen F, Patterson D G Jr, Sjodin A, Turner WE, W eihe P. Partition of environmental chemicals betw een maternal and fetal blood and tissues. Environ Sci Technol 2011; 45: 1121-6. 216. Y orifuji T, G randjean P, T suda T, K ash im a S, D oi H. C ancer excess after arsenic exposure from contam inated m ilk pow der. E nviron H ealth P rev M ed 2011; 16: 164-70. 217. Grandjean P, Herz K. M ethylm ercury and brain development: Im precision and underestim ation o f developm ental neurotoxicity in humans. M t Sinai J M ed 2011: 78: 107-18. 218. Pichery C, Bellanger M, Zm irou-N avier D, Glorennec P, Hartem ann P, G randjean P. Childhood lead exposure in France: benefit estim ation and partial cost-benefit analysis o f lead hazard control. E n v iro n H ealth 2011; 10: 44.1 111 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 219. W ohlfahrt-V eje C, M ain KM , Schmidt IM, Boas M, Jensen TK, Grandjean P, Skakkeb^k NE, Andersen HR. Low er birth w eight and increased body fat at school age in children prenatally exposed to m odern pesticides: A prospective study. E nviron H ealth 2011; 10: 79. 220. W ohlfahrt-V eje C, Andersen HR, Schmidt IM, Aksglaede L, S0 rensen K, Juul A, Jensen TK, Grandjean P, Skakkeb^k NE, M ain KM . Early Breast D evelopm ent in Girls after Prenatal Exposure to N on-Persistent Pesticides. Int J Androl 2012; 35: 273-82. 221. D algard C, Petersen M S, W eihe P, Grandjean P. Vitam in D status in relation to type 2 diabetes development. Diabetes Care 2011; 34: 1284-8. 222. Julvez J, Debes F, W eihe P, Choi AL, Grandjean P. Thyroid dysfunction as a m ediator of organochlorine neurotoxicity in preschool children. Environ Health Perspect 2011; 119:1429-35. 223. Audouze K, Grandjean P. A pplication o f computational systems biology to explore environm ental toxicity hazards. Environ H ealth Perspect 2011; 119: 1754-9. 224. Grandjean P, Andersen EW , Budtz-J0 rgensen E, Nielsen F, M 0 lbak K, W eihe P, Heilm ann C. D ecreased serum vaccine antibody concentrations in children exposed to perfluorinated compounds. JAM A 2012; 307: 391-7. 225. Grandjean P, Eriksen M L, Ellegaard O, W allin JA. The M atthew effect in environm ental science publication: A bibliom etric analysis of chemical substances in journal articles. Environ H ealth 2011; 10: 96. 226. V estergaard S, N ielsen F, A ndersson A M , H j0 llund N H , G randjean P, A ndersen H R, Jensen TK. Association betw een perfluorinated compounds and time to pregnancy in a prospective cohort of D anish couples attem pting to conceive. H um an Reproduct 2012; 27: 873 80. 227. W ohlfahrt-V eje C, Andersen HR, Jensen TK, Grandjean P, Skakkebaek NE, M ain KM. Smaller genitals at school age in boys w hose m others were exposed to non-persistent pesticides in early pregnancy. Int J Androl 2012; 35: 265-72. 228. Grandjean P, W eihe P, N ielsen F, H einzow B, Debes F, Budtz-J0 rgensen E. Neurobehavioral deficits at age 7 years associated w ith prenatal exposure to toxicants from maternal seafood diet. Neurotoxicol Teratol 2012; 34: 466-72. 229. Grandjean P, G ranlund C, K j^ r IM, Jensen TK, S0 rensen N, Andersson AM , Juul A, Skakkeb^k NE, Budtz-J0 rgensen E, W eihe P. Reproductive hormone profile and pubertal developm ent in 14-year-old boys prenatally exposed to polychlorinated biphenyls. Reprod Toxicol 2012; 34: 498-503. 230. K aragas M R , C hoi A L, O ken E, H o rv at M , Schoeny R , K am ai E, G randjean P, K orrick S. Evidence on the hum an health effects of low level m ethylm ercury exposure. Environ Health Perspect 2012; 120: 799-806. 231. G randjean P, O z o n o ff D. P o rtrait o f the jo u rn al as a young adult. E nviron H ealth. 2012; 11: 30. 232. Budtz-J0 rgensen E, Bellinger D, Lanphear B, Grandjean P, International Pooled Lead Study Investigators. An international pooled analysis for obtaining a benchm ark dose for environm ental lead exposure in children. Risk Anal 2013; 33: 450-61. 233. F ^ rch K, H 0 jlund K, Vind BF, Vaag A, D algard C, N ielsen F, Grandjean P. Increased serum concentrations of persistent organic pollutants among prediabetic individuals: potential role o f altered substrate oxidation patterns. J Clin Endocrinol M etab 2012; 97: E1705-13. 234. Yorifuji T, M urata K, Bjerve K, Choi AL, W eihe P, G randjean P. Visual evoked potentials in children prenatally exposed to m ethylm ercury. N eurotoxicology 2013; 37: 15-8. 235. P ichery C, B ellanger M , Z m irou-N avier D, F rery N , C ordier S, R oue-L eG all A, H artem ann12 112 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN P, Grandjean P. Econom ic evaluation o f health consequences o f prenatal m ethylm ercury exposure in France. E nviron H ealth 2012; 11: 53. 236. Andersen HR, W ohlfahrt-Veje C, Dalgard C, Christiansen L, M ain KM , Christine N ellem ann C, M urata K, Jensen TK, Skakkeb^k NE, Grandjean P. Paraoxonase 1 polymorphism and prenatal pesticide exposure associated with adverse cardiovascular risk profiles at school age. PLoS ONE 2012; 7(5): e36830. 237. Choi AL, Sun G, Zhang Y, Grandjean P. Developmental fluoride neurotoxicity: A system atic review and m eta-analysis. Environ H ealth Perspect 2012; 120: 1362-8. 238. M ozaffarian D, Shi P, M orris JS, Grandjean P, Siscovick D, Spiegelman D, W illett W, R im m E, C urhan G, F orm an J. M ercury exposure and risk o f hypertension in U S m en and wom en in tw o prospective cohorts. Hypertension 2012; 60: 645-52. 239. W u H, Bertrand KA, Choi AL, Hu FB, Laden F, Grandjean P, Sun Q. Plasm a levels of persistent organic pollutants and risk of type 2 diabetes: a prospective analysis in the N urses' H ealth Study and m eta-analysis. Environ H ealth Perspect 2013; 121: 153-61. 240. Barouki B, G luckm an PD, G randjean P, H anson M, H eindel JJ. Developmental origins of non-com m unicable diseases and dysfunctions: Im plications for research and public health. E nvironm ental H ealth 2012: 11: 42. 241. Julvez J, D avey-S m ith G, G olding J, R ing S, St. P ourcain B, G onzalez JR, G randjean P. Prenatal m ethylm ercury exposure and genetic predisposition to cognitive deficit at age 8 years. Epidemiology 2013; 24: 643-50. 242. Balbus JM, Barouki R, Birnbaum LS, Etzel RA, Gluckman PD, Grandjean P, Hancock C, H anson M A, Heindel JJ, H offm an K, Jensen GK, K eeling A, N eira M , Rabadan-D iehl C, Ralston J, Tang KC. Early-life prevention o f non-com m unicable diseases (Com ment). Lancet 2013; 381: 3-4. 243. D ietz R, Sonne C, Basu N, Braune B, O'H ara T, Letcher RJ, Scheuham m er T, A ndersen M, Andreasen C, Andriashek D, Asmund G, Aubail A, Baag0 e H, Born EW, Chan HM , Derocher AE, Grandjean P, K nott K, Kirkegaard M, Krey A, Lunn N, M essier F, Obbard M, Olsen MT, O stertag S, P eacock E, R enzoni A, R ig et FF, Skaare JU, Stern G, Stirling I, T aylor M , W iig O, W ilson S, A ars J. W h at are the toxicological effects o f m ercury in A rctic biota? Sci Total Environ 2013; 443: 775-790. 244. Bellanger M, Pichery C, Aerts D, Berglund M, Castano A, Cejchanova M, Crettaz P, Davidson F, Esteban M, Fischer ME, Gurzau AE, Halzlova K, Katsonouri A, Knudsen LE, K olossa-Gehring M, Koppen G, Ligocka D, M iklavcic A, Reis MF, Rudnai P, Tratnik JS, W eihe P, Budtz-J0 rgensen E, Grandjean P. Econom ic benefits o f m ethylm ercury exposure control in Europe: M onetary value o f neurotoxicity prevention. E nviron H ealth 2013; 12: 3. 245. H alling J, Petersen M S, J0 rgensen N, Jensen TK, G randjean P, W eihe P. Sem en quality and reproductive horm ones in Faroese m en - a cross-sectional population-based study o f 481 men. B M J O pen 2013; 3: e001946. 246. Grandjean P, Budtz-J0 rgensen E. Im m unotoxicity o f perfluorinated alkylates: Calculation o f b enchm ark doses b ased on serum concentrations in children. E nviron H ealth 2013; 12: 35. 247. Choi AL, M ogensen UB, Bjerve K, W eihe P, Grandjean P, Budtz-J0 rgensen E. Negative confounding by essential fatty acids in m ethylm ercury neurotoxicity associations. Neurotoxicol Teratol 2014; 42: 85-92. 248. M ozaffarian D, Shi P, M orris JS, Grandjean P, Siscovick DS, Spiegelman D, Hu FB. M ethylmercury exposure and incident diabetes mellitus in US men and wom en in two prospective cohorts. D iabetes C are 2013; 36: 3578-84.13 113 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 249. A udouze K, B ru n ak S, G randjean P. C om putational approach to chem ical etiologies o f diabetes. Sci C om m 2013; 3: 2712. 250. Fonseca M F, Hacon SS, G randjean P, Choi AL, Bastos W R. Iron status as a covariate in m ethylm ercury-associated neurotoxicity risk. Chem osphere 2014; 100: 89-96. 251. Grandjean P, Clapp R. Changing interpretation of hum an health risks from perfluorinated compounds. Publ Health Rep 2014:129; 482-5. 252. Grandjean P, Landrigan PJ. Neurobehavioural effects of developm ental toxicity. Lancet N eurol 2014; 13: 330-8. 253. Kim BM, Choi A, H a EH, Pedersen L, Nielsen F, W eihe P, Hong YC, Budtz-J0 rgensen E, Grandjean P. Effect o f hem oglobin and selenium on partition o f m ercury betw een maternal and cord blood. Environ Res 2014; 132: 407-12. 254. Grandjean P, O zonoff D. Transparency and translation o f science in a modern world. E nviron H ealth 2013; 12: 70. 255. Tang-Peronard JL, H eitm ann BL, Andersen HR, Steuerwald U, Grandjean P, W eihe P, Jensen TK. A ssociation betw een prenatal polychlorinated biphenyl exposure and obesity developm ent at ages 5 and 7 y: a prospective cohort study o f 656 children from the Faroe Islands. Am J Clin N utrit 2014; 99: 5-13 256. Timm erm ann CAG, Rossing LI, Grantved A, Ried-Larsen M, Dalgard C, Andersen LB, Grandjean P, N ielsen F, Svendsen KD, Scheike T, Jensen TK. Adiposity and glycem ic control in children exposed to perfluorinated compounds. J Clin Endocrinol M etab 2014; 99: E608-14. 257. Julvez J, Grandjean P. Genetic susceptibility to m ethylm ercury developm ental neurotoxicity m atters. Front G enet 2013; 4: 278. 258. V esterholm Jensen D, Christensen JH, V irtanen HE, S kakkeb^k N E, M ain KM , Toppari J, Veje CV, Andersson AM, N ielsen F, Grandjean P, Jensen TK. N o association between exposure to perfluorinated compounds and congenital cryptorchidism : a nested case-control study among 215 boys from D enm ark and Finland. Reproduction 2014; 147: 411-7. 259. Li M, Sherman LS, Blum JD, Grandjean P, M ikkelsen B, W eihe P, Sunderland EM , Shine JP. A ssessing sources o f hum an m ethylm ercury exposure using stable m ercury isotopes. Environ Sci Technol 2014; 48: 8800-6. 260. Grandjean P, Herz KT. Trace elem ents as paradigms of developmental neurotoxicants. J Trace Elem M ed Biol 2015; 31: 130-4. 261. Grandjean P, W eihe P, Debes F, Choi AL, Budtz-J0 rgensen E. N eurotoxicity from prenatal and postnatal exposure to m ethylmercury. Neurotoxicol Teratol 2014; 43: 39-44. 262. Grandjean P, Clapp R. Perfluorinated alkyl substances: em ergence o f insights into health risks. N ew Solutions 2015; 25: 147-63. 263. Osuna CE, Grandjean P, W eihe P, El-Fawal HAN. Autoantibodies associated with prenatal and childhood exposure to environm ental chem icals in Faroese children. Toxicol Sci 2014; 142: 158-66. 264. M ogensen UB, Grandjean P, Heilm ann C, Nielsen F, W eihe P, Budtz-J0 rgensen E. Structural equation m odeling of im m unotoxicity associated with exposure to perfluorinated com pounds. E nviron H ealth 2015; 14: 47. 265. Andersen HR, Debes F, W ohlfahrt-V eje C, M urata K, Grandjean P. Occupational pesticide exposure in early pregnancy and neurobehavioral function in children at school age. N eurotoxicol Teratol 2015; 47: 1-9. 266. K vist L, G iw ercm an A, W eihe P, Jensen TK, G randjean P, H alling J, Petersen M S, G iw ercm an YL. E xposure to persistent organic pollutants and sperm sex chrom osom e ratio in14 114 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN m en from the Faroe Islands. Environ Int 2014; 73: 359-64. 267. Jensen TK, Timm erm ann AG, Rossing LI, Ried-Larsen M, Grantved A, Andersen LB, Dalgaard C, Hansen OH, Scheike T, N ielsen F, Grandjean P. Polychlorinated biphenyl exposure and glucose m etabolism in D anish children at age 9 years. J Clin Endocrinol M etab 2014; 99: E2643-51. 268. Choi AL, Zhang Y, Sun G, Bellinger D, W ang K, Yang XJ, Li JS, Zheng Q, Fu Y, Grandjean P. Association of cognitive deficits with prenatal exposure to fluoride in Chinese children: a pilot study. Neurotoxicol Teratol 2015; 47: 96-101. 269. M 0 rck TA, N ielsen F, Nielsen JKS, Siersma V, Grandjean P, K nudsen LE. PFAS concentrations in plasm a samples from D anish school children and their mothers. Chem osphere 2015; 129: 203-9. 270. K ioum ourtzoglou M A, Roberts AL, N ielsen F, Shelley Tw oroger SS, Grandjean P, W eisskopf M G. W ithin-person reproducibility o f red blood cell m ercury over a 10- to 15-year period am ong w om en in the N urses' H ealth Study II. J Exp Sci Environ Epidem iol 2016; 26: 219-23. 271. W u H, Grandjean P, H u FB, Sun Q. Consum ption o f white rice and brow n rice and urinary inorganic arsenic concentration. Epidem iology 2015: 26: e65-7. 272. Jensen TK, Andersen LB, Kyhl HB, N ielsen F, Christensen HT, Grandjean P. Association betw een perfluorinated compounds and m iscarriage in a case-control study o f D anish pregnant w om en. P L oS O ne 2015; 10: e0123496. 273. Trasande L, Z oeller RT, H ass U, K ortenkam p A, G randjean P, M yers JP, D iG angi J, B ellanger M , H auser R, L egler J, Skakkebaek N , H eindel JJ. E stim ating burden and disease costs of exposure to endocrine disrupting chemicals in the European Union. J Clin Endocrinol M etab 2015; 100: 1245-55. 274. Bellanger M, D em eneix B, Grandjean P, Zoeller RT, Trasande L. N eurobehavioral deficits, diseases and associated costs o f exposure to endocrine disrupting chemicals in the European Union. J Clin Endocrinol M etab 2015; 100: 1256-66. 275. T ang-P eronard JL, H eitm an n B L , Jensen TK , V inggaard A M , M adsbad S, Steuerw ald U, Grandjean P, W eihe P, Nielsen F, Andersen HR. Prenatal exposure to persistent organic pollutants is associated w ith increased insulin levels in 5-year-old girls. Environ Res 2015; 142: 407-13. 276. Timm erm ann CAG, Osuna CE, Steuerwald U, W eihe P, Poulsen LK, Grandjean P. Asthma and allergy in children w ith and w ithout prior m easles mumps, and rubella vaccination. Pediatr Allergy Immunol 2015; 26: 742-9. 277. T0 ttenborg SS, Choi AL, Bjerve KS, W eihe P, G randjean P. Effect o f seafood m ediated PCB on desaturase activity and PU FA profile in Faroese septuagenarians. Environ Res 2015; 140: 699-703. 278. Petersen M S, H alling J, W eihe P, Jensen TK, G randjean P, N ielsen F, J0 rgensen N. Spermatogenic capacity in fertile men with elevated exposure to polychlorinated biphenyls. Environ Res 2015; 138: 345-51. 279. Grandjean P. Toxicology research for precautionary decision-m aking and the role of Hum an & Experim ental Toxicology. Hum Exp Toxicol 2015; 34: 1231-7. 280. Pearce NE, B lair A, Vineis P, Ahrens W, Andersen A, Anto JM, Arm strong BK, Baccarelli AA, Beland FA, Berrington A, Bertazzi PA, Birnbaum LS, Brownson RC, Bucher JR, Cantor KP, Cardis E, Cherrie JW, Christiani DC, Cocco P, Coggon D, Com ba P, Dem ers PA, Dem ent JM , D ouw es J, E isen E A , E ngel L S, F enske R A , F lem ing LE, F letch er T, F ontham E, F orastiere15 115 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN F, Frentzel-Beym e R, Fritschi L, Gerin M, Goldberg M, Grandjean P, Grimsrud TK, Gustavsson P, H aines A, H artge P, H ansen J, H auptm ann M , H eederik D, Hem m inki K, H em on D, HertzPicciotto I, H oppin JA, H u ff J, Jarvholm B, K ang D, K aragas M R, K jaerheim K, K juus H, Kogevinas M, Kriebel D, Kristensen P, Krom hout H, Laden F, Lebailly P, LeM asters G, Lubin JH , L ynch CF, L ynge E, `t M annetje A, M cM ichael A J, M cL aughlin JR, M arrett L, M artuzzi M , M erchant JA, M erler E, M erletti F, M iller A, M irer FE, M onson R, N ordby KC, Olshan AF, Parent M E, Perera FP, Perry M J, Pesatori AC, Pirastu R, Porta M, Pukkala E, Rice C, R ichardson DB, R itter L, R itz B, R onckers CM , R ushton L, Rusiecki JA, R usyn I, Sam et JM, Sandler D P, de Sanjose S, S chernham m er E, Seniori C ostantini A, Seixas N , Shy C, Siem iatycki J, Silvermann DT, Simonato L, Smith AH, Smith M T, Spinelli JJ, Spitz M R, Stallones L, Stayner LT, Steenland K, Stenzel M, Stewart BW , Stewart PA, Symanski E, Terracini B, Tolbert PE, V ainio H, V ena J, V erm eulen R, Victora CG, W ard EM , W einberg CR, W eisenburger D, W esseling C, W eiderpass E, Zahm SH. IARC monographs: 40 years o f evaluating carcinogenic hazards to hum ans. Environ H ealth Perspect 2015; 123: 507-14. 281. Zong G, Grandjean P, W u H, Sun Q. Circulating persistent organic pollutants and body fat distribution, evidence from NHANES 1999-2004. Obesity 2015; 23: 1903-10. 282. Debes F, W eihe P, Grandjean P. Cognitive deficits at age 22 years associated with prenatal exposure to m ethylm ercury. Cortex 2016; 74: 358-69. 283. M ogensen UB, Grandjean P, Nielsen F, W eihe P, Budtz-J0 rgensen E. Breastfeeding as an exposure pathw ay for perfluorinated alkylates. Environ Sci Technol 2015; 49: 10466-73. 284. K ielsen K, Sham in Z, R yder LP, N ielsen F, G randjean P, B udtz-J0 rgensen E, H eilm ann C. Antibody response to booster vaccination w ith tetanus and diphtheria in adults exposed to perfluorinated alkylates. J Im m unotoxicol 2016; 13: 270-3. 285. G randjean P, B arouki R , B ellinger D, C asteleyn L, C hadw ick LH , C ordier S, E tzel RA , Gray KA, H a EH, Junien C, Karagas M, K aw am oto T, Lawrence BP, Perera F, Prins G, Puga A, Rosenfeld CS, Sherr D, Sly P, Suk W, Sun Q, Toppari J, van den Hazel P, W alker CL, Heindel JJ. Life-long im plications o f developm ental exposure to environm ental stressors: N ew perspectives. Endocrinology 2015; 156: 3408-15. 286. Heindel JJ, Balbus J, Birnbaum L, Brune-D risse M L, Grandjean P, Gray K, Landrigan PJ, Sly PD, Suk W , Cory-Slechta D, Thom pson C, H anson M. Developm ental origins o f health and disease: integrating environm ental influences. Endocrinology 2015; 156: 3416-21. 287. Egsm ose EL, Brauner EV, Frederiksen M, M 0 rck TA, Siersma VD, Hansen PW , Nielsen F, Grandjean P, Knudsen LE. Associations betw een plasm a concentrations of PCB 28 and possible indoor exposure sources in D anish school children and m others. Environ Intern 2016; 87: 13-9. 288. Perry M J, Y oung H A , G randjean P, H alling J, Petersen M S, Sheena EM , Parisa K, W eihe P. Sperm aneuploidy in men with elevated lifetime exposure to dichlorodiphenyldichloroethylene (DDE) and polychlorinated biphenyl (PCB) pollutants. Environ H ealth Perspect 2016; 124: 951-6. 289. Julvez J, Paus T, B ellinger D, Eskenazi B, Tiem eier H, Pearce N, Ritz B, W hite T, R am chandani P, G ispert JD , D esrivieres S, B ro u w er R, B o u ch er O, A lem any S, L opez-V icente M , Suades-G onzalez E, Forns J, G randjean P, Sunyer J. E nvironm ent and B rain D evelopm ent: Challenges in the Global Context. N euroepidem iology 2016; 46: 79-82. 290. Yorifuji T, K ato T, Ohta H, Bellinger DC, M atsuoka K, Grandjean P. Neurological and neuropsychological functions in adults with a history of developmental arsenic poisoning from contam inated m ilk powder. Neurotoxicol Teratol 2016; 53: 75-80. 116 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 291. Sunderland EM , Driscoll CT Jr, H am m itt JK, G randjean P, Evans JS, Blum JD, Chen CY, E vers D C , Jaffe D A , M ason RP, G oho S, Jacobs W . B enefits o f regulating hazardous air pollutants from coal and oil-fired utilities in the U nited States. Environ Sci Technol 2016; 50:2117-20. 292. Grandjean P. Learning from Bernardino Ram azzini, a tribute to the M agister from Carpi and to the Fellows of the Collegium Ramazzini. Eur J Oncol 2016: 21: 51-60. 293. Vandenberg LN, Agerstrand M, Beronius A, Beausoleil C, Bergm an A, Bero LA, Bornehag CG, B oyer CS, C ooper GS, C otgreave I, Gee D, G randjean P, G uyton KZ, H ass U, H eindel JJ, Jobling S, K idd K A , K ortenkam p A, M acleod M R , M artin OV, N o rin d er U, S cheringer M , Thayer KA, Toppari J, W haley P, W oodruff TJ, R uden C. A proposed fram ew ork for the systematic review and integrated assessm ent (SYRINA) of endocrine disrupting chemicals. E nviron H ealth 2016; 15: 74. 294. Trasande L, Zoeller RT, H ass U, Kortenkam p A, Grandjean P, M yers JP, D iGangi J, H unt PM , R udel R, Sathyanarayana S, B ellan g er M , H au ser R, L eg ler J, S kakkebaek N E , H eindel JJ. Burden o f disease and costs o f exposure to endocrine disrupting chemicals in the European Union: an updated analysis. A ndrology 2016; 4: 565-72. 295. Dalgard C, Petersen M S, Steuerwald U, W eihe P, Grandjean P. Umbilical cord serum 25hydroxyvitam in D concentrations and relation to birthweight, head circum ference and infant length at age 14 days. P aediatr Perinat Epidem iol 2016; 30: 238-45. 296. Grandjean P. Paracelsus Revisited: The dose concept in a complex world. Basic Clin Pharm acol Toxicol 2016; 119: 126-32. 297. T inggaard J, W ohlfahrt-V eje C, H usby S, C hristiansen L, S kakkebaek N E , Jensen TK, Grandjean P, M ain KM, Andersen HR. Prenatal pesticide exposure and PON1 genotype associated with adolescent body fat distribution evaluated by dual X-ray absorptiometry (DXA). A ndrology 2016; 4: 735-44. 298. Zong G, Grandjean P, W ang X, Sun Q. Lactation history, serum concentrations of persistent organic pollutants, and m aternal risk o f diabetes. Environ Res 2016; 150: 282-8. 299. Grandjean P, Heilm ann C, W eihe P, Nielsen F, M ogensen UB, Budtz-J0 rgensen E. Serum vaccine antibody concentrations in adolescents exposed to perfluorinated compounds. Environ Health Perspect (in press). 300. Hu XC, Andrews D, Lindstrom AB, Bruton TA, Schaider LA, Grandjean P, Lohm ann R, Carignan CC, Blum A, Balan SA, Higgins CP, Sunderland EM. Detection of poly- and perfluoroalkyl substances (PFASs) in U.S. drinking w ater linked to industrial sites, m ilitary fire training areas and w astew ater treatm en t plants. E nviron Sci T echnol L ett 2016 3: 344-350. 301. Timmermann CAG, Budtz-J0 rgensen E, Petersen M S, W eihe P, Steuerwald U, Nielsen F, Jensen TK, Grandjean P. Shorter duration of breastfeeding at elevated exposures to perfluoroalkyl substances. Reproduct Toxicol 2017; 68: 164-170. 302. Lind DV, Priskorn L, Lassen TH, Nielsen F, Kyhl HB, Kristensen DM , Christesen HT, J0 rgensen JS, Grandjean P, Jensen TK. Prenatal exposure to perfluoroalkyl substances and anogenital distance at 3 m onths o f age as m arker o f endocrine disruption. Reproduct Toxicol 2017; 68: 200-206. 303. O ulhote Y, Sham im Z, K ielsen K, W eihe P, G randjean P, R yder LP, H eilm ann C. C hildren's w hite blood cell counts in relation to developm ental exposures to m ethylm ercury and persistent organic pollutants. Reproduct Toxicol 2017; 68: 207-214. 117 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 304. K arlsen M, Grandjean P, W eihe P, Steuerwald U, Oulhote Y, Valvi D. Early-life exposures to persistent organic pollutants in relation to overw eight in preschool children. Reproduct Toxicol 2017; 68: 145-153. 305. D alsag er L, C hristensen N , H usby S, K yhl H, N ielsen F, H 0 st A, G randjean P, Jensen TK. Association betw een prenatal exposure to perfluorinated compounds and symptoms of infections at age 1-4years among 359 children in the Odense Child Cohort. Environ Int 2016; 96: 58-64. 306. Oulhote Y, Steuerwald U, Debes F, W eihe P, Grandjean P. Behavioral difficulties in 7-year old children in relation to developm ental exposure to perfluorinated alkyl substances. Environ Int 2016; 97: 237-45. 307. W eihe P, D ebes F, H alling J, Petersen M S, M uckle G, O dland J 0 , D udarev A, A yotte P, Dewailly E, Grandjean P, Bonefeld-J0 rgensen E. Health effects associated w ith measured levels o f contam inants in the Arctic. Int J Circum polar Health 2016; 75: 33805. 308. Vandenberg LN, Agerstrand M, Beronius A, Beausoleil C, Bergm an A, Bero LA, Bornehag CG, B oyer CS, C ooper GS, C otgreave I, Gee D, G randjean P, G uyton KZ, H ass U, H eindel JJ, Jobling S, K idd K A , K ortenkam p A, M acleod M R , M artin OV, N o rin d er U, S cheringer M , Thayer KA, Toppari J, W haley P, W oodruff TJ, R uden C. A proposed fram ew ork for the systematic review and integrated assessm ent (SYRINA) of endocrine disrupting chemicals. E nviron H ealth 2016; 15: 74. 309. O ulhote Y, D ebes F, V estergaard S, W eihe P, G randjean P. A erobic fitness and neurocognitive function scores in young Faroese adults and potential m odification by prenatal m ethylm ercury exposure. Environ H ealth Perspect (Environ H ealth Perspect 2017; 125: 677-683. 310. K irk LE, J0 rgensen JS, Nielsen F, Grandjean P. Role of hair-mercury analysis and dietary advice in lowering m ethylm ercury exposure in pregnant women. Scand J Publ Health 2017; 45: 444-51. 311. Timmermann CAG, Budtz-J0 rgensen E, Jensen TK, Osuna CE, Petersen MS, Steuerwald U, N ielsen F, Poulsen LK, W eihe P, Grandjean P. A ssociation betw een perfluoroalkyl substance exposure and asthm a and allergic disease in children as m odified by M M R vaccination. J Im m unotoxicol 2017; 14: 39-49. 312. Sunderland EM , Driscoll CT Jr, H am m itt JK, G randjean P, Evans JS, Blum JD, Chen CY, E vers D C , Jaffe D A , M ason RP, G oho S, Jacobs W . B enefits o f regulating hazardous air pollutants from coal and oil-fired utilities in the U nited States. Environ Sci Technol. 2016; 50: 2117-20. 313. D alsag er L, C hristensen N , H usby S, K yhl H, N ielsen F, H 0 st A, G randjean P, Jensen TK. Association betw een prenatal exposure to perfluorinated compounds and symptoms of infections at age 1-4years among 359 children in the Odense Child Cohort. Environ Int 2016; 96: 58-64. 314. W eihe P, D ebes F, H alling J, Petersen M S, M uckle G, O dland J 0 , D udarev A, A yotte P, Dewailly E, Grandjean P, Bonefeld-J0 rgensen E. Health effects associated w ith measured levels o f contam inants in the Arctic. Int J Circum polar Health 2016; 75: 33805. 315. Yorifuji T, M atsuoka K, Grandjean P. H eight and blood chem istry in adults w ith a history o f developm ental arsenic poisoning from contam inated m ilk powder. Environ R es 2017; 155: 86-91. 316. Grandjean P, Heilm ann C, W eihe P, Nielsen F, M ogensen UB, Budtz-J0 rgensen E. Serum Vaccine Antibody Concentrations in Adolescents Exposed to Perfluorinated Compounds. Environ H ealth Perspect 2017; 125: 077018. 118 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 317. Valvi D, Oulhote Y, W eihe P, D algard C, Bjerve KS, Steuerwald U, G randjean P. Gestational diabetes and offspring birth size at elevated environm ental pollutant exposures. Environ Int 2017; 107: 205-215. 318. Grandjean P, Heilm ann C, W eihe P, Nielsen F, M ogensen UB, Timm erm ann A, BudtzJ0 rgensen E. Estim ated exposures to perfluorinated compounds in infancy predict attenuated vaccine antibody concentrations at age 5 years. J Im m unotoxicol 2017; 14: 188-195. Books 24. Grandjean P (ed). Prenatal program m ing and toxicity. Basic Clin Pharmacol Toxicol. 2008; 102(2): 71-273. 25. G ee D, G randjean P, H ansen SF, van den H ove S, M acG arvin M , M artin J, N ielsen G, Q uist D, Stanners D, eds. Late Lessons from Early W arnings, volum e II (EEA Report N o 1/2013). Copenhagen, European Environm ent Agency, 2013, 746 pp. 26. Grandjean P. Only one chance. H ow Environm ental Pollution Im pairs Brain D evelopm ent and H ow to Protect the Brains o f the N ext Generation. N ew York: Oxford U niversity Press, 2013 (232 p p ). 27. Grandjean P, Herm ann P. Kemi pa hjernen - gar ud over enhver forstand. K 0 benhavn: Gyldendal, 2015 (334 sider). 28. Grandjean P. Cerveaux en danger (Brains in danger, in French). Translated by Odile D em ange. Paris: B uchet Chastel, 2016 (336 pp.). Book chapters and other publications 205. G randjean P, N ielsen JB. M ercury. In: Lippm an M, ed. Environm ental Toxicants: H um an E xposures and T heir H ealth E ffects, 3rd ed. N e w Y ork: W iley, 2009, pp. 811-22. 206. Landrigan P, N ordberg M, Lucchini R, N ordberg G, Grandjean P, Iregren A, Alessio L. The Declaration o f Brescia on Prevention o f the N eurotoxicity o f M etals. M ed Lav 2006; 97: 811-4. (Also published in Am J Ind M ed 2007 50: 709-11). 207. Grandjean P. Industrikem ikaliers pavirkning af nervesystem ets udvikling. U geskr L ^g er 2007; 169: 2782-4. 208. G randjean P, K eiding N. The precautionary principle. In: M elnick EL, E verett BS, eds. Encyclopedia o f Quantitative Risk Assessm ent and Analysis. Chichester: W iley, 2008, pp. 1290 3. 209. B udtz-J0 rgensen E, G randjean P. M ercury/m ethylm ercury risk. In: M elnick EL, E verett BS, eds. Encyclopedia o f Quantitative Risk A ssessm ent and Analysis. Chichester: W iley, 2008. 210. G randjean P. M ercury. In: H eggenhougen HK, ed. E ncyclopedia o f Public Health. Oxford: Elsevier, 2008, Vol. 4, pp. 434-42. 211. G randjean P. H ealth significance o f m etal exposures. In: W allace RB, ed. M axcy-RosenauLast Public Health & Preventive M edicine, 15th ed. N ew York, NY: M cGraw -H ill 2007, pp. 603-17. 212. G randjean P, B ellin g er D, B ergm an A, C ordier S, D avey-S m ith G, E skenazi B, G ee D, Gray K, H anson M, van den Hazel P, H eindel JJ, H einzow B, H ertzPicciotto I, H u H, H uang TTK, K old Jensen T, Landrigan PJ, M cM illen IC, M urata K, Ritz B, Schoeters G, Skakkeb^k N E , Skerfving S, W eihe P. T he F aroes statem ent: H um an health effects o f developm ental exposure to chem icals in our environm ent. Basic Clin Pharm acol Toxicol 2008; 102: 73-5. 119 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 213. Grandjean P. M ethylmercury toxicity and functional program m ing (correspondence). R eproduct Toxicol 2008; 25: 134. 214. G randjean P. Early vulnerability, lifelong im pacts. San Francisco M edicine. 2008; 81: 17-8. 215. Grandjean P, Heindel JJ. In utero and early-life conditions and adult health and disease (letter). N Engl J M ed 2008; 359: 1523. 216. Budtz-J0 rgensen E, K eiding N, Grandjean P. Approaches to handling uncertainty when setting environm ental exposure standards. In: B aveye P, M ysiak J, L aba M , eds. U ncertainties in environmental m odelling and consequences for policy making. Dordrecht, The Netherlands: Springer, 2009, pp. 267-80. 217. Grandjean P, Choi AL, W eihe P, M urata K. M ethylm ercury neurotoxicology: From rare poisonings to silent pandemic. In W ang C, Slikker W Jr, eds: D evelopm ental N eurotoxicological Research: Principles, M odels, Techniques, Strategies and M echanisms. N ew York: W iley, 2010, pp 335-56. 218. Straif K, Benbrahim -Tallaa L, Baan R, Grosse Y, Secretan B, El Ghissassi F, Bouvard V, Guha N, Freem an C, Galichet L, Cogliano V; W HO International Agency for Research on C ancer M onograph W orking Group. A review o f hum an carcinogens--part C: m etals, arsenic, dusts, and fibres. L an cet O ncol 2009; 10: 453-4. 219. G randjean P, Yorifuji T. M ercury (C hapter 8). In: Bingham E, Cohrssen B, eds. P atty 's T oxicology, 6th ed. N e w Y ork: W iley 2012, Vol. 1, pp 213-27. 220. T akaro TK , D avis D, V an R ensburg S, Jroyo A guilar R S, ... G randjean P et al. (108 authors). Scientists appeal to Quebec Prem ier Charest to stop exporting asbestos to the developing w orld. In t J O ccup E nviron H ealth 2010 16: 242-9. 221. D arney S, F o w ler B, G randjean P, H eindel J, M attison D, Slikker W Jr. Prenatal program m ing and toxicity II (PPTO X II): role o f environm ental stressors in the developm ental origins o f disease. Reprod Toxicol 2011; 31: 271. Also published in Journal o f Developm ental O rigins o f H ealth and D isease 2011; 2: 2. 222. Choi A, G randjean P. H um an health significance o f dietary exposures to m ethylm ercury. In: Liu G, Cai Y, O'Driscoll N, eds. Environm ental Chem istry and Toxicology o f M ercury. Chichester: W iley, 2012, pp. 545-67. 223. Grandjean P. Exposure to environm ental chemicals as a risk factor for diabetes developm ent. In: B ourguignon J-P, Jegou B, K erdelhue B, Toppari J, Christen Y, Eds. M ulti System Endocrine Disruption. Berlin: Springer 2011, pp. 91-9. 224. Julvez J, Yorifuji T, Choi AL, G randjean P. Epidem iological evidence on m ethylm ercury neurotoxicity. In: A schner M , C eccatelli S, eds. M ethylm ercury and N eurotoxicity. Berlin: Springer, 2012, pp. 13-35. 225. Grandjean P. Strengths and lim itations o f H B M - Im precision matters. Int J H yg Environ Health 2012; 215: 94. 226. G randjean P. Larry N eedham and the partition ratio. Chem osphere 2011; 85: 142. 227. W eihe P, Grandjean P. Cohort studies o f Faroese children concerning potential adverse health effects after the m others' exposure to m arine contam inants during pregnancy. Acta Vet Scand 2012; 54(Suppl 1): S7. 228. Fox DA, Grandjean P, de Groot D, Paule M. Developm ental origins of adult diseases and neurotoxicity: Epidem iological and experimental studies. N eurotoxicology 2012; 33: 810-6. 229. London L, Beseler C, Bouchard Mf, Bellinger DC, Colosio C, Grandjean P, Harari R, K ootbodien T, K rom hout H, Little F, M eijster T, M oretto A, R ohlm an DS, Stallones L. Neurobehavioural and neurodevelopm ental effects of pesticide exposures. Neurotoxicology 120 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 2012; 33: 887-96. 230. B al-P rice A K , C oecke S, C osta L, C rofton K M , F ritsche E, G oldberg A, G randjean P, L ein PJ, Li A, L ucchini R, M undy W R , P ad illa S, P ersico A, Seiler A E M , K reysa J. C onference Report: Advancing the Science o f D evelopm ental N eurotoxicity (DNT) Testing for B etter Safety Evaluation. Altex 2012: 29: 202-15. 231. Grandjean P, Heilm ann C. Perfluorinated compounds and imm unotoxicity in children Reply (Letter). JA M A 2012; 307: 1910-1. 232. Schug TT, B arouki R, G luckm an P, G randjean P, H anson M, H eindel JJ.PPT O X III: Environm ental Stressors in the Developm ental Origins of Disease: Evidence and M echanisms. Toxicol Sci 2013; 131: 343-50. 233. Andersen HR, W ohlfahrt-V eje C, Debes F, Nielsen F, Jensen TK, Grandjean P, M ain KM. Langtidseffekter af prenatal pesticideksponering (Long-term effects of prenatal pesticide exposure, in Danish). Copenhagen: M ilj0 styrelsen (Danish Environm ental Protection Agency), 2012. 234. Grandjean P. Blyforgiftning i forebyggelse og forskning (Leder) [Lead poisoning in prevention and research (Editorial)]. U geskr Laeger 2012; 174: 2693. 235. Grandjean P, Pichery C, Bellanger M , B udtz-J0 rgensen E. Calculation o f m ercury's effects on neurodevelopm ent (letter). Environ H ealth Perspect 2012; 120: a452. 236. Grandjean P, K eiding N. (2013) Precautionary Principle. In: El-Shaaraw i AH, Piegorsch W (eds), Encyclopedia o f Environm etrics. Chichester, UK: John W iley, 2013. DOI: 10.1002/9780470057339.vnn011. 237. Grandjean P. Science for precautionary decision-m aking. In: Gee D, Grandjean P, H ansen SF, van den H ove S, M acG arvin M , M artin J, N ielsen G, Q uist D, Stanners D. L ate L essons from Early W arnings, volum e II (EEA Report N o 1/2013). Copenhagen, European Environm ent Agency, 2013, pp. 517-35. 238. G randjean P. Opinion: Toxicants and the Brain. The Scientist 2013 (June 17): 36043. 239. Choi AL, G randjean P, Sun G, Z hang Y. D evelopm ental fluoride neurotoxicity: Choi et al. respond (Letter). Environ H ealth Perspect 2013; 121: A70. 240. Grandjean P. Opinion: Problem s with Hidden COI. The Scientist 2013 (October 28): 37934. 241. Grandjean P, Budtz-J0 rgensen E. Epidemiological approaches to m etal toxicology (C hapter 13). In: N ordberg GF, Fow ler B, N ordberg M , Friberg LT, eds. H an dbook on the toxicology o f m etals, V olum e 1, 4th ed. A m sterdam : E lsevier, 2014, pp. 2 6 5 79. 242. Landrigan PJ, Lucchini R, K otelchuck D, Grandjean P. Principles for prevention of toxic effects from m etals (C hapter 24). In: N ordberg GF, Fow ler B, N ordberg M , eds. H andbook on the toxicology o f metals, 4th ed. Am sterdam: Elsevier, 2014, pp. 507-28. 243. Grandjean P. Developmental origins of diseases: challenge for risk assessm ent of chemicals (EU RO TO X abstract). Toxicol Lett 2013; 221 Suppl: S15. 244. Grandjean P. M ercury (C hapter 29). In: Landrigan PJ, Etzel RA, eds. C hildren's Environm ental Health. N ew York: Oxford U niversity Press, 2014, pp. 273-80. 245. Heilm ann C, Jensen L, W eihe P, Nielsen F, Knudsen LE, Budtz-J0 rgensen E, M 0 lbak K, Grandjean P. Persistente fluorforbindelser reducerer im m unfunktionen (Persistent perfluorinated com pounds cause im m unotoxic effects, in Danish). U geskr Laeg 2015; 177: 660-3. 246. G randjean P. Chem ical brain drain: insidious and pervasive. In: Breyer, H, ed. Giftfreies Europa. Brussels, 2014, pp. 133-40. 247. Grandjean P. M ercury (article 02853). In: Caplan M , ed. R eference M odule in Biom edical 121 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN Sciences. Elsevier, 2015. 248. G randjean P, Landrigan PJ. N eurodevelopm ental toxicity: still m ore questions than answers - A uthors' response. L an cet N eurol 2014; 13: 648-9. 249. Grandjean P. Prenatal prevention (letter). Science 2014; 345: 1462. 250. Birnbaum LS, Grandjean P. Alternatives to PFASs: Perspectives on the science (editorial). Environ H ealth Perspect 2015; 123: A104-5. 251. Oulhote Y, Grandjean P. A ssociation betw een child poverty and academ ic achievem ent (letter). JA M A Pediatr 2016; 170: 179-80. 252. K ielsen K, Shamim Z, Ryder LP, Grandjean P, Heilm ann C. Vaccination efficacy and environm ental pollution. In: Esser C (ed.). Environm ental Influences on the Im m une System. Vienna: Springer, 2016, pp. 181-203. 253. Trasande L, A ttina T, Skakkebaek NE, Juul A, Porta M , Soto AM, V andenberg L, Sathyanarayana S, F letch er T, D em eneix B, B ergm an A, C ohn BA , B ellan g er M , G ore A C, L egler J, B ourguignon JP, Slam a R, Toppari J, B lum berg B, M yers JP, Z oeller RT, K ortenkam p A, D iG angi J, Philippe G randjean P, R uss H auser R, R udel R. Endocrine disruptors: Refereed science to guide action on ED Cs (Correspondence). N ature 2016; 536: 30. 254. M ie A, G uyot EK , K ahl J, R em bialkow ska E, A ndersen H R , G randjean, P, G unnarsson S. Health im plications o f organic food and organic agriculture. Science and Technology Options A ssessm ent Panel, General for Parliam entary Research Services (D G EPRS) o f the European Parliament, 2016. 255. Grandjean P, Kishi R, Kogevinas M; International Society for Environm ental Epidemiology (ISEE). Prevention of developmental neurotoxicity. Epidem iology 2017; 28: 157-158. 122 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN EXHIBIT D CITED PUBLICATIONS 1. E uropean F ood Safety A uthority: O pinion o f the Scientific Panel on C ontam inants in the Food chain on Perfluorooctane sulfonate (PFOS), perfluorooctanoic acid (PFOA) and th eir salts. The EFSA Journal 2008, 653:1-131. 2. European Food Safety Authority: G uidance o f the Scientific Com m ittee on U se o f the b enchm ark dose approach in risk assessm ent. The EFSA Journal 2009, 1150:1-72. 3. C8 Science Panel: The Science Panel W ebsite. In. Edited by F letcher T, Steenland K, Savitz D; 2013. 4. A gency for Toxic Substances and D isease Registry: D raft toxicological profile for perfluoroalkyls. In. A tlanta, GA: A gency for Toxic Substances and D isease Registry; 2015. 5. N ational T oxicology Program : Im m unotoxicity A ssociated w ith E xposure to Perfluorooctanoic Acid (PFOA) or Perfluorooctane Sulfonate (PFOS). In., vol. 2017. Raleigh, NC: National Toxicology Program; 2016. 6. International A gency for R esearch on Cancer: Som e C hem icals U sed as Solvents and in P olym er M anufacture. P erfluorooctanoic acid. In: IARCMonographs On The Evaluation O f Carcinogenic Risks To Humans. vol. 110. Lyon: International A gency for R esearch on Cancer; 2016: 37-110. 7. Post GB, C ohn PD , C ooper KR: Perfluorooctanoic acid (PFO A ), an em erging drinking w ater contam inant: a critical review o f recent literature. Environ Res 2012, 116:93-117. 8. Steenland K, F letch er T, Savitz DA : E pidem iologic evidence on the health effects o f perfluorooctanoic acid (PFO A ). Environ Health Perspect 2010, 118(8):1100-1108. 9. W hite SS, Fenton SE, H ines EP: Endocrine disrupting properties o f perfluorooctanoic acid. J SteroidBiochem Mol Biol 2011, 127(1-2):16-26. 10. N e w Jersey D rinking W ater Q uality Institute H ealth E ffects Subcom m ittee: H ealth-based m axim um contam inant levels upport docum ent: Perfluorooctanoic acid (PFOA ). In. Edited by Protection NJDoE. Trenton, NJ: N ew Jersey D rinking W ater Quality Institute; 2016. 11. K au ck E A , D iesslin AR: Som e P roperties o f P erfluorocarboxylic A cids. Industrial & Engineering Chemistry 1951, 43(10):2332-2334. 12. Paul A G , Jones K C , Sw eetm an AJ: A first global production, em ission, and environm ental inventory for perfluorooctane sulfonate. Environ Sci Technol 2009, 43(2):386-392. 13. K ey BD , H ow ell RD , C riddle CS: F luorinated organics in the biosphere. Environ Sci Technol 1997, 31(9):2445-2454. 14. B u tt CM , B erg er U, B ossi R, T om y GT: L evels and trends o f poly- and perfluorinated com pounds in the arctic environm ent. Sci Total Environ 2010, 408(15):2936-2965. 15. H ou de M , D e Silva A O , M u ir D C , L etch er RJ: M onitoring o f perfluorinated com pounds in aquatic biota: an up d ated review . Environ Sci Technol2011, 45(19):7962-7973. 16. M innesota D ep artm en t o f H ealth: E ast M etro P F C C om m unity U pdates. In. St-Paul, M N: Environm ental Public Health Tracking and Biomonitoring, M innesota Departm ent of Health; 2013. 123 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 17. A gency fo r T oxic Substances and D isease R egistry: D raft toxicological profile for perfluoroalkyls. In. A tlanta, GA: A gency for Toxic Substances and D isease Registry; 2009. 18. M innesota D ep artm en t o f H ealth: P ublic H ealth A ssessm ent for perfluorochem ical contam ination in Lake Elm o and Oakdale, W ashington County, M innesota. In. Atlanta, GA: Agency for Toxic Substances and Disease Registry; 2008. 19. X iao X, U lrich B A , C hen B, H iggins CP: Sorption o f P oly- and Perfluoroalkyl Substances (PFASs) Relevant to Aqueous Film -Form ing Foam (AFFF)-Impacted G roundw ater by B iochars and A ctivated Carbon. Environ Sci Technol 2017, 51(11):6342-6351. 20. H olzer J, Goen T, Just P, R eupert R, Rauchfuss K, K raft M, M uller J, W ilhelm M: Perfluorinated compounds in fish and blood o f anglers at Lake M ohne, Sauerland area, G erm any. Environ Sci Technol 2011, 45(19):8046-8052. 21. Delinsky AD, Strynar MJ, M cCann PJ, Varns JL, M cM illan L, N akayam a SF, Lindstrom AB: Geographical distribution of perfluorinated compounds in fish from M innesota lakes and rivers. Environ Sci Technol 2010, 44(7):2549-2554. 22. Delinsky AD, Strynar MJ, N akayam a SF, Varns JL, Ye X, M cCann PJ, Lindstrom AB: D eterm ination of ten perfluorinated compounds in bluegill sunfish (Lepomis m acrochirus) fillets. Environ Res 2009, 109(8):975-984. 23. Cardno Entrix: M ississippi River Pool 2 PFOS Assessm ent - Fish Tissues and W aterSum m er 2011. In. Okem os, MI: Cardno Entrix; 2011. 24. Satoh JI: M olecular netw ork analysis o f hum an m icroR N A targetom e: from cancers to A lzheim er's disease. BioData mining 2012, 5(1):17. 25. M ondal D, L opez-E spinosa M J, A rm strong B, Stein CR, F letcher T: R elationships o f perfluorooctanoate and perfluorooctane sulfonate serum concentrations betw een motherchild pairs in a population w ith perfluorooctanoate exposure from drinking water. Environ Health Perspect 2012, 120(5):752-757. 26. Kato K, Calafat AM, W ong LY, W anigatunga AA, Caudill SP, N eedham LL: Polyfluoroalkyl com pounds in pooled sera from children participating in the National H ealth and N utrition E xam ination Survey 2001-2002. Environ Sci Technol 2009, 43(7):2641-2647. 27. C ariou R, V eyrand B, Y am ada A, B errebi A, Z alko D, D urand S, P o llono C, M archand P, L eblanc JC, A ntignac JP et al: Perfluoroalkyl acid (PFA A ) levels and profiles in breast m ilk, m aternal and cord serum o f F rench w o m en and th eir new borns. Environ Int 2015, 84:71-81. 28. N eedham LL, G randjean P, H einzow B, Jorgensen PJ, N ielsen F, Patterson DG, Jr., Sjodin A, Turner W E, W eihe P: Partition o f environm ental chem icals betw een m aternal and fetal blood and tissues. Environ Sci Technol 2011, 45(3):1121-1126. 29. L iu J, Li J, Liu Y, Chan H M , Zhao Y, Cai Z, W u Y: Com parison on gestation and lactation exposure o f perfluorinated com pounds for new borns. Environ Int 2011, 37(7):1206-1212. 30. M ogensen UB, Grandjean P, Nielsen F, W eihe P, Budtz-Jorgensen E: Breastfeeding as an E xposure Pathw ay for P erfluorinated A lkylates. Environ Sci Technol 2015, 49(17):10466-10473. 124 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 31. B utenhoff JL, Kennedy GL, Jr., Fram e SR, O 'Connor JC, Y ork RG: The reproductive toxicology o f am m onium perfluorooctanoate (A PFO ) in the rat. Toxicology 2004, 196(1- 2):95-116. 32. H inderliter PM , M ylchreest E, G annon SA, B utenhoff JL, K ennedy GL, Jr.: Perfluorooctanoate: Placental and lactational transport pharm acokinetics in rats. Toxicology 2005, 211(1-2):139-148. 33. W orld H ealth Organization: Fourth W H O -Coordinated Survey of H um an M ilk for Persistent O rganic Pollutants in Cooperation w ith UNEP. In. Geneva: W orld H ealth Organization; 2007. 34. G randjean P, Jensen AA: B reastfeeding and the w eanling's dilem m a. Am JPublic Health 2004, 94(7):1075; author reply 1075-1076. 35. Grandjean P, Andersen EW , Budtz-Jorgensen E, Nielsen F, M olbak K, W eihe P, H eilm ann C: Serum vaccine antibody concentrations in children exposed to perfluorinated com pounds. JAMA : thejournal o f the American Medical Association 2012, 307(4):391-397. 36. M ondal D, W eldon RH, Arm strong BG, Gibson LJ, Lopez-Espinosa M J, Shin HM, Fletcher T: Breastfeeding: a potential excretion route for m others and im plications for infant exposure to perfluoroalkyl acids. Environ Health Perspect 2014, 122(2):187-192. 37. M innesota Departm ent of Health: M D H issues new guidance on chemicals in some private wells, city w ater in E ast M etro, Bem idji. In., vol. 2017. St.Paul, M N: M innesota D epartm ent of Health; 2017. 38. G uruge K S, M anage PM , Y am anaka N , M iyazaki S, T aniyasu S, Y am ashita N: Speciesspecific concentrations of perfluoroalkyl contam inants in farm and pet animals in Japan. Chemosphere 2008, 73(1 Suppl):S210-215. 39. B raunig J, B aduel C, H effernan A, R otander A, D onaldson E, M ueller JF: Fate and redistribution o f perfluoroalkyl acids th rough A F F F -im pacted groundw ater. Sci Total Environ 2017, 596-597:360-368. 40. K o w alczy k J, E hlers S, O berhausen A, T ischer M , F u rst P, Schafft H, L ahrssenW iederholt M: Absorption, distribution, and m ilk secretion o f the perfluoroalkyl acids PFB S, PFH xS, PFO S, and PFO A by dairy cow s fed naturally contam inated feed. Journal o f agricultural andfood chemistry 2013, 61(12):2903-2912. 41. Bjorklund JA, Thuresson K, De W it CA: Perfluoroalkyl compounds (PFCs) in indoor dust: concentrations, hum an exposure estim ates, and sources. Environ Sci Technol 2009, 43(7):2276-2281. 42. Taves DR: E vidence th at there are tw o form s o f fluoride in hum an serum . Nature 1968, 217(5133):1050-1051. 43. Guy WS: Fluorocom pound(s) of Hum an Plasma: Analysis, Prevalence, Purification and Characterization. Rochester, NY: University of Rochester School of M edicine and Dentistry; 1972. 44. G riffith FD , L ong JE: A nim al toxicity studies w ith am m onium perfluorooctanoate. Am IndHygAssoc J 1980, 41(8):576-583. 45. Franko J, M eade BJ, Frasch HF, Barbero AM , Anderson SE: Derm al penetration potential o f perfluorooctanoic acid (PFO A ) in hum an and m ouse skin. Journal o f toxicology and environmental health Part A 2012, 75(1):50-62. 125 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 46. Centers for D isease Control and Prevention: Fourth national report on hum an exposure to environm ental chem icals, U pdated tables. In. A tlanta, GA: C enters for D isease Control and Prevention; 2012. 47. Guy W S, Taves DR, Brey WS: Organic fluorocompounds in hum an-plasm a - prevalence and characterization. ACS Symposium Series 1976(28):117-134. 48. O zo n o ff D: M edical aspects o f the hazardous w aste problem . The Americanjournal o f forensic medicine andpathology 1982, 3(4):343-348. 49. Olsen GW, Church TR, M iller JP, Burris JM, Hansen KJ, Lundberg JK, Armitage JB, H erron RM , M edhdizadehkashi Z, N obiletti JB et al: Perfluorooctanesulfonate and other fluorochem icals in the serum o f A m erican R ed C ross adult blood donors. Environ Health Perspect 2003, 111(16):1892-1901. 50. Calafat AM , K uklenyik Z, Reidy JA, Caudill SP, Tully JS, N eedham LL: Serum concentrations o f 11 polyfluoroalkyl com pounds in the U .S. population: data from the national health and nutrition exam ination survey (N H A N E S). Environ Sci Technol 2007, 41(7):2237-2242. 51. Jin H, Zhang Y, Jiang W , Zhu L, M artin JW: Isom er-Specific Distribution o f Perfluoroalkyl Substances in B lood. Environ Sci Technol 2016, 50(14):7808-7815. 52. Perez F, N adal M , N avarro-O rtega A, Fabrega F, D om ingo JL, Barcelo D, Farre M: A ccum ulation o f perfluoroalkyl substances in hum an tissues. Environ Int 2013, 59:354 362. 53. K ato K, W ong LY, Jia LT, K uklenyik Z, Calafat AM: Trends in Exposure to Polyfluoroalkyl C hem icals in the U .S. Population: 1999-2008. Environ Sci Technol2011. 54. M innesota D epartm ent o f Health: East M etro perfluorochem ical biom onitoring project. In. St. Paul, M N : E nvironm ental P ublic H ealth T racking and B iom onitoring, M innesota D epartm ent of Health; 2009. 55. Egeghy PP, Lorber M: An assessm ent o f the exposure o f Am ericans to perfluorooctane sulfonate: a com parison o f estim ated intake w ith values inferred from N H A N ES data. J Expo Sci Environ Epidemiol 2011, 21(2):150-168. 56. Lorber M, Egeghy PP: Simple intake and pharm acokinetic m odeling to characterize exposure o f A m ericans to perfluoroctanoic acid, PFO A . Environ Sci Technol 2011, 45(19):8006-8014. 57. Olsen GW , Burris JM, Ehresm an DJ, Froehlich JW , Seacat AM , B utenhoff JL, Zobel LR: Half-life of serum elimination of perfluorooctanesulfonate,perfluorohexanesulfonate, and perfluorooctanoate in retired fluorochem ical production w orkers. Environ Health Perspect 2007, 115(9):1298-1305. 58. A ndersen M E, Clewell HJ, 3rd, Tan YM, B utenhoff JL, Olsen GW: Pharm acokinetic m odeling o f saturable, renal resorption o f perfluoroalkylacids in m onkeys--probing the determ inants o f long plasm a half-lives. Toxicology 2006, 227(1-2):156-164. 59. V erner M A, Loccisano AE, M orken N H , Yoon M , W u H, M cD ougall R, M aisonet M, M arcus M , K ishi R, M iyashita C et al: A ssociations o f Perfluoroalkyl Substances (PFA S) w ith Low er Birth W eight: An Evaluation o f Potential Confounding by Glom erular F iltration R ate U sing a P hysiologically B ased P harm acokinetic M odel (PB PK ). Environ Health Perspect 2015, 123(12):1317-1324. 60. Olsen GW , Lange CC, Ellefson M E, M air DC, Church TR, Goldberg CL, Herron RM, M edhdizadehkashi Z, N obiletti JB, R ios JA et al: Tem poral trends o f perfluoroalkyl 126 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN concentrations in A m erican R ed C ross adult b lood donors, 2000-2010. Environ Sci Technol 2012, 46(11):6330-6338. 61. Genuis SJ, Liu Y, Genuis QI, M artin JW: Phlebotom y treatm ent for elim ination o f perfluoroalkyl acids in a highly exposed fam ily: a retrospective case-series. PLoS One 2014, 9(12):e114295. 62. Lorber M, Eaglesham GE, Hobson P, Toms LM , M ueller JF, Thom pson JS: The effect of ongoing blood loss on hum an serum concentrations o f perfluorinated acids. Chemosphere 2015, 118:170-177. 63. M innesota Departm ent of Health: Environm ental Public Health Tracking & B iom onitoring. In: Report to the Minnesota Legislature. St.Paul, M N: M innesota D epartm ent of Health; 2011. 64. M innesota D epartm ent o f H ealth: E ast M etro PFC3 B iom onitoring Project. In. St. Paul, MN: Environm ental Public Health Tracking and Biomonitoring, M innesota Departm ent of Health; 2015. 65. Em m ett EA, Shofer FS, Zhang H, Freem an D, Desai C, Shaw LM: Com m unity exposure to perfluorooctanoate: relationships betw een serum concentrations and exposure sources. J Occup Environ Med 2006, 48(8):759-770. 66. Olsen GW , Burris JM, Burlew M M , M andel JH: Plasm a cholecystokinin and hepatic enzymes, cholesterol and lipoproteins in amm onium perfluorooctanoate production w orkers. Drug Chem Toxicol 2000, 23(4):603-620. 67. Chang SC, Das K, Ehresm an DJ, Ellefson M E, Gorm an GS, H art JA, N oker PE, Tan YM , L ieder PH , L au C et al: C om parative pharm acokinetics o f perfluorobutyrate in rats, mice, m onkeys, and hum ans and relevance to hum an exposure via drinking water. Toxicol Sci 2008, 104(1):40-53. 68. H offm an K, W ebster TF, Bartell SM, W eisskopf M G, Fletcher T, V ieira VM: Private drinking w ater wells as a source o f exposure to perfluorooctanoic acid (PFOA) in com m unities surrounding a fluoropolym er production facility. Environ Health Perspect 2011, 119(1):92-97. 69. Lindh CH, Rylander L, Toft G, Axm on A, Rignell-Hydbom A, Giwercman A, Pedersen HS, G oalczyk K, L udw icki JK, Z vyezday V et al: B lood serum concentrations o f perfluorinated compounds in m en from Greenlandic Inuit and European populations. Chemosphere 2012, 88(11):1269-1275. 70. W eihe P, K ato K, Calafat AM, N ielsen F, W anigatunga AA, N eedham LL, Grandjean P: Serum concentrations of polyfluoroalkyl compounds in Faroese w hale m eat consumers. Environ Sci Technol 2008, 42(16):6291-6295. 71. Lindstrom AB, Strynar M J, Libelo EL: Polyfluorinated compounds: past, present, and future. Environ Sci Technol 2011, 45(19):7954-7961. 72. Gilliland FD, M andel JS: M ortality am ong em ployees o f a perfluorooctanoic acid production plant. J Occup M ed 1993, 35(9):950-954. 73. Leonard RC, K reckm ann KH, Sakr CJ, Symons JM: Retrospective cohort m ortality study of workers in a polym er production plant including a reference population of regional w orkers. Ann Epidemiol 2008, 18(1):15-22. 74. Sakr CJ, Symons JM, Kreckm ann KH, Leonard RC: Ischaemic heart disease m ortality study among w orkers w ith occupational exposure to amm onium perfluorooctanoate. Occup Environ Med 2009, 66(10):699-703. 127 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 75. Steenland K, D e d d e n s J, Salvan A, Stayner L: Negative bias in exposure-response trends in occupational studies: modeling the healthy workers survivor effect. Am J Epidemiol 1996, 143(2):202-210. 76. Lund in JI, Alexander B H , Olsen G W , C h u r c h T R : A m m o n i u m perfluorooctanoate production and occupational mortality. Epidemiology 2009, 20(6):921-928. 77. Grandjean P, Bellinger D, B e r g m a n A, Cordier S, D a ve y- Sm it h G, Eskenazi B, G e e D, Gray K, H a n s o n M , va n den Hazel P et al: T h e Faroes statement: h u m a n health effects of developmental exposure to chemicals in our environment. Basic Clin Pharmacol Toxicol 2008, 102(2):73-75. 78. Hines EP, Wh i t e SS, Stanko JP, Gibbs-Flournoy E A , L a u C, Fenton SE: Phenotypic dichotomy following developmental exposure to perfluorooctanoic acid ( P F O A ) in female C D - 1 mice: L o w doses induce elevated serum leptin and insulin, and overweight in mid-life. Mol Cell Endocrinol 2009, 304(1-2):97-105. 79. Grandjean P: Science for precautionary decision-making. In: Late Lessonsfrom Early Warnings. Edited by G e e D, Grandjean, P., Hansen, S.F., van den Hove, S., MacGarvin, M., Martin, J., Nielsen, G., Quist, D., Stanners, D., vol. II. Copenhagen: European Environment Agency; 2013: 517-535. 80. Michaels D: D o u b t is their product: h o w industry's assault on science threatens your health. Oxfo rd ;N e w York: Oxford University Press; 2008. 81. B l u m A, Balan SA, Scheringer M , Trier X, G o l d e n m a n G, Cousins IT, D i a m o n d M , Fletcher T, Higgins C, L i n d e m a n A E et al: T h e M a dr id Statement on Poly- and Perfluoroalkyl Substances (PFASs). Environ Health Perspect 2015, 123(5):A107-111. 82. European Environment Agency: Late lessons from early warnings: the precautionary principle 1896-2000. In: Environmental issue report No 22. Copenhagen; 2001. 83. Zobel L R , Olsen G W , Butenhoff JL: Perfluorinated c o m p o u n d s and immunotoxicity in children. JAMA : thejournal o f the American Medical Association 2012, 307(18):1910; author reply 1910-1911. 84. Ubel F A , Sorenson SD, R o a c h D E : Health status of plant workers exposed to fluorochemicals--a preliminary report. Am IndH yg Assoc J 1980, 41(8):584-589. 85. H o d g e H C , Smith F A : Occupational fluoride exposure. J Occup M ed 1977, 19(1):12-39. 86. N e e d l e m a n H L : Clair Patterson and Robert Kehoe: t w o views of lead toxicity. Environ Res 1998, 78(2):79-85. 87. Tsai PL, Hatfield T H : Global benefits from the phaseout of leaded fuel. J Environ Health 2011, 74(5):8-14. 88. Olsen G W , Gilliland F D , B u r l e w M M , Burris J M , M a n d e l JS, M a n d e l JH: A n epidemiologic investigation of reproductive h o r m o n e s in m e n with occupational exposure to perfluorooctanoic acid. J Occup Environ M ed 1998, 40(7):614-622. 89. W e n CP, Tsai SP: A n a t o m y of the health worker effect - a critique of s u m m a r y statistics employed in occupational epidemiology. Scand J Work Environ Health 1982, 8 Suppl 1:48-52. 90. Clapp R, Hoppin, P.: Perfluorooctanoic Acid. In: Defending Science. T h e Project on Scientific K n o w l e d g e and Public Policy; 2011. 91. Gutshall D M , Pilcher G D , Langley A E : Effect of thyroxine supplementation on the response to perfluoro-n-decanoic acid ( P F D A ) in rats. J Toxicol Environ Health 1988, 24(4):491-498. 128 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. Langley A E , Pilcher G D : Thyroid, bradycardic and hypothermic effects of perfluoro-n- decanoic acid in rats. J Toxicol Environ Health 1985, 15(3-4):485-491. Raleigh K K , Alexander B H , Olsen G W , Ra machandran G, M o r e y SZ, Church TR, L o ga n P W , Scott LL, Allen E M : Mortality and cancer incidence in a m m o n i u m perfluorooctanoate production workers. Occup Environ M ed 2014, 71(7):500-506. Hardell E, K a r r m a n A, v a n Bavel B, B a o J, Carlberg M , Hardell L: Case-control study on perfluorinated alkyl acids ( P F A A s ) and the risk of prostate cancer. Environ Int 2014, 63:35-39. Mc Ga ri ty T O , W a g n e r W E : Be nd in g Science: H o w special interests corrupt public health research. Boston: Harvard University Press; 2008. Fei C, Olsen J: Prenatal exposure to perfluorinated chemicals and behavioral or coordination problems at age 7 years. Environ Health Perspect 2011, 119(4):573-578. Eriksen K T , Sorensen M , M c L a u g h l i n JK, Lipworth L, Tjonneland A, O v er va d K, Raaschou-Nielsen O: Perfluorooctanoate and perfluorooctanesulfonate plasma levels and risk of cancer in the general Danish population. J Natl Cancer Inst 2009, 101(8):605-609. Qazi M R , Abedi M R , Nelson B D , DePierre J W , Abedi-Valugerdi M : Dietary exposure to perfluorooctanoate or perfluorooctane sulfonate induces hypertrophy in centrilobular hepatocytes and alters the hepatic i m m u n e status in mice. Int Immunopharmacol 2010, 10(11):1420-1427. Peters J M , Gonzalez FJ: W h y toxic equivalency factors are not suitable for perfluoroalkyl chemicals. Chem Res Toxicol 2011, 24(10):1601-1609. B a c h C C , Henriksen TB, Bossi R, B e c h B H , Fuglsang J, Olsen J, N o h r E A : Perfluoroalkyl A c id Concentrations in Bl o o d Samples Subjected to Transportation and Processing Delay. PLoS One 2015, 10(9):e0137768. Grandjean P, Budtz-Jorgensen E: A n ignored risk factor in toxicology: T h e total imprecision of exposure assessment. Pure Appl Chem 2010, 82(2):383-391. Jorgensen A W , Hilden J, Gotzsche PC: Cochrane reviews co mp a r e d with industry supported meta-analyses and other meta-analyses of the s a m e drugs: systematic review. BMJ 2006, 333(7572):782. Lesser LI, Ebbeling C B , G o o z n e r M , Wypij D, L u d w i g D S : Relationship between funding source and conclusion a m o n g nutrition-related scientific articles. PLoS M ed 2007, 4(1):e5. Jackler R K : Testimony b y otolaryngologists in defense of tobacco companies 2009-2014. Laryngoscope 2015, 125(12):2722-2729. E g il ma n D S , Ardolino EL, H o w e S, Bird T: Deconstructing a state-of-the-art review of the asbestos brake industry. New solutions : ajournal o f environmental and occupational health policy : NS 2011, 21(4):545-571. Grandjean P, Eriksen M L , Ellegaard O, Wallin JA: T h e M a t t h e w effect in environmental science publication: a bibliometric analysis of chemical substances in journal articles. Environ Health 2011, 10:96. Grandjean P, Barouki R, Bellinger D C , Casteleyn L, C h a d w i c k L H , Cordier S, Etzel R A , Gray K A , H a E H , Junien C et al: Life-Long Implications of Developmental Exposure to Environmental Stressors: N e w Perspectives. Endocrinology 2015, 156(10):3408-3415. Giesy JP, K a n n a n K: Global distribution of perfluorooctane sulfonate in wildlife. Environ Sci Technol2001, 35(7):1339-1342. 129 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 109. H ansen KJ, Clem en LA, Ellefson M E, Johnson HO: Com pound-specific, quantitative characterization o f organic fluorochem icals in biological m atrices. Environ Sci Technol 2001, 35(4):766-770. 110. B uck RC, Franklin J, B erger U, C onder JM, C ousins IT, de V oogt P, Jensen AA, K annan K, M abury SA, van Leeuwen SP: Perfluoroalkyl and polyfluoroalkyl substances in the environm ent: term inology, classification, and origins. Integrated environmental assessment and management 2011, 7(4):513-541. 111. National Research Council: Science and decisions: advancing risk assessment. W ashington, D.C.: National Academ y Press; 2009. 112. C8 Science Panel: Status Report: PFO A and im m une biom arkers in adults exposed to PFO A in drinking w ater in the m id Ohio valley. In. Edited by Fletcher T, Steenland K, Savitz D; 2009. 113. Grandjean P, H eilm ann C, W eihe P, N ielsen F, M ogensen UB, Tim m erm ann A, BudtzJorgensen E: Estim ated exposures to perfluorinated compounds in infancy predict attenuated v accine antibody concentrations at age 5-years. J Immunotoxicol 2017, 14(1):188-195. 114. van Loveren H, Germ olec D, K oren HS, Luster M I, N olan C, R epetto R, Smith E, Vos JG, V ogt RF: Report o f the Bilthoven Symposium: A dvancem ent o f Epidem iological Studies in Assessing the H um an Health Effects o f Im m unotoxic Agents in the E n vironm ent and the W orkplace. Biomarkers 1999, 4(2):135-157. 115. W eisglas-K uperus N , P atandin S, B erbers G A , Sas TC, M u ld er PG , Sauer PJ, H ooijkaas H: Im m unologic effects o f background exposure to polychlorinated biphenyls and dioxins in D utch preschool children. Environ Health Perspect 2000, 108(12):1203-1207. 116. Heilm ann C, Grandjean P, W eihe P, N ielsen F, Budtz-Jorgensen E: Reduced antibody responses to vaccinations in children exposed to polychlorinated biphenyls. PLoS Med 2006, 3(8):e311. 117. Stolevik SB, N ygaard UC, N am ork E, H augen M , M eltzer HM , A lexander J, Knutsen HK , A aberge I, V ainio K, van L overen H et al: Prenatal exposure to polychlorinated biphenyls and dioxins from the maternal diet may be associated with im m unosuppressive effects th at p ersist into early childhood. Food Chem Toxicol 2013, 51:165-172. 118. Jusko TA, D e Roos AJ, Lee SY, Thevenet-M orrison K, Schwartz SM, V erner M A, M urinova LP, D robna B, K ocan A, F abisikova A et al: A B irth C ohort Study o f M aternal and Infant Serum PCB-153 and DD E Concentrations and Responses to Infant T uberculosis V accination. Environ Health Perspect 2016, 124(6):813-821. 119. Steenland K, Zhao L, W inquist A, Parks C: U lcerative colitis and perfluorooctanoic acid (PFOA) in a highly exposed population o f com m unity residents and w orkers in the mid- O hio valley. Environ Health Perspect 2013, 121(8):900-905. 120. M ogensen UB, B udtz-J0 rgensen E, H eilm ann C, N ielsen F, W eihe P, G randjean P: Structural equation m odeling of im m unotoxicity associated with exposure to perfluorinated com pounds. Environ Health 2015, 14(47). 121. D ietert RR: Developm ental im m unotoxicology (DIT): w indow s o f vulnerability, im m une dysfunction and safety assessm ent. J Immunotoxicol 2008, 5(4):401-412. 122. Grandjean P, H eilm ann C, W eihe P, N ielsen F, M ogensen UB, Budtz-Jorgensen E: Serum Vaccine Antibody Concentrations in Adolescents Exposed to Perfluorinated Com pounds. Environ Health Perspect 2017, 125(7):077018. 130 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. Grandjean P, H e i l m a n n C: Perfluorinated c o m p o u n d s and immunotoxicity in children - Reply. JAMA : thejournal o f the American Medical Association 2012, 307:1910-1911. Stein C R , M c G o v e r n KJ, Pajak A M , Maglione PJ, W o l f f M S : Perfluoroalkyl and polyfluoroalkyl substances and indicators of i m m u n e function in children aged 12-19 y: National Health and Nutrition Examination Survey. Pediatr Res 2016, 79(2):348-357. L o ok er C, Luster MI, Calafat A M , Johnson VJ, Burleson G R , Burleson F G , Fletcher T: Influenza vaccine response in adults exposed to perfluorooctanoate and perfluorooctanesulfonate. ToxicolSci 2014, 138(1):76-88. Kielsen K, S h a m i m Z, R y d e r LP, Nielsen F, Grandjean P, Budtz-Jorgensen E, H e i l m a n n C: Antibody response to booster vaccination with tetanus and diphtheria in adults exposed to perfluorinated alkylates. J Immunotoxicol 2016, 13(2):270-273. Dalsager L, Christensen N, H u s b y S, K y h l H, Nielsen F, Host A, Grandjean P, Jensen T K : Association between prenatal exposure to perfluorinated c o m p o u n d s and s y m p t o m s of infections at age 1-4years a m o n g 359 children in the O d en se Child Cohort. Environ Int 2016, 96:58-64. G r a n u m B, H a u g LS, N a m o r k E, Stolevik SB, T h o m s e n C, Aaberge IS, v a n Loveren H, Lo vi k M , N y g a a r d U C : Pre-natal exposure to perfluoroalkyl substances m a y be associated with altered vaccine antibody levels and immune-related health outcomes in early childhood. J Immunotoxicol 2013, 10(4):373-379. Fei C, M c L a u g h l i n JK, Lipworth L, Olsen J: Prenatal exposure to P F O A and P F O S and risk of hospitalization for infectious diseases in early childhood. Environ Res 2010, 110(8):773-777. Goudarzi H, Miyashita C, O k a d a E, Kashino I, C h e n CJ, Ito S, Araki A, Kobayashi S, Matsuura H, Kishi R: Prenatal exposure to perfluoroalkyl acids and prevalence of infectious diseases up to 4years of age. Environ Int 2017, 104:132-138. Steenland K, Z h a o L, Winquist A: A cohort incidence study of workers exposed to perfluorooctanoic acid (PFOA). Occup Environ Med 2015, 72(5):373-380. Fairley KJ, Purdy R, Kearns S, An de rs on SE, M e a d e BJ: Exposure to the immunosuppressant, perfluorooctanoic acid, enhances the murine IgE and airway hyperreactivity response to ovalbumin. Toxicol Sci 2007, 97(2):375-383. W a n g IJ, Hsieh W S , C h e n C Y , Fletcher T, Lien G W , Chiang H L , Chiang CF, W u T N , C h e n PC: T h e effect of prenatal perfluorinated chemicals exposures on pediatric atopy. Environ Res 2011, 111(6):785-791. H a n s e n L G , Halk en S, Host A, Moller K, Osterballe O: Prediction of allergy from family history and cord blood IgE levels. A follow-up at the age of 5 years. C o r d blood IgE. IV. Pediatr Allergy Immunol 1993, 4(1):34-40. O k a d a E, Sasaki S, Saijo Y, W a s h i n o N, Miyashita C, Kobayashi S, Konishi K, Ito Y M , Ito R, Nakata A et al:Prenatal exposure to perfluorinated chemicals and relationship with allergies and infectious diseases in infants. Environ Res 2012, 112:118-125. T i m m e r m a n n C A , Budtz-Jorgensen E, Jensen T K , O s u n a CE, Petersen M S , Steuerwald U, Nielsen F, Poulsen L K , W e i h e P, Grandjean P: Association between perfluoroalkyl substance exposure and asthma and allergic disease in children as modified b y M M R vaccination. J Immunotoxicol 2017, 14(1):39-49. Qi n X D , Qian Z M , D h a r m a g e SC, Perret J, Geiger SD, R i g d o n SE, H o w a r d S, Z e n g X W , H u L W , Y a n g B Y et al: Association of perfluoroalkyl substances exposure with impaired lung function in children. Environ Res 2017, 155:15-21. 131 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 138. O ulhote Y, Sham im Z, K ielsen K, W eihe P, G randjean P, R y d er LP, H eilm ann C: Children's w hite blood cell counts in relation to developm ental exposures to m ethylm ercury and p ersistent organic pollutants. Reprod Toxicol 2017, 68:207-214. 139. Pennings JL, Jennen DG, N ygaard UC, N am ork E, H aug LS, van Loveren H, Granum B: Cord blood gene expression supports that prenatal exposure to perfluoroalkyl substances causes depressed im m une functionality in early childhood. J Immunotoxicol 2016, 13(2):173-180. 140. Goldenthal EI, Jessup DC, Geil RG, M ehring JS: Final Report, N inety Day Subacute Rhesus M onkey Toxicity Study, International Research and D evelopm ent Corporation, Study No. 137-090, N ovem ber 10, 1978, U.S. E PA A dm inistrative R ecord, AR226-0447. In.; 1978. 141. D eW itt JC, Peden-A dam s M M , K eller JM, Germ olec DR: Im m unotoxicity of perfluorinated com pounds: recent developm ents. Toxicol Pathol 2012, 40(2):300-311. 142. Bollands AD, Low e KC: Effects o f a perfluorocarbon emulsion, Fluosol-DA , on rat lym phoid tissue and im m unological com petence. Comp Biochem Physiol C 1986, 85(2):309-312. 143. Peden-A dam s M M , K eller JM , E udaly JG, B erger J, G ilkeson GS, K eil DE: Suppression o f hum oral im m unity in m ice follow ing exposure to perfluorooctane sulfonate. Toxicol Sci 2008, 104(1):144-154. 144. D ew itt JC, Copeland CB, Strynar M J, Luebke RW : Perfluorooctanoic acid-induced im m unom odulation in adult C 57B L /6J or C 57B L /6N fem ale m ice. Environ Health Perspect 2008, 116(5):644-650. 145. Guruge KS, H ikono H, Shim ada N, M urakam i K, H asegaw a J, Yeung LW , Y am anaka N, Yam ashita N: Effect of perfluorooctane sulfonate (PFOS) on influenza A virus-induced m ortality in fem ale B6C 3F1 m ice. The Journal o f toxicological sciences 2009, 34(6):687- 691. 146. Keil DE, M ehlm ann T, B utterw orth L, Peden-A dam s M M : Gestational exposure to perfluorooctane sulfonate suppresses im m une function in B 6C 3F1 m ice. Toxicol Sci 2008, 103(1):77-85. 147. Corsini E, Sangiovanni E, A vogadro A, Galbiati V, Viviani B, M arinovich M , Galli CL, Dell'A gli M, Germ olec DR: In vitro characterization o f the im m unotoxic potential of several perfluorinated com pounds (PFCs). Toxicol ApplPharmacol 2012, 258(2):248- 255. 148. U.S. Environm ental Protection Agency: Health Effects D ocum ent for Perfluorooctanoic A cid (PFOA). In. W ashington, D.C.: U.S. EPA ; 2014. 149. U.S. Environm ental Protection Agency: Health Effects D ocum ent for Perfluorooctane Sulfonate (PFOS). In. W ashington, D.C.: U.S. EPA; 2014. 150. Grandjean P, Budtz-Jorgensen E: Im m unotoxicity o f perfluorinated alkylates: calculation o f b enchm ark doses b ased on serum concentrations in children. Environ Health 2013, 12:35. 151. Chang ET, Adam i HO, Boffetta P, W edner HJ, M andel JS: A critical review of perfluorooctanoate and perfluorooctanesulfonate exposure and immunological health conditions in hum ans. CritRev Toxicol 2016, 46(4):279-331. 152. Qazi M R, N elson BD, D epierre JW, Abedi-V alugerdi M: 28-Day dietary exposure of m ice to a low total dose (7 mg/kg) o f perfluorooctanesulfonate (PFOS) alters neither the cellular compositions of the thym us and spleen nor humoral im m une responses: does the 132 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN route o f adm inistration play a pivotal role in P F O S -induced im m unotoxicity? Toxicology 2010, 267(1-3):132-139. 153. D eW itt JC, Shnyra A, B adr M Z, Loveless SE, Hoban D, Fram e SR, Cunard R, A nderson SE, M eade BJ, Peden-A dam s M M et al: Im m unotoxicity o f perfluorooctanoic acid and perfluorooctane sulfonate and the role of peroxisom e proliferator-activated receptor alpha. CritRev Toxicol 2009, 39(1):76-94. 154. Corsini E, A vogadro A, Galbiati V, dell'Agli M, M arinovich M , Galli CL, Germ olec DR: In vitro evaluation of the imm unotoxic potential of perfluorinated compounds (PFCs). Toxicol Appl Pharmacol 2011, 250(2):108-116. 155. K ielsen K, Shamim, Z., Ryder, L.P., Grandjean, P., Heilm ann, C. : Vaccination efficacy and environm ental pollution: Springer Verlag; 2016. 156. C8 Science Panel: Probable link evaluation o f pregnancy induced hypertension and preeclam psia. In. Edited by Fletcher T, Steenland K, Savitz D; 2011. 157. B uck Louis GM: Persistent environm ental pollutants and couple fecundity: an overview. Reproduction 2014, 147(4):R 97-R 104. 158. Fei C, M cL aughlin JK, L ipw orth L, O lsen J: M aternal levels o f perfluorinated chem icals and subfecundity. Hum Reprod2009, 24(5):1200-1205. 159. Velez M P, A rbuckle TE, Fraser W D: M aternal exposure to perfluorinated chem icals and reduced fecundity: the M IR E C study. Hum Reprod2015, 30(3):701-709. 160. V estergaard S, N ielsen F, A ndersson A M , H jollund N H , G randjean P, A ndersen H R, Jensen TK: Association betw een perfluorinated compounds and tim e to pregnancy in a prospective cohort o f D anish couples attem pting to conceive. Hum Reprod 2012, 27(3):873-880. 161. W hitw orth KW , H aug LS, Baird DD, B echer G, H oppin JA, Skjaerven R, Thom sen C, E ggesbo M, Travlos G, W ilson R et al: Perfluorinated com pounds and subfecundity in pregnant w om en. Epidemiology 2012, 23(2):257-263. 162. Spira N, Spira A, Schwartz D: Fertility o f couples follow ing cessation o f contraception. J Biosoc Sci 1985, 17(3):281-290. 163. Lopez-Espinosa M J, Fletcher T, Arm strong B, G enser B, D hatariya K, M ondal D, D ucatm an A, Leonardi G: A ssociation o f Perfluorooctanoic Acid (PFOA) and Perfluorooctane Sulfonate (PFOS) with age of puberty among children living near a chem ical plant. Environ Sci Technol 2011, 45(19):8160-8166. 164. Lopez-Espinosa M J, M ondal D, Arm strong BG, Eskenazi B, Fletcher T: Perfluoroalkyl Substances, Sex Horm ones, and Insulin-like Growth Factor-1 at 6-9 Years o f Age: A C ross-Sectional A nalysis w ithin the C8 H ealth Project. Environ Health Perspect 2016, 124(8):1269-1275. 165. Christensen KY, M aisonet M , Rubin C, H olm es A, Calafat AM , K ato K, Flanders W D, H eron J, M cG eehin M A, M arcus M: Exposure to polyfluoroalkyl chem icals during pregnancy is not associated w ith offspring age at m enarche in a contem porary British cohort. Environ Int 2011, 37(1):129-135. 166. K ristensen SL, Ram lau-H ansen CH, Ernst E, Olsen SF, B onde JP, Vested A, H alldorsson TI, B echer G, H aug LS, Toft G: Long-term effects o f prenatal exposure to perfluoroalkyl substances on fem ale reproduction. Hum Reprod2013, 28(12):3337-3348. 167. Z h o u W , Z hang L, T ong C, F ang F, Z hao S, T ian Y, T ao Y, Z hang J, Shanghai B irth C ohort S: P lasm a P erfluoroalkyl and P olyfluoroalkyl Substances C oncentration and 133 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. Menstrual Cycle Characteristics in Preconception W o m e n . Environ Health Perspect 2017, 125(6):067012. Rattan S, Z h o u C, Chiang C, M a h a l i n g a m S, B r e h m E, Flaws JA: Exposure to endocrine disruptors during adulthood: consequences for female fertility. JEndocrinol 2017, 233(3):R109-R129. K n o x SS, Jackson T, Javins B, Frisbee SJ, Shankar A, D u c a t m a n A M : Implications of early me no pa us e in w o m e n exposed to perfluorocarbons. J Clin EndocrinolMetab 2011, 96(6):1747-1753. Toft G, Jonsson B A , Lindh C H , G i w e r c m a n A, Spano M , Heederik D, Lenters V, V e r m eu le n R, Rylander L, Pedersen H S et al: Exposure to perfluorinated c o m p o u n d s and h u m a n s e m e n quality in arctic and European populations. Hum Reprod 2012, 27(8):2532- 2540. R a y m e r JH, Michael L C , Studabaker W B , Olsen G W , Sloan CS, Wilcosky T, W a l m e r D K : Concentrations of perfluorooctane sulfonate ( P F O S ) and perfluorooctanoate ( P F O A ) and their associations with h u m a n s e m e n quality measurements. Reprod Toxicol 2012, 33(4):419-427. Joensen U N , Bossi R, Leffers H, Jensen A A , S kakkebaek N E , Jorgensen N: D o perfluoroalkyl c o m p o u n d s impair h u m a n s e m e n quality? Environ Health Perspect 2009, 117(6):923-927. Vested A, R a m l a u - H a n s e n C H , Olsen SF, B o n d e JP, Kristensen SL, Halldorsson TI, Becher G, H a u g LS, Ernst E H , Toft G: Associations of in utero exposure to perfluorinated alkyl acids with h u m a n s e m e n quality and reproductive h o r m o n e s in adult men. Environ Health Perspect 2013, 121(4):453-458, 458e451-455. B a c h C C , Vested A, Jorgensen K T , B o n d e JP, Henriksen TB, Toft G: Perfluoroalkyl and polyfluoroalkyl substances and measures of h u m a n fertility: a systematic review. Crit Rev Toxicol 2016, 46(9):735-755. D a r r o w L A , H o w a r d s PP, Winquist A, Steenland K: P F O A and P F O S serum levels and miscarriage risk. Epidemiology 2014, 25(4):505-512. Jensen T K , Andersen L B , K y h l H B , Nielsen F, Christesen H T , Grandjean P: Association between Perfluorinated C o m p o u n d Exposure and Miscarriage in Danish Pregnant W o m e n . PLoS One 2015, 10(4):e0123496. Stein C R , Savitz D A : S e r u m perfluorinated c o m p o u n d concentration and attention deficit/hyperactivity disorder in children 5-18 years of age. Environ Health Perspect 2011, 119(10):1466-1471. D a r r o w L A , Stein C R , Steenland K: S e r u m perfluorooctanoic acid and perfluorooctane sulfonate concentrations in relation to birth outcomes in the M i d - O h i o Valley, 2 0 0 5 2010. Environ Health Perspect 2013, 121(10):1207-1213. Starling AP , Engel S M , Richardson D B , Baird D D , H a u g LS, Stuebe A M , Klungsoyr K, H a r m o n Q, Becher G, T h o m s e n C et al: Perfluoroalkyl substances during pregnancy and validated preeclampsia a m o n g nulliparous w o m e n in the N o r w e g i a n M o t h e r and Child Cohort Study. Am J Epidemiol 2014, 179(7):824-833. Apelberg BJ, Witter FR, H e r b st ma n JB, Calafat A M , Halden R U , N e e d h a m LL, G o l d m a n L R : Co r d serum concentrations of perfluorooctane sulfonate ( P F O S ) and perfluorooctanoate ( P F O A ) in relation to weight and size at birth. Environ Health Perspect 2007, 115(11):1670-1676. 134 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 181. Fei C, M cL aughlin JK, Tarone RE, O lsen J: Perfluorinated chem icals and fetal growth: a study w ithin the D anish N ational B irth Cohort. Environ Health Perspect 2007, 115(11):1677-1682. 182. Olsen GW , B utenhoff JL, Zobel LR: Perfluoroalkyl chem icals and hum an fetal development: an epidem iologic review w ith clinical and toxicological perspectives. Reprod Toxicol 2009, 27(3-4):212-230. 183. Savitz DA: G uest editorial: biom arkers o f perfluorinated chem icals and birth weight. Environ Health Perspect 2007, 115(11):A 528-529. 184. G rieger JA, Clifton VL: A review o f the im pact o f dietary intakes in hum an pregnancy on infant birthw eight. Nutrients 2014, 7(1):153-178. 185. Lind DV, Priskorn L, Lassen TH, N ielsen F, Kyhl HB, K ristensen DM , Christesen HT, Jorgensen JS, Grandjean P, Jensen TK: Prenatal exposure to perfluoroalkyl substances and anogenital distance at 3 m onths o f age in a D anish m other-child cohort. Reprod Toxicol 2017, 68:200-206. 186. W hitw orth KW , H aug LS, Baird DD, B echer G, H oppin JA, Skjaerven R, Thom sen C, E ggesbo M, Travlos G, W ilson R et al: Perfluorinated com pounds in relation to birth w eig h t in the N o rw egian M o th er and C hild C ohort Study. Am J Epidemiol 2012, 175(12):1209-1216. 187. M in ato y a M , Itoh S, M iyashita C, A raki A, Sasaki S, M iu ra R, G oudarzi H, Iw asaki Y, Kishi R: Association of prenatal exposure to perfluoroalkyl substances with cord blood adipokines and birth size: The H okkaido Study on environm ent and children's health. Environ Res 2017, 156:175-182. 188. M aisonet M, Terrell M L, M cG eehin M A, Christensen KY, H olm es A, Calafat AM, M arcus M: M aternal concentrations of polyfluoroalkyl compounds during pregnancy and fetal and postnatal grow th in B ritish girls. Environ Health Perspect 2012, 120(10):1432- 1437. 189. K arlsen M , Grandjean P, W eihe P, Steuerwald U, O ulhote Y, Valvi D: Early-life exposures to persistent organic pollutants in relation to overw eight in preschool children. Reprod Toxicol 2017, 68:145-153. 190. Das KP, Grey BE, Zehr RD, W ood CR, B utenhoff JL, Chang SC, Ehresm an DJ, Tan YM , Lau C: Effects o f perfluorobutyrate exposure during pregnancy in the mouse. Toxicol Sci 2008, 105(1):173-181. 191. C8 Science Panel: Probable link evaluation o f birth defects. In. Edited by F letcher T, Steenland K, Savitz D; 2011. 192. Gore AC, Chappell VA, Fenton SE, Flaw s JA, N adal A, Prins GS, Toppari J, Zoeller RT: EDC-2: The Endocrine Society's Second Scientific Statem ent on Endocrine-D isrupting C hem icals. Endocr Rev 2015, 36(6):E1-E150. 193. M aisonet M, Calafat AM , M arcus M, Jaakkola JJ, Lashen H: Prenatal Exposure to Perfluoroalkyl A cids and Serum T estosterone C oncentrations at 15 Y ears o f A ge in Fem ale A L SPA C Study Participants. Environ Health Perspect 2015, 123(12):1325-1330. 194. Tsai M S, Lin CY, Lin CC, Chen M H, H su SH, Chien KL, Sung FC, Chen PC, Su TC: Association betw een perfluoroalkyl substances and reproductive hormones in adolescents and young adults. Internationaljournal o f hygiene and environmental health 2015, 218(5):437-443. 135 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 195. Fei C, M cL aughlin JK, L ipw orth L, O lsen J: M aternal concentrations o f perfluorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA) and duration of breastfeeding. Scand J Work Environ Health 2010, 36(5):413-421. 196. R om ano M E, Xu Y, Calafat AM, Yolton K, Chen A, W ebster GM, E liot M N, H ow ard CR, Lanphear BP, Braun JM: M aternal serum perfluoroalkyl substances during pregnancy and duration o f breastfeeding. Environ Res 2016, 149:239-246. 197. Tim m erm ann CA, Budtz-Jorgensen E, Petersen M S, W eihe P, Steuerwald U, N ielsen F, Jensen TK, Grandjean P: Shorter duration of breastfeeding at elevated exposures to perfluoroalkyl substances. Reprod Toxicol 2017, 68:164-170. 198. Biegel LB, Liu RC, H urtt M E, Cook JC: Effects o f am m onium perfluorooctanoate on L eydig cell function: in vitro, in vivo, and ex vivo studies. Toxicol Appl Pharmacol 1995, 134(1):18-25. 199. G orrochategui E, P erez-A lbaladejo E, C asas J, L acorte S, P orte C: P erfluorinated chemicals: differential toxicity, inhibition of arom atase activity and alteration of cellular lipids in hum an placental cells. Toxicol Appl Pharmacol 2014, 277(2):124-130. 200. Kjeldsen LS, Bonefeld-Jorgensen EC: Perfluorinated compounds affect the function of sex horm one receptors. Environmental science andpollution research international 2013, 20(11):8031-8044. 201. Dankers AC, Roelofs MJ, Piersma AH, Sweep FC, Russel FG, van den Berg M, van Duursen M B, M asereeuw R: Endocrine disruptors differentially target ATP-binding cassette transporters in the blood-testis barrier and affect Leydig cell testosterone secretion in vitro. ToxicolSci 2013, 136(2):382-391. 202. M aras M , V anparys C, M uylle F, R obbens J, B erger U, B arber JL, B lust R, D e Coen W: Estrogen-like properties o f fluorotelom er alcohols as revealed by m cf-7 breast cancer cell proliferation. Environ Health Perspect 2006, 114(1):100-105. 203. B enninghoff AD, Bisson W H, Koch DC, Ehresm an DJ, Kolluri SK, W illiams DE: Estrogen-like activity o f perfluoroalkyl acids in vivo and interaction w ith hum an and rainbow tro u t estrogen receptors in vitro. Toxicol Sci 2011, 120(1):42-58. 204. Z hao B, H u GX, C hu Y, Jin X, G ong S, A kingbem i B T , Z hang Z, Z irkin B R , G e RS: Inhibition o f hum an and rat 3beta-hydroxysteroid dehydrogenase and 17beta- hydroxysteroid dehydrogenase 3 activities by perfluoroalkylated substances. Chem Biol Interact 2010, 188(1):38-43. 205. M acon M B, V illanueva LR, Tatum -G ibbs K, Z ehr RD, Strynar M J, Stanko JP, W hite SS, H elfant L, Fenton SE: Prenatal perfluorooctanoic acid exposure in CD-1 mice: low-dose developm ental effects and internal dosim etry. Toxicol Sci 2011, 122(1):134-145. 206. Post GB, Louis JB, Cooper KR, Boros-Russo BJ, Lippincott RL: Occurrence and potential significance o f perfluorooctanoic acid (PFOA) detected in N ew Jersey public drinking w ater system s. Environ Sci Technol 2009, 43(12):4547-4554. 207. Gilliland FD, M andel JS: Serum perfluorooctanoic acid and hepatic enzymes, lipoproteins, and cholesterol: a study o f occupationally exposed men. Am JIndM ed 1996, 29(5):560-568. 208. Grandjean P, Gronlund C, K jaer IM, Jensen TK, Sorensen N, Andersson AM, Juul A, Skakkebaek NE, Budtz-Jorgensen E, W eihe P: Reproductive hormone profile and pubertal developm ent in 14-year-old boys prenatally exposed to polychlorinated biphenyls. Reprod Toxicol 2012, 34(4):498-503. 136 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 209. W ohlfahrt-V eje C, Andersen HR, Jensen TK, Grandjean P, Skakkebaek NE, M ain KM: Smaller genitals at school age in boys w hose m others were exposed to non-persistent pesticides in early pregnancy. Int J Androl 2012, 35(3):265-272. 210. W ohlfahrt-V eje C, Andersen HR, Schmidt IM, Aksglaede L, Sorensen K, Juul A, Jensen TK, Grandjean P, Skakkebaek NE, M ain KM: Early breast developm ent in girls after prenatal exposure to non-persistent pesticides. Int J Androl 2012, 35(3):273-282. 211. Araujo AB, D ixon JM, Suarez EA, M urad M H, Guey LT, W ittert GA: Clinical review: Endogenous testosterone and mortality in men: a systematic review and meta-analysis. J Clin EndocrinolMetab 2011, 96(10):3007-3019. 212. Yu W G, Liu W , Jin YH, Liu XH, W ang FQ, Liu L, N akayam a SF: Prenatal and postnatal im pact o f perfluorooctane sulfonate (PFOS) on rat development: a cross-foster study on chem ical burden and th y roid horm one system . Environ Sci Technol 2009, 43(21):8416- 8422. 213. C8 Science Panel: Probable link evaluation o f thyroid disease. In. Edited by F letcher T, Steenland K, Savitz D; 2012. 214. Porterfield SP: Thyroidal dysfunction and environm ental chem icals--potential im pact on brain developm ent. Environ Health Perspect 2000, 108 Suppl 3:433-438. 215. Z o eller R T , R o v et J: T im ing o f thyroid horm one action in the developing brain: clinical observations and experim ental findings. Journal of neuroendocrinology 2004, 16(10):809-818. 216. C rofton KM : Thyroid disrupting chem icals: m echanism s and m ixtures. Int J Androl 2008, 31(2):209-223. 217. Olsen GW , Burris JM, Burlew M M , M andel JH: Epidem iologic assessm ent o f worker serum perfluorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA) concentrations and m edical surveillance exam inations. J Occup Environ Med 2003, 45(3):260-270. 218. K nox SS, Jackson T, Frisbee SJ, Javins B, D ucatm an AM: Perfluorocarbon exposure, gender and thyroid function in the C8 H ealth Project. The Journal o f toxicological sciences 2011, 36(4):403-410. 219. W inquist A, Steenland K: Perfluorooctanoic acid exposure and thyroid disease in com m unity and w o rk er cohorts. Epidemiology 2014, 25(2):255-264. 220. L opez-E spinosa M J, M ondal D, A rm strong B, B loom M S, F letcher T: T hyroid function and perfluoroalkyl acids in children living near a chem ical plant. Environ Health Perspect 2012, 120(7):1036-1041. 221. M elzer D, Rice N, Depledge MH, Henley W E, Galloway TS: Association betw een serum perfluorooctanoic acid (PFOA) and thyroid disease in the U.S. National H ealth and N utrition E xam ination Survey. Environ Health Perspect 2010, 118(5):686-692. 222. W en LL, Lin LY, Su TC, Chen PC, Lin CY: Association betw een serum perfluorinated chem icals and thyroid function in U.S. adults: the National H ealth and N utrition E xam ination Survey 2007-2010. J Clin Endocrinol Metab 2013, 98(9):E 1456-1464. 223. L eung AM , K orevaar TI, Peeters RP, Z oeller RT, K ohrle J, D untas LH , B rent GA, D em eneix BA: Exposure to Thyroid-Disrupting Chemicals: A Transatlantic Call for A ction. Thyroid 2016, 26(4):479-480. 224. B utenhoff JL, Bjork JA, Chang SC, Ehresm an DJ, Parker GA, Das K, Lau C, Lieder PH, van Otterdijk FM, W allace KB: Toxicological evaluation of ammonium perfluorobutyrate in rats: tw enty-eight-day and ninety-day oral gavage studies. Reprod Toxicol 2012, 33(4):513-530. 137 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 225. B utenhoff JL, Kennedy GL, Jr., Chang SC, Olsen GW: Chronic dietary toxicity and carcinogenicity study w ith amm onium perfluorooctanoate in Sprague-Dawley rats. Toxicology 2012, 298(1-3):1-13. 226. Chang SC, Thibodeaux JR, Eastvold M L, Ehresm an DJ, B jork JA, Froehlich JW , Lau C, Singh RJ, W allace KB, B utenhoff JL: Thyroid horm one status and pituitary function in adult rats given oral doses o f perfluorooctanesulfonate (PFO S). Toxicology 2008, 243(3):330-339. 227. C hang S, A llen BC , A ndres K L , E hresm an D J, F alvo R, P ro v en ch er A, O lsen GW , B utenhoff JL: Evaluation o f Serum Lipid, Thyroid, and H epatic Clinical Chem istries in Association W ith Serum Perfluorooctanesulfonate (PFOS) in Cynomolgus M onkeys A fter Oral D osing W ith Potassium PFO S. ToxicolSci 2017, 156(2):387-401. 228. Taylor PN, Okosieme OE, M urphy R, Hales C, Chiusano E, M aina A, Joomun M, Bestw ick JP, Smyth P, Paradice R et al: M aternal perchlorate levels in w om en with borderline thyroid function during pregnancy and the cognitive developm ent of their offspring; D ata from the C ontrolled A ntenatal T hyroid Study. J Clin EndocrinolMetab 2014:jc20141901. 229. L am D W , L eR oith D: T he w orldw ide diabetes epidem ic. Curr Opin Endocrinol Diabetes Obes 2012, 19(2):93-96. 230. M itchell NS, Catenacci VA, W yatt HR, Hill JO: Obesity: overview o f an epidemic. Psychiatr Clin North Am 2011, 34(4):717-732. 231. C8 Science Panel: Probable link evaluation o f diabetes. In. E dited by F letcher T, Steenland K, Savitz D; 2012. 232. Kuo CC, M oon K, Thayer KA, N avas-A cien A: Environmental chemicals and type 2 diabetes: an updated system atic review o f the epidem iologic evidence. Curr Diab Rep 2013, 13(6):831-849. 233. Liu Y, Peterson KE: M aternal Exposure to Synthetic Chem icals and Obesity in the O ffspring: R ecent Findings. Curr Environ Health Rep 2015, 2(4):339-347. 234. M acN eil J, Steenland N K , Shankar A, D ucatm an A: A cross-sectional analysis o f type II diabetes in a com m unity w ith exposure to perfluorooctanoic acid (PFO A ). Environ Res 2009, 109(8):997-1003. 235. A bbott BD , W ood CR, W atkins AM , Tatum -G ibbs K, D as KP, Lau C: Effects o f perfluorooctanoic acid (PFOA) on expression o f peroxisom e proliferator-activated receptors (PPAR) and nuclear receptor-regulated genes in fetal and postnatal CD-1 m ouse tissues. Reprod Toxicol 2012, 33(4):491-505. 236. Lin CY, Chen PC, Lin YC, Lin LY: Association among serum perfluoroalkyl chemicals, glucose hom eostasis, and m etabolic syndrom e in adolescents and adults. Diabetes Care 2009, 32(4):702-707. 237. Zong G, Grandjean P, W ang X, Sun Q: Lactation history, serum concentrations of persistent organic pollutants, and m aternal risk o f diabetes. Environ Res 2016, 150:282 288. 238. Timm erm ann CA, Rossing LI, Grantved A, Ried-Larsen M, Dalgard C, Andersen LB, G randjean P, N ielsen F, Svendsen KD , Scheike T et al: A diposity and glycem ic control in children exposed to perfluorinated com pounds. J Clin Endocrinol Metab 2014, 99(4):E608-614. 138 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 239. A ndersen CS, Fei C, G am borg M , N o h r EA , Sorensen TI, O lsen J: P renatal exposures to perfluorinated chem icals and anthropom etry at 7 years o f age. Am J Epidemiol 2013, 178(6):921-927. 240. H alldorsson TI, R ytter D, H aug LS, B ech BH , D anielsen I, B echer G, H enriksen TB, Olsen SF: Prenatal exposure to perfluorooctanoate and risk o f overw eight at 20 years of age: a prospective cohort study. Environ Health Perspect 2012, 120(5):668-673. 241. M ora AM, Oken E, Rifas-Shim an SL, W ebster TF, Gillman M W , Calafat AM, Ye X, Sagiv SK: Prenatal Exposure to Perfluoroalkyl Substances and Adiposity in Early and M id-C hildhood. Environ Health Perspect 2017, 125(3):467-473. 242. H oyer BB, Ram lau-Hansen CH, Vrijheid M, Valvi D, Pedersen HS, Zviezdai V, Jonsson BA, Lindh CH, Bonde JP, Toft G: Anthropom etry in 5- to 9-Year-Old Greenlandic and Ukrainian Children in Relation to Prenatal Exposure to Perfluorinated Alkyl Substances. Environ Health Perspect 2015, 123(8):841-846. 243. Braun JM: Early-life exposure to EDCs: role in childhood obesity and neurodevelopm ent. Nat Rev Endocrinol 2017, 13(3):161-173. 244. Audouze K, Grandjean P: Application of computational systems biology to explore environm ental toxicity hazards. Environ Health Perspect 2011, 119(12):1754-1759. 245. H oltcam p W : O besogens: an environm ental link to obesity. Environ Health Perspect 2012, 120(2):a62-68. 246. L v Z, Li G, Li Y, Y ing C, C hen J, C hen T, W ei J, L in Y, Jiang Y, W ang Y et al: G lucose and lipid hom eostasis in adult rat is im paired by early-life exposure to perfluorooctane sulfonate. Environ Toxicol2013, 28(9):532-542. 247. Scharmach E, Buhrke T, Lichtenstein D, Lam pen A: Perfluorooctanoic acid affects the activity o f the hepatocyte n u clear factor 4 alpha (H N F4alpha). Toxicology letters 2012, 212(2):106-112. 248. Barouki R, Gluckm an PD, Grandjean P, H anson M, Heindel JJ: D evelopm ental origins of non-com m unicable disease: im plications fo r research and public health. Environ Health 2012, 11(1):42. 249. Ogden CL, Carroll M D, K it BK, Flegal KM: Prevalence o f childhood and adult obesity in the U nited States, 2011-2012. JAMA : thejournal o f the American Medical Association 2014, 311(8):806-814. 250. Gregg EW , Shaw JE: Global Health Effects o f O verw eight and Obesity. N Engl J Med 2017, 377(1):80-81. 251. W eiss B: The intersection o f neurotoxicology and endocrine disruption. Neurotoxicology 2012, 33(6):1410-1419. 252. M ariussen E: N eurotoxic effects o f perfluoroalkylated compounds: m echanism s o f action and environm ental relevance. Arch Toxicol 2012, 86(9):1349-1367. 253. G randjean P, L andrigan PJ: D evelopm ental neurotoxicity o f industrial chem icals. Lancet 2006, 368(9553):2167-2178. 254. Fei C, M cL aughlin JK, L ipw orth L, O lsen J: Prenatal exposure to perfluorooctanoate (PFOA) and perfluorooctanesulfonate (PFOS) and maternally reported developmental m ilestones in infancy. Environ Health Perspect 2008, 116(10):1391-1395. 255. H offm an K, W ebster TF, W eisskopf M G, W einberg J, Vieira VM: Exposure to polyfluoroalkyl chem icals and attention deficit/hyperactivity disorder in U.S. children 12 15 years o f age. Environ Health Perspect 2010, 118(12):1762-1767. 139 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 256. Stein CR, Savitz DA, Bellinger DC: Perfluorooctanoate exposure in a highly exposed comm unity and parent and teacher reports o f behaviour in 6-12-year-old children. Paediatric andperinatal epidemiology 2014, 28(2):146-156. 257. Gump BB, W u Q, D um as AK, K annan K: Perfluorochem ical (PFC) exposure in children: associations w ith im paired response inhibition. Environ Sci Technol 2011, 45(19):8151- 8159. 258. Vuong AM, Yolton K, W ebster GM, Sjodin A, Calafat AM , Braun JM, Dietrich KN, Lanphear BP, Chen A: Prenatal polybrom inated diphenyl ether and perfluoroalkyl substance exposures and executive function in school-age children. Environ Res 2016, 147:556-564. 259. Oulhote Y, Steuerwald U, Debes F, W eihe P, Grandjean P: Behavioral difficulties in 7- year old children in relation to developm ental exposure to perfluorinated alkyl substances. Environ Int 2016, 97:237-245. 260. Liew Z, Ritz B, Bonefeld-Jorgensen EC, Henriksen TB, N ohr EA, Bech BH, Fei C, Bossi R, von E hrenstein OS, Streja E et al: Prenatal exposure to perfluoroalkyl substances and the risk o f congenital cerebral palsy in children. Am JEpidemiol 2014, 180(6):574-581. 261. Johansson N, Fredriksson A, Eriksson P: Neonatal exposure to perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA) causes neurobehavioural defects in adult m ice. Neurotoxicology 2008, 29(1):160-169. 262. Grandjean P: Only one chance. H ow Environm ental Pollution Im pairs Brain D evelopm ent - and H ow to Protect the Brains o f the N ext Generation. N ew York: Oxford University Press; 2013. 263. C8 Science Panel: Probable Link Evaluation for heart disease (including high blood pressure, high cholesterol, coronary artery disease). In. E dited by F letcher T, Steenland K, Savitz D; 2012. 264. Olsen GW, Burris JM, M andel JH, Zobel LR: Serum perfluorooctane sulfonate and hepatic and lipid clinical chemistry tests in fluorochem ical production employees. J Occup Environ Med 1999, 41(9):799-806. 265. G allo V, Leonardi G, G enser B, L opez-Espinosa M J, Frisbee SJ, K arlsson L, D ucatm an AM , Fletcher T: Serum perfluorooctanoate (PFOA ) and perfluorooctane sulfonate (PFOS) concentrations and liver function biom arkers in a population w ith elevated PFO A exposure. Environ Health Perspect 2012, 120(5):655-660. 266. D arrow LA, Groth AC, W inquist A, Shin HM , Bartell SM, Steenland K: M odeled Perfluorooctanoic Acid (PFOA) Exposure and Liver Function in a M id-O hio Valley C om m unity. Environ Health Perspect 2016, 124(8):1227-1233. 267. Lin CY, Lin LY, Chiang CK, W ang W J, Su YN, Hung KY, Chen PC: Investigation of the associations betw een low-dose serum perfluorinated chem icals and liver enzym es in U S adults. The Americanjournal o fgastroenterology 2010, 105(6):1354-1363. 268. Sakr CJ, Kreckm ann KH, Green JW, Gillies PJ, Reynolds JL, Leonard RC: Cross sectional study o f lipids and liver enzymes related to a serum biom arker o f exposure (ammonium perfluorooctanoate or APFO) as part of a general health survey in a cohort of occupationally exposed w orkers. J Occup Environ M ed 2007, 49(10):1086-1096. 269. Sakr CJ, Leonard RC, K reckm ann KH, Slade M D, Cullen MR: Longitudinal study of serum lipids and liver enzymes in workers w ith occupational exposure to amm onium perfluorooctanoate. J Occup Environ Med 2007, 49(8):872-879. 140 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 270. Seacat AM , Thom ford PJ, Hansen KJ, Olsen GW , Case M T, B utenhoff JL: Subchronic toxicity studies on perfluorooctanesulfonate potassium salt in cynomolgus monkeys. Toxicol Sci 2002, 68(1):249-264. 271. Bjork JA, B utenhoff JL, W allace KB: M ultiplicity of nuclear receptor activation by P F O A and P FO S in prim ary hum an and rodent hepatocytes. Toxicology 2011, 288(1- 3):8-17. 272. W an HT, Zhao YG, W ei X, Hui KY, Giesy JP, W ong CK: PFO S-induced hepatic steatosis, the m echanistic actions on b eta-oxidation and lipid transport. Biochim Biophys Acta 2012, 1820(7):1092-1101. 273. Ikeda T, A iba K, Fukuda K, Tanaka M: The induction o f peroxisom e proliferation in rat liver by perfluorinated fatty acids, m etabolically inert derivatives o f fatty acids. J Biochem 1985, 98(2):475-482. 274. Kennedy GL, B utenhoff JL, Olsen GW, O 'C onnor JC, Seacat AM , Perkins RG, Biegel LB , M urphy SR, F arrar D G : T he toxicology o f perfluorooctanoate. Crit Rev Toxicol 2004, 34(4):351-384. 275. Schwim m er JB, D eutsch R, Kahen T, Lavine JE, Stanley C, Behling C: Prevalence of fatty liver in children and adolescents. Pediatrics 2006, 118(4):1388-1393. 276. Pacifico L, Poggiogalle E, Cantisani V, M enichini G, Ricci P, Ferraro F, Chiesa C: P ediatric nonalcoholic fatty liver disease: A clinical and laboratory challenge. World J Hepatol 2010, 2(7):275-288. 277. M ukherjee D: Perfluorooctanoic Acid Exposure and Cardiovascular Disease: Potential Role and Preventive M easures Com m ent on "Perfluorooctanoic Acid and Cardiovascular D isease in U S A dults". Arch Intern Med 2012:1-2. 278. Olsen GW , Zobel LR: A ssessm ent o f lipid, hepatic, and thyroid param eters w ith serum perfluorooctanoate (PFO A ) concentrations in fluorochem ical production workers. Int Arch Occup Environ Health 2007, 81(2):231-246. 279. C osta G, Sartori S, C onsonni D: T hirty years o f m edical surveillance in perfluooctanoic acid production w orkers. J Occup Environ Med 2009, 51(3):364-372. 280. Shankar A, X iao J, D ucatm an A: Perfluorooctanoic A cid and Cardiovascular D isease in U S A dults. Arch Intern Med 2012:1-7. 281. Frisbee SJ, Shankar A, K nox SS, Steenland K, Savitz DA, Fletcher T, D ucatm an AM: Perfluorooctanoic acid, perfluorooctanesulfonate, and serum lipids in children and adolescents: results from the C8 H ealth Project. Arch Pediatr Adolesc Med 2010, 164(9):860-869. 282. Steenland K, T inker S, F risbee S, D ucatm an A, V accarino V: A ssociation o f perfluorooctanoic acid and perfluorooctane sulfonate w ith serum lipids among adults living near a chem ical plant. Am J Epidemiol 2009, 170(10):1268-1278. 283. Nelson JW, Hatch EE, W ebster TF: Exposure to polyfluoroalkyl chemicals and cholesterol, body w eight, and insulin resistance in the general U .S. population. Environ Health Perspect 2010, 118(2):197-202. 284. Kerger BD, Copeland TL, DeCaprio AP: Tenuous dose-response correlations for com m on disease states: case study o f cholesterol and perfluorooctanoate/sulfonate (PFO A /P FO S ) in the C8 H ealth Project. Drug Chem Toxicol 2011, 34(4):396-404. 285. A lexander BH, Olsen GW , Burris JM, M andel JH, M andel JS: M ortality o f em ployees of a perfluorooctanesulphonyl fluoride m anufacturing facility. Occup Environ M ed 2003, 60(10):722-729. 141 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 286. M in JY, Lee KJ, Park JB, M in KB: Perfluorooctanoic acid exposure is associated with elevated hom ocysteine and hypertension in U S adults. Occup Environ Med 2012, 69(9):658-662. 287. Shankar A, Xiao J, Ducatm an A: Perfluoroalkyl chem icals and chronic kidney disease in U S adults. Am J Epidemiol 2011, 174(8):893-900. 288. Steenland K, T inker S, Shankar A, D ucatm an A: A ssociation o f perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS) with uric acid am ong adults with elevated com m unity exposure to PFO A . Environ Health Perspect 2010, 118(2):229-233. 289. U.S. Environm ental Protection Agency.: Perfluorooctanoic A cid (PFOA ) and Fluorinated Telomers. 2006. 290. A lexander BH, Olsen GW : Bladder cancer in perfluorooctanesulfonyl fluoride m anufacturing w orkers. Ann Epidemiol 2007, 17(6):471-478. 291. Steenland K, W oskie S: C ohort m ortality study o f w orkers exposed to perfluorooctanoic acid. Am J Epidemiol 2012, 176(10):909-917. 292. M innesota Cancer Surveillance System: Cancer incidence in D akota and W ashington counties. In: MCSSEpidemiology Report 2007:1. St. Paul, M N: M innesota C ancer Surveillance System; 2007. 293. V ieira VM , H offm an K, Shin H M , W einberg JM , W ebster TF, F letcher T: Perfluorooctanoic acid exposure and cancer outcom es in a contam inated community: a geographic analysis. Environ Health Perspect 2013, 121(3):318-323. 294. Barry V, W inquist A, Steenland K: Perfluorooctanoic acid (PFOA) exposures and incident cancers am ong adults living near a chem ical plant. Environ Health Perspect 2013, 121(11-12):1313-1318. 295. Grandjean P, Olsen JH, Jensen OM, Juel K: Cancer incidence and m ortality in workers exposed to fluoride. J Natl Cancer Inst 1992, 84(24):1903-1909. 296. C8 Science Panel: Status Report: PFO A and A dult Thyroid D isease in the m id-Ohio Valley. In. E dited by Fletcher T, Steenland K, Savitz D; 2011. 297. Institute o f M edicine (U.S.). Com m ittee on B reast Cancer and the Environm ent: The Scientific Evidence Research M ethodology and Future Directions.: Breast cancer and the environm ent : a life course approach. W ashington, DC: National Academ ies Press; 2012. 298. Science Advisory Board: SAB Review of E PA 's D raft Risk Assessm ent of Potential H um an H ealth E ffects A ssociated w ith P FO A and Its Salts. In. W ashington, DC: U.S. Environmental Protection Agency; 2006. 299. Bonefeld-Jorgensen EC, Long M, Bossi R, Ayotte P, Asm und G, K ruger T, Ghisari M, M ulvad G, K ern P, N zulum iki P et al: P erfluorinated com pounds are related to breast cancer risk in G reenlandic Inuit: a case control study. Environ Health 2011, 10:88. 300. W ielsoe M , K ern P, Bonefeld-Jorgensen EC: Serum levels o f environm ental pollutants is a risk factor fo r b reast cancer in Inuit: a case control study. Environ Health 2017, 16(1):56. 301. Biegel LB, H urtt M E, Fram e SR, O 'C onnor JC, Cook JC: M echanism s o f extrahepatic tum or induction by peroxisom e proliferators in m ale CD rats. ToxicolSci 2001, 60(1):44- 55. 302. G lynn A, B erg er U , B ig n ert A, U llah S, A une M , L ignell S, D arnerud PO : P erfluorinated alkyl acids in blood serum from prim iparous w om en in Sweden: serial sampling during pregnancy and nursing, and tem poral trends 1996-2010. Environ Sci Technol 2012, 46(16):9071-9079. 142 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 303. B irnbaum LS, G randjean P: A lternatives to PFA Ss: P erspectives on the Science. Environ Health Perspect 2015, 123(5):A 104-105. 304. Bow m an JS: Response to "Com m ent on 'Fluorotechnology Is Critical to M odern Life: T he F luoroC ouncil C ounterpoint to the M adrid Statem ent'". Environ Health Perspect 2015, 123(7):A170-171. 305. B ogdanska J, Sundstrom M , B ergstrom U, B org D, A bedi-V alugerdi M , B ergm an A, D eP ierre J, N obel S: T issue distribution o f 35S -labelled perfluorobutanesulfonic acid in adult m ice follow ing dietary exposure fo r 1-5 days. Chemosphere 2014, 98:28-36. 306. M innesota Departm ent of Health: Health Based Guidance for Groundwater: Perfluorohexane sulfonate. In. St.Paul, M N: H ealth R isk A ssessm ent Unit, Environm ental Health Division, M innesota Departm ent of Health; 2009. 307. Forem an JE, Chang SC, Ehresm an DJ, B utenhoff JL, Anderson CR, Palkar PS, Kang BH, Gonzalez FJ, Peters JM: Differential hepatic effects of perfluorobutyrate m ediated by m ouse and hum an PPA R-alpha. ToxicolSci 2009, 110(1):204-211. 308. B utenhoff JL, Chang SC, Ehresm an DJ, York RG: Evaluation of potential reproductive and developmental toxicity of potassium perfluorohexanesulfonate in Sprague Dawley rats. Reprod Toxicol 2009, 27(3-4):331-341. 309. Tao L, Kannan K, W ong CM, A rcaro KF, B utenhoff JL: Perfluorinated compounds in hum an m ilk from M assachusetts, U .S.A . Environ Sci Technol 2008, 42(8):3096-3101. 310. Gwinn MR, Axelrad DA, Bahadori T, Bussard D, Cascio W E, D eener K, Dix D, Thomas RS, K avlock RJ, Burke TA: Chemical R isk Assessment: Traditional vs Public Health Perspectives. Am J Public Health 2017, 107(7):1032-1039. 311. D ong Z, B ahar M M , Jit J, K ennedy B, Priestly B, N g J, Lam b D, Liu Y, D uan L, N aidu R: Issues raised by the reference doses for perfluorooctane sulfonate and perfluorooctanoic acid. Environ Int 2017, 105:86-94. 312. R ainieri S, C onlledo N , L angerholc T, M adorran E, Sala M , B arran co A: T oxic effects o f perfluorinated com pounds at hum an cellular level and on a m odel vertebrate. Food Chem Toxicol 2017, 104:14-25. 313. D esforges JP, L evin M , Jasperse L, D e G uise S, E ulaers I, L etch er R J, A cquarone M , N ordoy E, F olkow L, H am m er Jensen T et al: E ffects o f polar bear and killer w hale derived contam inant cocktails on m arine m am m al im m unity. Environ Sci Technol 2017. 314. U.S. Environm ental Protection Agency: Provisional health advisories for perfluorooctanoic acid (PFO A ) and perfluorooctane sulfonate (PFO S). In. W ashington, DC: U.S. Environm ental Protection Agency; 2009. 315. M innesota D ep artm en t o f H ealth: H ealth R isk L im its fo r Perfluorochem icals. In: Report to the Minnesota Legislature. St.Paul, M N: M innesota D epartm ent o f H ealth; 2008. 316. M innesota Departm ent of Health: 2011 Health Risk Limits for Groundwater: Perfluorobutane sulfonate. In. St.Paul, M N: H ealth R isk A ssessm ent Unit, Environm ental Health Division, M innesota Departm ent of Health; 2011. 317. Budtz-Jorgensen E, K eiding N, Grandjean P: Benchm ark dose calculation from epidem iological data. Biometrics 2001, 57(3):698-706. 318. Lau C, Thibodeaux JR, Hanson RG, Narotsky M G, Rogers JM, Lindstrom AB, Strynar M J: E ffects o f perfluorooctanoic acid exposure during pregnancy in the m ouse. Toxicol Sci 2006, 90(2):510-518. 143 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862 27-CV-10-28862 Filed in Fourth Judicial District Court 11/17/2017 7:31 PM Hennepin County, MN 319. National Research Council (U.S.). Com m ittee on Pesticides in the Diets o f Infants and Children.: Pesticides in the diets o f infants and children. W ashington, D.C.: National Academ y Press; 1993. 144 CONFIDENTIAL - SUBJECT TO A PROTECTIVE ORDER ENTERED IN HENNEPIN COUNTY DISTRICT COURT, NO. 27-CV-10-28862