Document ZJBJLVyNgBqaJ6YREw5Bmz9N7

Comments Regarding the Health Effects Documents for Perfluorooctanoic Acid (PFOA) and Perfluorooctane Sulfonate (PFOS) Prepared on behalf of 3M I-94 and McKnight Road St. Paul, MN 55144-1000 April 29, 2014 Executive Summary These comments on the Health Effects Document for Perfluorooctanoic Acid (PFOA) and Health Effects Document for Perfluorooctane Sulfonate (PFOS) have been prepared on behalf of 3M by Gradient, an environmental consulting firm with particular expertise in toxicology and risk assessment. We note that, while this work has been sponsored by 3M, the conclusions are solely those of the authors. Because of the limited time to comment on both documents (60 days to analyze > 500 pages), rather than conducting a broad analysis, we have focused on the bases for the two reference doses (RfDs) (i.e., points-ofdeparture [PODs] and choice of uncertainty factors) and human concordance of the selected values. If more time becomes available, we may provide additional comments. Our overall conclusion is that, while the assessments are comprehensive in their coverage of the toxicology and epidemiology of PFOA and PFOS, there are a number of issues with the United States Environmental Protection Agency's (US EPA's) methodology and assumptions. The RfDs are not based on the most reliable and relevant PODs, and they apply uncertainty factors that do not adequately consider inter-species differences. The impact of these choices is that the proposed RfDs are overly stringent and yield results that lack concordance with human data. We present a brief summary of the basis for this overall conclusion below, with detailed support later in this document. US EPA's overall assessments, while comprehensive, present critical assumptions without sufficiently clarifying the basis for these assumptions With respect to both PFOA and PFOS, US EPA presents multiple PODs, selecting values based on consistency, but it does not articulate a clear argument supporting the reliability of the selected endpoints. The extent to which US EPA considered other important attributes, such as relevance, reliability, and adverse nature of the selected POD, is unclear. US EPA's analysis does not adequately consider human data There are many well-conducted PFOA and PFOS human studies available. Occupational cohorts have been followed since the 1970s with exposures well above those of the general population with no evidence of adverse effects. Overall, these studies do not provide evidence of adverse effects in humans. For our analysis of the human data, using US EPA's approach for cross-species pharmacokinetic extrapolation, we calculated serum concentrations of PFOA and PFOS in humans at the dose associated with the RfD and compared these concentrations with those identified in worker studies. PFOA serum concentrations in workers ranged from non-detect (ND) to 114,100 ng/mL, with means from 4,325 to 6,800 ng/mL versus the RfD associated serum level of 143 ng/mL (assuming steady state intake). PFOS serum concentrations in workers ranged from ND to 12,830 ng/mL, with means from 1,290 to 2,440 ng/mL versus the RfD associated with serum level of 370 ng/mL (assuming steady state intake). G:\Projects\214032_PFCs\TextProc\r042914r.docx ES-1  No associations with liver biomarkers or liver changes, the most consistent finding in animal studies, have been reliably found in worker studies at serum concentrations greater than PFOA or PFOS serum concentrations associated with the RfDs using US EPA's extrapolation approach. While some inconsistent associations in worker studies have been found regarding serum lipids and PFOA or PFOS, these findings do not provide evidence of adverse effects in humans and may have non-causal explanations. The PFOA RfD based on animal studies could be an order of magnitude or more higher and still be protective US EPA has inappropriately based the POD on increased liver weight, a non-adverse finding, based on the lack of toxic effects (e.g., histopathological changes indicative of cell damage) at the associated doses. Consideration of the non-adverse nature of findings would increase the POD. This would increase the RfD by approximately a factor of 3 or more. The mode of action for findings in animal studies likely involves peroxisome proliferatoractivated receptor (PPAR), as well responses related to constitutive androstane receptor (CAR) and pregnane X receptor (PXR) activation; humans are less sensitive than rodents to chemicals operating through these mechanisms. Considering the role of PPAR, as well as these other receptors, in the toxicological response would decrease the interspecies pharmacodynamic uncertainty factor. This would increase the PFOA RfD by approximately a factor of 3 or more. The combined impact of these considerations on the PFOA RfD would be an increase of approximately an order of magnitude or more. The PFOS RfD based on neurodevelopment effects is not appropriate While US EPA identified delayed neurodevelopment in the Butenhoff et al. (2009) study, this constitutes a weak finding, and the appropriateness of its use for developing an RfD is questionable. Epidemiological studies in humans do not confirm neurodevelopment effects of PFOS in humans. An ongoing study of PFOS in monkeys will be available soon, and needs to be considered by US EPA Compared to the rat, the monkey is the preferred animal model because of better biological concordance (both in general and for PPAR, CAR, and PXR agonists). A dose-escalation study in monkeys is presently being analyzed to address certain limitations in the Seacat et al. (2002) monkey study, such as the lack of a baseline for HDL measurements and assay specificity for monkey thyroid hormones. The ongoing study was designed to rigorously evaluate these endpoints using a repeat measures, dose-escalation framework. Through application of a primate-specific dose extrapolation factor, use of the ongoing primate study would lower the cross-species pharmacokinetic uncertainty factor for purposes of RfD derivation (i.e., 48 for the primate versus 123 for the rat). In addition, this study could identify a more reliable POD than presently available animal studies. G:\Projects\214032_PFCs\TextProc\r042914r.docx ES-2  Overall, consideration of this study could potentially yield a less stringent RfD with less uncertainty than the present RfD. G:\Projects\214032_PFCs\TextProc\r042914r.docx ES-3 1 Summary of US EPA's approach to derive RfDs Perfluorooctanoic acid (PFOA) The Health Effects Document for PFOA (US EPA, 2014a) proposed a reference dose (RfD) of 0.00002 mg/kg-day, based on increased liver weight in rodents as a marker of loss of homeostasis. The studies on which the United States Environmental Protection Agency (US EPA) relied observed statistically significant increases in liver weights at doses of 1 mg/kg-day in the parent generation during a reproduction and development assay (Lau et al., 2006; York et al., 2010) or at 0.64 mg/kg-day in a 13-week assay (Palazzolo, 1993). US EPA calculated a human equivalent dose (HED) from the serum concentrations in the animals at either the lowest observed adverse effect level (LOAEL) or no observed adverse effect level (NOAEL) for each study and used this value as a point of departure (POD). It then applied uncertainty factors (10 for intrahuman variability, 3 for species differences, 10 for LOAEL to NOAEL, and 10 for less than lifetime exposures) to derive the RfD. US EPA analyzed several other studies in this manner; it found the RfDs derived from the three described here to be consistent with each other and used them as the key studies. Perfluorooctanesulfonic acid (PFOS) The Health Effects Document for PFOS (US EPA, 2014b) proposed an RfD of 0.00003 mg/kg-day, based on neurodevelopmental delays in rats. US EPA derived this value from a NOAEL of 0.3 mg/kg-day in a study by Butenhoff et al. (2009) in which rat male offspring showed increased motor activity and decreased habituation on post-natal day (PND) 17 (but not PND 13, 21, or 61) following a maternal dose of 1 mg/kg-day. US EPA calculated a HED from an average serum concentration in rats at the NOAEL dose of 10.87 mg/L and used this value as a POD. It then applied an uncertainty factor of 30 (10 for intrahuman variability and 3 for interspecies differences) to derive the RfD. US EPA analyzed several other studies in this manner, but it chose the Butenhoff et al. (2009) study based on "the consistency of the response and with recognition of the use of developmental toxicity as the sensitive endpoint" (US EPA, 2014b). 2 US EPA's overall assessments, while comprehensive, present critical assumptions without sufficiently clarifying the basis for these assumptions The rationale for the methodology used by US EPA to develop the RfDs for PFOA and PFOS is unclear in places. For example, US EPA conducted several different analyses for choosing PODs based on different methods (NOAEL-based, benchmark dose [BMD]-based, and HED-based); each analysis resulted in differing RfDs across analyses and across studies within each analysis. US EPA ultimately chose the HED method because it described this method as providing the most consistent RfDs across studies. Consistency, in and of itself, is not a sufficient basis for selection of a POD; relevance, reliability, and adverse nature of the selected endpoint must also be considered. The extent to which US EPA considered these other factors is not clear, nor is it clear if consistency was the primary basis for determination of the selected POD value. We provide the following recommendations for enhancing the utility and scientific reliability of the RfDs for PFOA and PFOS and for improving the overall credibility of the process: 1 G:\Projects\214032_PFCs\TextProc\r042914r.docx  US EPA should provide a systematic review of selected studies using pre-established criteria for assessing strengths and limitations of individual studies. This review should provide a clear scientific rationale for selection of the peer-reviewed studies and other scientific reports selected for further analysis. US EPA should base its choice of critical endpoints for RfD determination on public health relevance as supported by human-relevant mode of action determinations. Choice of the critical endpoint should be based on an integration of epidemiological and toxicological evidence. If multiple RfDs can be derived for the critical endpoint, US EPA should apply a weight-ofevidence analysis to determine the most scientifically robust and relevant RfD. 3 US EPA's analysis does not adequately consider human data A substantial number of epidemiology studies have evaluated the relationship between exposure to perfluorinated compounds (PFCs) and health effects, including the relationship with several different disease outcomes, as well as disease biomarkers. In particular, studies conducted at PFC manufacturing facilities (3M and DuPont) have used both retrospective and prospective study designs for large cohorts (n = 1,400- ~6,000) to evaluate potential health effects associated with PFCs at exposures substantially higher than general population exposures. These studies should be relied on preferentially over crosssectional general population studies, which, by study design, are insufficient to determine causation. Moreover, as noted above, because the worker studies involved serum concentrations at least 20-fold above concentrations found in the general public, the occupational studies are more reliable for establishing associations between PFCs and health effects. Overall, the studies in workers have not demonstrated any consistent adverse effects. In particular, there has been no increase in liver disease risk, which US EPA has identified as a key endpoint of concern for both PFOA and PFOS. Changes in serum lipid levels (high-density lipoprotein [HDL]/ triglycerides) have been statistically elevated in some of the worker studies. These changes are inconsistently observed and of limited biological relevance because there is no evidence that PFCs are associated with cardiovascular disease (CVD) or metabolic syndrome. The sections below summarize studies in workers for these key endpoints (see also Tables 1 and 2). Liver Disease. Data from studies of workers with high exposures to PFOA and PFOS do not provide evidence of increased liver disease in humans (see Table 1). Several epidemiology studies (mortality incidence and episodes of care studies) conducted at 3M and DuPont plants demonstrate a lack of association of liver-related disorders with PFOA and/or PFOS exposure. For example, in the most extensive mortality studies conducted at the 3M plants (one in Decatur, Alabama, and one in Cottage Grove, Minnesota), there were no statistically significant associations between employment at the plant and mortality from liver disease (cirrhosis or cancer) (Alexander et al., 2003; Lundin et al., 2009). These findings are consistent with the most recent DuPont manufacturing PFOA plant studies (Leonard et al., 2008), as well as the episodes of care and questionnaire studies conducted by 3M (Alexander and Grice, 2006; Grice et al., 2007; Olsen et al., 2004; Alexander et al., 2014). Hepatic Enzymes. Data from studies of workers with high exposures to PFOA and PFOS indicate that these chemicals are not associated with consistent or adverse changes in liver enzyme parameters in humans (see Table 2). Clinical chemistries have been evaluated in PFC workers since the 1970s. For example, Olsen et al. (2003a) reported that none of the adjusted odds ratios for hepatic clinical chemistry tests from workers in Antwerp, Belgium, and Decatur, Alabama, were statistically significant (p < 0.05). Also, in a longitudinal analysis, there were no statistically significant coefficients for PFOS, PFOA, or total organic fluorine associated with changes in HDL or results of various liver function tests (i.e., for alkaline phosphatase, gamma-glutamyl transferase, aspartate aminotransferase [AST], alanine aminotransferase [ALT], total bilirubin, or direct bilirubin), after adjustment for potential confounders 2 G:\Projects\214032_PFCs\TextProc\r042914r.docx (Olsen et al., 2003a). In a recent longitudinal study of workers involved in the demolition of PFC manufacturing facilities, there was no association between PFOA or PFOS serum levels and the liver enzymes AST and ALT. In fact, one of the few statistically significant findings in the study was that AST levels increased in workers whose serum PFOA levels decreased during the study period (Olsen et al., 2012). Serum PFOA levels in this study covered a large range (< 1 to ~5,000 ng/mL PFOA and ~1 to 1,000 ng/mL PFOS). Serum Lipids. EPA did not consider serum lipids as a critical effect in their analysis. We discuss it here because there have been inconsistent findings of reduced serum lipids associated with PFOA and PFOS in some human studies. As discussed below, these findings are inconsistent and not clinically significant, and may have alternative non-causal, explanations (see, for example, the discussion in Olsen et al., 2012 of possible non-causal explanations for PFOA and PFOS-associated changes in serum lipids, such as a common impact on membrane transport for PFOA, PFOS and cholesterol). There is no reliable evidence that PFOA or PFOS causes adverse changes in serum lipids in workers with serum PFOA and PFOS levels much higher than those associated with the proposed RfDs. Overall, the associations between serum PFOA or PFOS and various serum lipid indicators are inconsistent across studies. When associations are observed, changes are modest and within normal values (i.e., not clinically meaningful). In some cases, the changes in lipid serum levels are in the opposite direction of those observed in animal studies. While some statistically significant associations were reported in worker studies between PFOA or PFOS serum concentrations and increased measures of some serum lipids, these observations were limited to certain studies of certain populations, such as the Decatur and Antwerp workers (Olsen et al., 2003a) and the Parkersburg workers (Sakr et al., 2007a,b), and several inconsistencies and contradictions exist regarding these findings (see Table 2). An ongoing clinical trial with PFOA as anti-cancer drug has found some significant associations between LDL cholesterol and PFOA serum levels. The levels in these patients range up to 495,000 ng/mL, a value over 70 times higher than the highest mean level measured in any worker study (Olsen et al., 2000). The highest dose administered was 1,200 mg/per week. In patients with the highest PFOA serum levels (~330,000 to 495,000 ng/mL1), an approximately 20% decrease in LDL cholesterol was measured; the decrease was about 10% in the mid-exposure group (serum levels ~165,000 to 330,000 ng/mL1). While the long-term significance of these changes, if any, will need to be monitored, at this point in the clinical trial the researchers have determined that these PFC serum levels were associated with a favorable toxicity profile and that the maximum tolerated dose (MTD) had not been achieved (MacPherson et al, 2010) Because of general symptoms common to drug-related toxicities (e.g. fatigue and nausea), a dose slightly below the maximum dose tested (1,200 mg/wk) in the Phase 1 trials of 1,000 mg/week was selected for the maximum dose in the Phase 2 trial (MacPherson et al., 2011). Cardiovascular Disease. One of the key health concerns associated with changes in serum lipids is CVD. It is noteworthy that, for workers exposed to high levels of PFCs, most studies show no association with risk of CVD (see Table 1). For example, in a recent study conducted at 3M's Cottage Grove plant, there was no association with an elevated risk of dying from CVD when compared to the general population of Minnesota (Lundin et al., 2009). An earlier study at DuPont's PFOA Parkersburg plant found an increased incidence of rheumatic heart disease, acute myocardial infarction, and atherosclerosis and aneurysm in PFOA-exposed workers (Leonard, 2003). However, several follow-up studies, some refined to specifically investigate the relationship between PFOA exposure and CVD, were not able to replicate initial findings; these more recent studies found no increase between PFOA exposure and mortality from ischemic heart disease (Leonard et al., 2008; Sakr et al., 2009; Steenland and Woskie, 2012). Although not described in detail in this text, diabetes, which is a metabolic syndrome with links to 1 Converted from concentrations reported in M. 3 G:\Projects\214032_PFCs\TextProc\r042914r.docx serum lipoproteins, was also not reliably elevated in occupational cohort studies of PFOA and PFOS (Alexander, 2001; Steenland and Woskie, 2012; Mandel and Johnson, 1995). 4 Consideration of human data indicates RfDs are overly conservative PFOA and PFOS serum concentrations at the US EPA-derived RfDs As summarized above, based on the available studies in workers, there is no relationship between PFOA and/or PFOS exposure and adverse health effects. In particular, there is no association with liver effects, including changes in liver enzyme levels. Despite this, US EPA has selected liver effects (from animal studies) as the key endpoint. The worker studies also show no biologically relevant increases in serum lipids, as well as no increase in CVD or metabolic syndrome (which could result from alteration in serum lipoproteins). To provide perspective on the serum levels associated with no adverse outcome in the worker studies and toxicity criteria provided by US EPA, we converted the RfD for PFOA and PFOS to equivalent human serum levels (assuming steady state). We performed this calculation using the same methodology US EPA used to calculate HEDs from modeled average animal serum concentrations. The equation is as follows: Human Equivalent Dose [HED] = average serum concentration (in mg/L) modeled from animal experiments X human clearance (0.00014 L/kg-day for PFOA and 0.000081 L/kg-day for PFOS). Based on this equation, the human serum concentrations of these chemicals at the RfDs can be calculated: Human serum concentration = RfD (mg/kg-day)/Clearance (L/kg-day). According to US EPA methods, the human serum concentrations at the RfDs are 143 ng/mL for PFOA and 370 ng/mL for PFOS. Comparison of worker study serum levels to the serum levels associated with US EPA-derived RfDs The worker (and pharmaceutical) studies involve exposures that are many times higher than the calculated serum levels that would be associated with the proposed RfDs for PFOA (143 ng/mL) and PFOS (370 ng/mL). Table 1 presents studies evaluating PFC exposure and disease, including the serum levels measured in the studies, if available. While some of the studies did not report serum levels, information available from biomonitoring studies at the plants where the epidemiological investigations were conducted can be used to approximate the serum levels in the health effects studies. These biomonitoring studies2 have reported PFOA arithmetic means of 4,325 to 6,800 ng/mL with maximum values over 114,100 ng/mL from the 3M Cottage Grove facility. These measurements are 30 to 47 times greater (based on mean levels) and range to almost 800 times higher (based on maximum levels) than the calculated serum level associated with the proposed RfD for PFOA (143 ng/mL). PFOS arithmetic means of 1,290 to 2,440 ng/mL, with maximum values of approximately 12,800 ng/mL, have also been reported in these worker populations from the 3M Decatur Facility. These mean levels are 3 to 7 times higher (based on mean levels) and range to approximately 35 times higher (based on maximum levels) than the calculated serum level associated with the proposed RfD for PFOS (370 ng/mL). 2 PFOA levels at Cottage Grove, where 3M epidemiological studies have been conducted. 4 G:\Projects\214032_PFCs\TextProc\r042914r.docx Exposure durations for workers exposed to PFOA or PFOS ranged from 6 months to more than 20 years. Given half-lives estimated for workers exposed to PFOS and PFOA of 3.8 and 5.4 years,3 respectively (Olsen et al., 2005, 2007), and assuming that 1) steady-state serum concentrations (e.g., > 90% steadystate concentration) can be achieved in 4 half-lives (Shen, 2013), and 2) exposure levels are relatively constant, steady-state serum concentrations for workers would be reached after approximately 15 years for workers exposed to PFOA, and 22 years for workers exposed to PFOS. Although consideration of whether workers are at steady state is important for estimating potential human doses based on serum levels, it is not an issue for comparison of serum effect levels between humans and animals. 5 The PFOA RfD based on animal studies could be an order of magnitude or more higher and still be protective Adaptive changes in the liver US EPA has designated increased liver weight as the most sensitive adverse effect on which to base the proposed RfD for PFOA (US EPA, 2014a). The studies that US EPA used to establish a POD for RfD development reported increased liver weights in rats (Palazzolo, 1993; Perkins et al., 2004; Butenhoff et al., 2004; York et al., 2010) and mice (Lau et al., 2006; Dewitt et al., 2008). At issue is the question of whether increased liver size by itself is an adverse effect. In its 2002 guidance document, the US EPA Health Effects Division Toxicology Science Advisory Council (TOXSAC) (2002) noted that a statistically significant increase in liver size alone is not a reliable indicator of hepatic toxicity. Similarly, Williams and Iatropoulos (2002) also described increases in liver size (and liver hypertrophy) as an adaptive effect, noting that hepatocyte viability will not be compromised if liver changes are within the context of maintaining homeostasis. This position is consistent with that of the Agency for Toxic Substances and Disease Registry (ATSDR). When evaluating liver effects as an endpoint for establishing health criteria, ATSDR notes that effects such as increased liver weight and elevated liver enzymes are likely adaptive changes which should be categorized as "less serious" if they do not occur in conjunction with "other effects showing a threat to the organism from serious damage to the liver" (Pohl and Chou, 2005). A recent report by the European Society of Toxicologic Pathology evaluated adverse vs. adaptive changes in the liver. This report concluded that "[h]epatomegaly as a consequence of hepatocellular hypertrophy without histologic or clinical pathology alterations indicative of liver toxicity is considered an adaptive and a non-adverse reaction" (Hall et al., 2012). The analysis further noted that "liver weight increases up to 150% of control values may be considered nonadverse in the context of safety evaluation for a chemical," and the authors ultimately proposed a framework for assessing the biological significance of liver changes whereby liver weight changes would only be considered adverse when accompanied by histological evidence of structural degeneration (e.g., hepatocyte necrosis, fibrosis, steatotic vascular degeneration). In the absence of liver pathology, the framework considered changes in ALT (2- to 3-fold increase) and biologically significant (emphasis added) changes in other biomarkers for liver dysfunction, including ALP, AST, cholesterol and triglycerides, to be adverse. 3 It should be noted that these half-lives are greater than those for the general population. This may be related, at least in part, to larger body burdens for workers as compared with the general population. 5 G:\Projects\214032_PFCs\TextProc\r042914r.docx The conclusions of Hall et al. (2012) and others are applicable to the findings in several of the studies with PFOA and PFOS in which increases in liver weight in monkeys (Butenhoff et al., 2002; Seacat et al., 2002) and rats (Palazzolo, 1993; Butenhoff et al., 2004; York et al., 2010) were not associated with a functional deficit and were completely reversible after a sufficient recovery period. For example, in a key study selected by US EPA to develop the PFOA RfD, Palazzolo (1993) found increased liver weights in rats at the 0.64 mg/kg-day dose but no statistically significant necrotic or inflammatory changes in any dose group (0.06, 0.64, 1.94, or 6.5 mg/kg-day). At doses up to 1.94 mg/kg-day, the percent changes in relative liver weights were less than 150%, and all changes were reversible. Clinical chemistry results were not available for this study, but a bioassay in rats found only slight increases in liver enzymes (less than two-fold) at the highest dose tested (14-16 mg/kg-day) at the end of the 2-year study period (Butenhoff et al., 2012). In light of this information, a LOAEL of 6.5 mg/kg-day (compared to US EPA's LOAEL of 0.64 mg/kg-day) could be identified from this study, based on a liver weight increase greater than 150%. Moreover, this is a conservative value; this increase was still likely adaptive, occurring in the absence of other adverse histological findings in the liver. The next lowest LOAEL that US EPA identified as a possible POD (see table 5.11 in US EPA PFOA Health Effects Document) was an immunotoxicology study in mice (Dewitt et al., 2008). In this study, US EPA identified a LOAEL of 0.94 mg/kg-day, based on increased absolute and relative liver weights. Actual liver weights were not reported in the paper, but, according to the authors, relative liver weights for all dose groups were elevated (p < 0.05) relative to controls by 135-160% 1 day after exposure ended and 122-145% 15 days after exposure ended. Given these mild increases in liver weights (up to 7.5 mg/kg-day, the highest dose tested), these changes should not be considered adverse unless liver pathology could also be documented (it was not). US EPA also identified a LOAEL of 1 mg/kg-day from a two-generation study in rats for a possible POD, based, in part, on increased liver weight (York et al., 2010). However, as with the other studies US EPA cited, the effects observed in the liver at 1 mg/kg-day and the next higher dose (3 mg/kg-day) may be adaptive. It was only at the two highest doses (10 and 30 mg/kg-day) that liver weights were greater than 150% of control. The authors noted no adverse liver pathology at any dose. Based on these findings, 10 mg/kg-day (or higher) should be considered the LOAEL for liver effects in this study. Loveless et al. (2006) conducted a 28-day study in mice and rats. In this study, US EPA identified a LOAEL of 1 mg/kg-day and a NOAEL of 0.3 mg/kg-day. The LOAEL selected was not based on liver weight increases alone; individual cell necrosis and focal necrosis in the animals was also noted at 1 mg/kg-day. Using this study, which more appropriately identified actual adverse effects in the liver, would result in a NOAEL 0.3 mg/kg-day, which is 5 times higher than the NOAEL of 0.06 mg/kg-day from Palazzolo (1993). Overall, the liver enlargements observed in the studies relied upon by US EPA should be considered adaptive, and they do not reflect appropriate endpoints on which to base the RfD for PFOA. Acknowledgement of the non-adverse nature of the findings would increase the POD, and likely the RfD, at least 3-fold. An additional consideration regarding liver enlargement in rodents is the applicability of this endpoint in humans based on the mechanism of action. Increased liver weight mediated by peroxisome proliferatoractivated receptor (PPAR), common in rodents, has not been observed in non-human primates at plausible exposure levels and is not expected to occur in humans (Klaunig et al., 2003; Lake, 1995). 6 G:\Projects\214032_PFCs\TextProc\r042914r.docx PPAR and Other Receptors Associated with a Species-Specific Response PPAR is a ligand-activated nuclear hormone receptor that controls expression of peroxisomal enzymes involved in lipid metabolism, such as acyl coenzyme A (CoA) oxidase, catalase, cytochrome P450s, and fatty acid binding protein. Natural ligands for PPAR include eicosanoids and fatty acids. PPAR can also be activated by structurally diverse synthetic ligands, including fibrates, phthalates, chlorinated solvents and PFCs. PPAR activators target the liver, where, in addition to enzyme induction, they can cause a substantial increase in the number and volume of peroxisomes (i.e., peroxisome proliferation) and hepatocellular hypertrophy and proliferation, which can result in increased liver weight (Berger and Moller, 2002; Gonzalez and Shah, 2008; Klaunig et al., 2003; Lee et al., 2003). There are important species differences, with rats and mice being much more sensitive to peroxisome proliferation following exposure to PPAR ligands than humans and other primates. In contrast to rats and mice, the increase in peroxisomes at therapeutic doses of fibrates (approximately 20-40 mg/kg-day) is modest, with either no or only slightly increased levels of peroxisomal enzymes even at doses 4-fold above therapeutic doses (Klaunig et al., 2003). In transgenic mice carrying a humanized PPAR, peroxisome proliferation (but not hepatomegaly or hepatocyte proliferation) is observed in response to treatment with PPAR activators (Yang et al., 2008). Recent studies have also demonstrated that the activation of CAR and PXR produce rodent-specific effects (Elcombe et al., 2014). The hepatocellular response to PFOA is likely mediated to a large extent by direct activation of PPAR and to a lesser extent by either direct or indirect activation of other nuclear receptors, including the constitutive androstane receptor (CAR) and pregnane X-receptor (PXR). Rosen et al. (2008) found that, in mice exposed to PFOA in drinking water, 85% of gene expression changes in the liver were PPARdependent. Many of the PPAR-independent genes identified by Rosen et al. are activated by CAR. Elcombe et al. (2010) found that, in rats exposed to PFOA in the diet, hepatocellular hypertrophy and hyperplasia occurred in conjunction with biochemical (e.g., CYP4A1 and peroxisomal -oxidation) and clinical (e.g., decreased serum total cholesterol and triglycerides) markers of PPAR activation. Moreover, these PPAR-associated effects observed in rats exposed to PFOA were comparable to those observed with the prototypical PPAR agonist Wy 14,643. Elcombe et al. (2010) also observed biochemical effects indicative of activation of CAR, as indicated by CYPCB1/2 protein expression, and CAR/PXR, as indicated by CYP3A1 protein expression. Consistent with involvement of PPAR in PFOA's hepatic effects, results from several studies provide evidence that humans are likely less susceptible than rodents or mice to these hepatic effects. Using COS-1 cells transiently transfected with a luciferase reporter gene under control of either a mouse or human PPAR ligand binding domain, Wolf et al. (2008) observed that mouse PPAR was more sensitive than human PPAR to activation by PFOA by approximately 3-fold. Bjork and Wallace (2009) observed increased expression of the PPAR target genes acyl-CoA oxidase and acyl-CoA thioesterase in primary rat hepatocytes, but the authors reported no increase in expression of the corresponding human genes in primary human hepatocytes or HepG2/C3a hepatoma cells. In an in vivo study of mice exposed to PFOA via gavage, Nakamura et al. (2009) observed increased expression of PPAR target genes and increased levels of hepatic lipids in wild-type (wt) mice, but the authors saw no such increases in mice either lacking PPAR (PPAR null mice) or with a humanized PPAR (hPPAR). As discussed by US EPA (2014), studies using PPAR-null mice provide evidence that PFOA can cause hepatic effects in the absence of PPAR. However, the hepatic effects of PFOA observed in PPAR-null mice are generally less robust than effects in wt mice. For example, for most of the genes studied in an in vitro gene profiling study, the magnitude of gene expression changes in livers of wt mice exposed to 1 or 3 mg/kg PFOA were generally less than that observed for PPAR-null mice (Wolf et al., 2008). In a 7 G:\Projects\214032_PFCs\TextProc\r042914r.docx developmental toxicity study, the LOAEL for an increase in relative liver weight in adults was 1 mg/kg in wt mice vs. 3 mg/kg in PPAR-null mice. The difference in sensitivity to PFOA was more pronounced for the pups, with a LOAEL for increased relative liver weight of 0.1 mg/kg in wt pups vs. 3 mg/kg in PPAR-null pups (Abbott et al., 2007). In mice exposed to 5 mg/kg PFOA in the peripubertal period (PND 21-49), Zhao et al. (2010) observed significant effects on expression of drug and hormone metabolizing genes in the liver of wt but not PPAR-null mice. In mice treated with 5 mg/kg from gestation day (GD) 1-17, Albrecht et al., 2013 observed significantly increased relative liver weight on both GD18 and GD20 for wt mice, but only on GD18 for PPAR-null and hPPAR mice. In contrast to the aforementioned studies with PPAR-null mice, a study by Minata et al. (2010) observed comparable increases in liver weight in wt and PPAR-null mice exposed to approximately 5.4, 10.8, and 21.6 mg/kg-day PFOA for 4 weeks. Although the reason for this discrepancy is not clear, in terms of human relevance, interpretation of results from studies with PPAR-null mice should consider inherent differences between wt and PPAR-null mice that could influence response to chemical exposure. For example, Aoyama et al. (1998) reported that mitochondrial fatty acid metabolism is defective in PPARnull mice, which could make them susceptible to disruption of fatty acid homeostasis. In this respect, studies with hPPAR mice - which maintain normal fatty acid metabolism - would be more informative regarding potential effects in humans (Yang et al., 2008). As discussed above, studies by Nakamura et al. (2009) and Albrecht et al. (2013) indicate that hPPAR mice are less susceptible than wt mice to the hepatic effects of PFOA. Taken together, results from in vitro and in vivo studies provide solid evidence that PPAR, as well as the activation of CAR and PXR, are involved in mediating liver effects of PFOA and humans would be less sensitive than rats and mice to these effects. As such, we recommend that the inter-species pharmacodynamic uncertainty factor US EPA uses to calculate the RfD for PFOA be reduced from 3 to 1. Conclusion A revised analysis of the RfD for PFOA that does not rely on adaptive liver changes as the sensitive endpoint will result in a point-of-departure that would be 3-fold or higher. This adjustment, in combination with a reduction of the interspecies pharmacodynamic uncertainty factor from 3 to 1 (justified because humans would be less sensitive than rats to PPAR) would potentially increase the PFOA RfD by approximately an order of magnitude or more. 6 The PFOS RfD based on neurodevelopmental effects in rats is not appropriate While US EPA identified delayed neurodevelopment in the Butenhoff et al. (2009) study as a critical endpoint for its POD, this constitutes a weak finding, and the appropriateness of its use for developing an RfD is questionable. US EPA should reconsider its use of this study, and neurodevelopment as an endpoint for the development of an RfD, for several reasons: The finding in the Butenhoff et al. (2009) study may not be dose-related. For example, the increased motor activity, indicating an absence of habituation to the environment, observed in males at PND 17 was reported to be statistically significant at the p < 0.05 level. There are, however, more than 200 comparisons in this report, and there were no corrections made for multiple comparisons. At a significance level of 0.05, one could expect to find about 10 results that are statistically significant by chance alone. The developmental effect was transient, occurring on PND 17 but not PNDs 13, 21, or 61. 8 G:\Projects\214032_PFCs\TextProc\r042914r.docx  With respect to the animal studies, EPA needs to conduct a weight-of-evidence analysis, consistent with its own guidance (US EPA, 1991, 1998), and reconsider neurodevelopment as a critical effect. It should be noted for example, that certain other studies, e.g., Lau et al. (2003) and Luebker et al. (2005), do not confirm neurodevelopmental findings. There is no confirmed evidence for neurodevelopmental effects from PFOS in humans. Two studies of the association between neurodevelopment and maternal PFOS levels found no evidence of association (Fei et al., 2008; Stein and Savitz, 2011), while two other studies provided only limited evidence of an association (Hoffman et al., 2010; Chen et al., 2013). None of these studies controlled for certain confounders such as heritability factors (Bellinger, 2009; Comings et al., 2005; Millichap, 2008) which are strongly associated with neurodevelopmental effects in humans. Overall, there is no confirmed evidence for an association between PFOS and neurodevelopment. Because of the lack of confirmed animal and human data for neurodevelopmental effects, we propose that, if US EPA uses animal data for RfD development, it consider using liver toxicity (specifically, indicators of adverse effects) as an endpoint rather than neurodevelopment. Liver effects of PFOS have been evaluated in rats (Seacat et al., 2003; Curran et al., 2008; Goldenthal, 1978) and mice (Bijland et al., 2011), and a new study in monkeys is currently underway (discussed in Section 7). 7 Ongoing studies of PFOS in monkeys will be available soon, and should be considered in the assessment US EPA selected endpoints from studies in rodents and monkeys as the basis for the PFOS RfD. Compared to the rat, the monkey is the preferred animal model because of better biological concordance with humans, both in general and specifically for PFOS-related toxicity. Based on similarities in genetics and physiology, in response to chemical insults, non-human primates are generally expected to mount responses to chemical agents more similar to humans than are rodents (Burbacher and Grant, 2001; Kane et al., 2006). This is likely the case for PFOS, which exerts its toxicity, in part, through a PPAR mechanism - a rodent-specific response. Kane et al. (2006) state that the cynomolgus monkey is "superior to rodent models with respect to the preclinical evaluation of PPAR agonists." Seacat et al. (2002) evaluated the effects of PFOS in cynomolgus monkeys. Animals were dosed orally for 182 days with 0.03, 0.15, or 0.75 mg/kg-day PFOS (6 animals/sex/group, except for 0.03 mg/kg-day group, which has 4 animals/sex). For the 0.15 and 0.75 mg/kg-day dose groups, recovery animals (2 animals/sex/group) were monitored for 1 year after treatment cessation. In the 0.75 mg/kg-day dose group, Seacat et al. (2002) observed mortality in two male monkeys on days 155 and 179. The authors considered these findings to be PFOS-related, although they also note that the cause of morbidity or mortality for these two monkeys could not be determined unequivocally. Other key findings in the highdose group included decreased body weights, increased liver weights (which were reversible) decreased serum total cholesterol, lowered T3 concentrations, and lowered estradiol concentrations. The authors observed a decrease in HDL in males at the low- and high-dose group and in females at the mid- and high-dose group. Although they observed statistically significantly increased hepatic peroxisome activity, as measured by palmitoyl CoA oxidase activity, in the females at the 0.75 mg/kg-day dose group, the authors noted that this finding did not exceed the criteria for biological significance of a two-fold increase over the control values. At the end of the dosing period, sacrificed animals in the 0.75 mg/kg-day dose group showed evidence of hepatocellular hypertrophy and lipid vaculolation; overall, the authors saw no peroxisomal or cell proliferation. With respect to the decreases in HDL, these results were difficult to interpret due to the lack of baseline values for HDL and robust clinical normal data for cynomolgus monkeys. In the recovery animals, no clinical or hepatic effects were seen. 9 G:\Projects\214032_PFCs\TextProc\r042914r.docx The methodological uncertainties with respect to measurement of HDL and TSH in the Seacat et al. (2002) monkey study and the small numbers of animals in the recovery group limit the ability to draw robust conclusions from this study regarding those clinical measurements. Thus, it is important that US EPA consider developing findings from an ongoing analysis of PFOS effects in cynomolgus monkeys that addresses uncertainties from the earlier study (in-phase study was completed in April 2014). The aim of this study is to determine if, using a primate model, there are PFOS-related changes from baseline in certain clinical chemistry parameters identified from previous work as potentially associated with PFOS, i.e., serum concentrations of cholesterol and thyroid-related hormones (Ley, 2014). The study will also measure other clinical chemistry parameters of potential toxicological interest as based on rodent studies (e.g., liver enzymes). The dosing and measurement schedule in this study has taken advantage of understanding developed from prior toxicological (Seacat et al., 2002) and pharmacokinetic (Chang et al., 2012) studies in monkeys, and it is expected to be able to identify clinically non-adverse serum alterations associated with PFOS; such alterations could potentially be applied to develop a more robust and humanrelevant NOAEL as a POD for a PFOS RfD. In addition, US EPA could use a lower interspecies pharmacokinetic extrapolation factor (48 vs. 123) if it relied on monkey studies instead of rat studies; this would be expected to yield a less stringent RfD with less uncertainty. 8 Conclusions While the Health Effect Documents provide comprehensive coverage of the toxicology and epidemiology of PFOA and PFOS, there are a number of issues with US EPA's methodology and assumptions in these documents. The RfDs are not based on the most reliable and relevant PODs and apply uncertainty factors that do not adequately consider inter-species differences. For example, the PFOA assessment does not appropriately consider the role of PPAR, as well as other receptors, for the interspecies pharmacodynamic uncertainty factor. The PFOS POD is based on an endpoint (neurodevelopmental effects in rats) that may not be dose-related and has not been observed in humans. Because of these choices, the proposed RfDs are overly stringent and yield results that lack concordance with human data. In addition, the basis for some of US EPA's assumptions (e.g., selection of a particular POD) is not described sufficiently. The overall time US EPA allotted for public comment is inadequate to conduct a full review. Because of this limited time (only 60 days to analyze > 500 pages), we have focused on the bases for the two RfDs (i.e., PODs and uncertainty factors) and human concordance of the selected values. If more time were available, we would conduct a more detailed analysis of a number of other topics, such as the relevance and potential impact of studies other than those used as the basis for the RfDs and the calculation methods for the PODs. Given the importance of this assessment to risk management decisions for PFOA and PFOS, such as developing drinking water criteria, we recommend that US EPA provide at least another 60 days for stakeholder input. We also recommend that a rigorous peer review of the documents be conducted and that such peer review involve individuals with significant expertise in the key technical areas (such as toxicological mode of action for liver effects, and PFOA and PFOS pharmacokinetics). 10 G:\Projects\214032_PFCs\TextProc\r042914r.docx References Abbott, BD; Wolf, CJ; Schmid, JE; Das, KP; Zehr, RD; Helfant, L; Nakayama, S; Lindstrom, AB; Strynar, MJ; Lau, C. 2007. "Perfluorooctanoic acid induced developmental toxicity in the mouse is dependent on expression of peroxisome proliferator activated receptor-alpha." Toxicol. Sci. 98(2):571581. Albrecht, PP; Torsell, NE; Krishnan, P; Ehresman, DJ; Frame, SR; Chang, SC; Butenhoff, JL; Kennedy, GL; Gonzalez, FJ; Peters, JM. 2013. "A species difference in the peroxisome proliferator-activated receptor alpha-dependent response to the developmental effects of perfluorooctanoic acid." Toxicol. Sci. 131(2):568-582. Alexander, B; Raleigh, K; Ramachandran, G; Church, T; Scott, L; Mongin, S; Logan, P; Morey, S; Olsen, G. 2014. "Mortality experience of APFO production workers." Toxicologist 138(1):584. Abstract No. 2211. Alexander, BH. 2001. "Final Report, Mortality Study of Workers Employed at the 3M Cottage Grove Facility." University of Minnesota, School of Public Health, April 26. Alexander, BH; Grice, M. 2006. "Self-Reported Medical Conditions in Perfluorooctanesulfonyl Fluoride Manufacturing Workers (Final)." University of Minnesota, School of Public Health. Submitted to US EPA Docket No. AR-226. 26p., February 1. Alexander, BJ; Olsen, GW; Burris, JM; Mandel, JH; Mandel, JS. 2003. "Mortality of employees of a perflurooctanesulfonyl fluoride manufacturing facility." Occup. Environ. Med. 60:722-729. Aoyama, T; Peters, JM; Iritani, N; Nakajima, T; Furihata, K; Hashimoto, T; Gonzalez, FJ. 1998."Altered constitutive expression of fatty acid-metabolizing enzymes in mice lacking the peroxisome proliferatoractivated receptor alpha (PPARalpha)." J. Biol. Chem. 273(10):5678-5684. Bellinger, DC. 2009. "Interpreting epidemiologic studies of developmental neurotoxicity: Conceptual and analytic issues." Neurotoxicol. Teratol. 31(5):267-274. Berger, J; Moller, DE. 2002. "The mechanisms of action of PPARs." Annu. Rev. Med. 53:409-435. Bijland, S; Rensen, PC; Pieterman, EJ; Maas, AC; van der Hoorn, JW; van Erk, MJ; Havekes, LM; Willems van Dijk, K; Chang, SC; Ehresman, DJ; Butenhoff, JL; Princen, HM. 2011."Perfluoroalkyl sulfonates cause alkyl chain length-dependent hepatic steatosis and hypolipidemia mainly by impairing lipoprotein production in APOE*3-Leiden CETP mice." Toxicol. Sci.123(1):290-303. Bjork, JA; Wallace, KB. 2009. "Structure-activity relationships and human relevance for perfluoroalkyl acid-induced transcriptional activation of peroxisome proliferation in liver cell cultures." Toxicol. Sci. 111(1):89-99. Burbacher, TM; Grant, KS. 2001. "Nonhuman primates as animal models for toxicology research." Curr. Protoc. Toxicol. Unit 1.1:1-9. Butenhoff, J; Costa, G; Elcombe, C; Farrar, D; Hansen, K; Iwai, H; Jung, R; Kennedy, G; Lieder, P; Olsen, G; Thomford, P. 2002. "Toxicity of ammonium perfluorooctanoate in male cynomolgus monkeys after oral dosing for 6 months." Toxicol. Sci. 69:244-257. 11 G:\Projects\214032_PFCs\TextProc\r042914r.docx Butenhoff, JL; Ehresman, DJ; Chang, SC; Parker, GA; Stump, DG. 2009. "Gestational and lactational exposure to potassium perfluorooctanesulfonate (K(+)PFOS) in rats: Developmental neurotoxicity." Reprod. Toxicol. 27(3-4):319-330. Butenhoff, JL; Kennedy, GL; Frame, SR; O'Connor, JC; York, RG. 2004. "The reproductive toxicology of ammonium perfluorooctanoate (APFO) in the rat." Toxicology 196:95-116. Chang, SC; Noker, PE; Gorman, GS; Gibson, SJ; Hart, JA; Ehresman, DJ; Butenhoff, JL. 2012. "Comparative pharmacokinetics of perfluorooctanesulfonate (PFOS) in rats, mice, and monkeys." Reprod. Toxicol. 33(4):428-440. Chen, MH; Ha, EH; Liao, HF; Jeng, SF; Su, YN; Wen, TW; Lien, GW; Chen, CY; Hsieh, WS; Chen, PC. 2013. "Perfluorinated compound levels in cord blood and neurodevelopment at 2 years of age." Epidemiology 24(6):800-808. Comings, DE; Chen, TJ; Blum, K; Mengucci, JF; Blum, SH; Meshkin, B. 2005. "Neurogenetic interactions and aberrant behavioral co-morbidity of attention deficit hyperactivity disorder (ADHD): Dispelling myths." Theor. Biol. Med. Model. 23(2):50. Costa, G; Sartori, S; Consonni, D. 2009. "Thirty years of medical surveillance in perfluorooctanoic production workers." J. Occup. Environ. Med. 51(3):364-372. Curran, I; Hierlihy, SL; Liston, V; Pantazopoulos, P; Nunnikhoven, A; Tittlemier, S; Barker, M; Trick, K; Bondy, G. 2008. "Altered fatty acid homeostasis and related toxicologic sequelae in rats exposed to dietary potassium perfluorooctanesulfonate (PFOS)." J. Toxicol. Environ. Health A71(23):1526-1541. DeWitt, JC; Copeland, CB; Strynar, MJ; Luebke, RW. 2008. "Perfluorooctanoic acid-induced immunomodulation in adult C57BL/6J or C57BL/6N female mice." Environ. Health Perspect.116(5):644-650. Elcombe, CR; Elcombe, BM; Foster, JR; Farrar, DG; Jung, R; Chang, SC; Kennedy, GL; Butenhoff, JL. 2010. "Hepatocellular hypertrophy and cell proliferation in Sprague-Dawley rats following dietary exposure to ammonium perfluorooctanoate occurs through increased activation of the xenosensor nuclear receptors PPARa and CAR/PXR." Arch. Toxicol. 84(10):787-798. Elcombe, CR; Peffer, RC; Wolf, DC; Bailey, J; Bars, R; Bell, D; Cattley, RC; Ferguson, SS; Geter, D; Goetz, A; Goodman, JI; Hester, S; Jacobs, A; Omiecinski, CJ; Schoeny, R; Xie, W; Lake, BG. 2014. "Mode of action and human relevance analysis for nuclear receptor-mediated liver toxicity: A case study with phenobarbital as a model constitutive androstane receptor (CAR) activator." Crit. Rev. Toxicol. 44(1):64-82. Fei, C; McLaughlin, JK; Lipworth, L; Olsen, J. 2008. "Prenatal exposure to perfluorooctanoate (PFOA) and perfluorooctanesulfonate (PFOS) and maternally reported developmental milestones in infancy." Environ. Health Perspect. 116(10):1391-1395. Gilliland, FD; Mandel, JS. 1993. "Mortality among employees of a perfluorooctanoic acid production plant." J. Occup. Med. 35(9):950-954. Gilliland, FD; Mandel, JS. 1996. "Serum perfluorooctanoic acid and hepatic enzymes, lipoproteins and cholesterol: A study of occupationally exposed men." Am. J. Ind. Med. 29(5):560-568. 12 G:\Projects\214032_PFCs\TextProc\r042914r.docx Goldenthal, EI; Jessup, DC; Geil, RG; Jefferson, ND; Arceo, RJ [International Research and Development Corp.]. 1978. "Ninety Day Subacute Rat Toxicity Study." Report to 3M Co. Submitted to US EPA Docket No. AR-226. November 10. Gonzalez, FJ; and Shah, YM. 2008. "PPAR: Mechanism of species differences and hepatocarcinogenesis of peroxisome proliferators." Toxicology 246:2-8. Grice, MM; Alexander, BH; Hoffbeck, R; Kampa, DM. 2007. "Self-reported medical conditions in perfluorooctanesulfonyl fluoride manufacturing workers." J. Occup. Environ. Med. 49(7):722-729. Hall, AP; Elcombe, CR; Foster, JR; Harada, T; Kaufmann, W; Knippel, A; Kuttler, K; Malarkey, DE; Maronpot, RR; Nishikawa, A; Nolte, T; Schulte, A; Strauss, V; York, MJ. 2012. "Liver hypertrophy: A review of adaptive (adverse and non-adverse) changes - Conclusions from the 3rd International ESTP Expert Workshop." Toxicol. Pathol. 40(7):971-994. Hoffman, K; Webster, TF; Weisskopf, MG; Weinberg, J; Vieira, VM. 2010. "Exposure to polyfluoroalkyl chemicals and attention deficit hyperactivity disorder in U.S. children aged 12-15 years." Environ. Health Perspect. 118:1762-1767. Kane, CD; Francone, OL; Stevens, KA. 2006. "Differential regulation of the cynomolgus, human, and rat acyl-CoA oxidase promoters by PPARalpha." Gene 380(2):84-94. Klaunig, JE; Babich, MA; Baetcke, KP; Cook, JC; Corton, JC; David, RM; DeLuca, JG; Lai, DY; McKee, RH; Peters, JM; Roberts, RA; Fenner-Crisp, PA. 2003. "PPAR-alpha agonist-induced rodent tumors: Modes of action and human relevance." Crit. Rev. Toxicol. 33(6):655-780. Lake, BG. 1995. "Mechanisms of hepatocarcinogenicity of peroxisome-proliferating drugs and chemicals." Annu. Rev. Pharmacol. Toxicol. 35:483-507. Lau, C; Thibodeaux, JR; Hanson, RG; Narotsky, MG; Rogers, JM; Lindstrom, AB; Strynar, MJ. 2006. "Effects of perfluorooctanoic acid exposure during pregnancy in the mouse." Toxicol. Sci.90(2):510-518. Lau, C; Thibodeaux, JR; Hanson, RG; Rogers, JM; Grey, BE; Stanton, ME; Butenhoff, JL; Stevenson, LA. 2003. "Exposure to perfluorooctane sulfonate during pregnancy in rat and mouse. II. Postnatal evaluation." Toxicol. Sci. 74:382-392. Lee, CH; Olson, P; Evans, RM. 2003. "Minireview: Lipid metabolism, metabolic diseases, and peroxisome proliferator-activated receptors." Endocrinology 144(6):2201-2207. Leonard, RC. 2003. "Epidemiology Surveillance Report: Cancer Incidence for Washington Works Site, 1959-2001." January 27. Leonard, RC; Kreckmann, KH; Sakr, CJ; Symons, JM. 2008. "Retrospective cohort mortality study of workers in a polymer production plant including a reference population of regional workers."Ann. Epidemiol. 18(1):15-22. Ley, C. [3M]. 2014. Email to B. Beck (Gradient) re: PFOS Study. April 29. 13 G:\Projects\214032_PFCs\TextProc\r042914r.docx Loveless, SE; Finlay, C; Everds, NE; Frame, SR; Gillies, PJ; O'Connor, JC; Powley, CR; Kennedy, GL. 2006. "Comparative responses of rats and mice exposed to linear/branched, linear, or branched ammonium perfluorooctanoate (APFO)." Toxicology 220(2-3):203-217. Luebker, DJ; Case, MT; York, RG; Moore, JA; Hansen, KJ; Butenhoff, JL. 2005. "Two-generation reproduction and cross-foster studies of perfluorooctanesulfonate (PFOS) in rats." Toxicology 215(12):126-148. Lundin, JI; Alexander, BH; Olsen, GW; Church, TR. 2009. "Ammonium perfluorooctanoate production and occupational mortality." Epidemiology 20:921-928. MacPherson, IR; Bissett, D; Petty, RD; Tait, B; Samuel, LM; MacDonald, J; Smith, M; Birse-Archbold, JA; Barnett, AL; Wolf, CR; Elcombe, CR; Jeynes-Ellis, A; Evans, TRJ. 2011. "A first-in-human phase I clinical trial of CXR1002 in patients (pts) with advanced cancer." J. Clin. Oncol. 29(15(Suppl.)):3063. Macpherson, IRJ; Bissett, D; Tait, B; Samuel, L; MacDonald, J; Barnett, AL; Wolf, CR; Elcombe, CR; Jeynes-Ellis, A; Evans, TRJ. 2010. "A Phase I Clinical Trial of CXR1002 in Patients with Advanced cancer." In 22nd EORTC-NCI-AACR Symposium, 1p. Mandel, JS; Johnson, RA. 1995. "Final Report: Mortality Study of Employees at 3M Plant in Decatur, Alabama." University of Minnesota, School of Public Health. 49p., March 13. Mandel, JS; Schuman, LM. 1989. "Mortality Study at the 3M Chemolite Plant." Report to 3M Co. (St. Paul, MN), January. Millichap, JG. 2008. "Etiologic classification of attention-deficit/hyperactivity disorder." Pediatrics 121(2):e358-e365. Minata, M; Harada, KH; Karrmann, A; Hitomi, T; Hirosawa, M; Murata, M; Gonzalez, FJ; Koizumi, A. 2010. "Role of peroxisome proliferator-activated peceptor-alpha in hepatobiliary injury induced by ammonium perfluorooctanoate in mouse liver." Ind. Health 48(1):96-107 Nakamura, T; Ito, Y; Yanagiba, Y; Ramdhan, DH; Kono, Y; Naito, H; Hayashi, Y; Li, Y; Aoyama, T; Gonzalez, FJ; Nakajima, T. 2009. "Microgram-order ammonium perfluorooctanoate may activate mouse peroxisome proliferator-activated receptor alpha, but not human PPARalpha." Toxicology265(1-2):27-33. Olsen, G; Ehresman, D; Froehlich, J; Burris, J; Butenhoff, J. [3M Co.]. 2005. "Evaluation of the Half-life (T1/2) of Elimination of Perfluorooctanesulfonate (PFOS), Perfluorohexanesulfonate (PFHS) and Perfluorooctanoate (PFOA) from Human Serum." TOX017. Presented at Fluoros: An International Symposium on Fluorinated Alkyl Organics in the Environment, Toronto, Canada. 1p Olsen, GW; Burlew, MM; Marshall, JC; Burris, JM; Mandel, JH. 2004. "Analysis of episodes of care in a perfluorooctanesulfonyl fluoride production facility." J. Occup. Environ. Med. 46(8):837-846 Olsen, GW; Burris, JM; Burlew, MM; Mandel, JH. 2000. "Plasma cholecystokinin and hepatic enzymes, cholesterol and lipoproteins in ammonium perfluorooctanoate production workers." Drug Chem. Toxicol. 23(4):603-620. 14 G:\Projects\214032_PFCs\TextProc\r042914r.docx Olsen, GW; Burris, JM; Burlew, MM; Mandel, JH. 2003a. "Epidemiologic assessment of worker serum perfluorooctanesulfonate (PFOS) and perflurooctanoate (PFOA) concentrations and medical surveillance examinations." J. Occup. Environ. Med. 45(3):260-270. Olsen, GW; Burris, JM; Ehresman, DJ; Froehlich, JW; Seacat, AM; Butenhoff, JL; Zobel, LR. 2007. "Half-life of serum elimination of perfluorooctanesulfonate, perfluorohexanesulfonate, and perfluorooctanoate in retired fluorochemical production workers." Environ. Health Perspect. 115(9):1298-1305. Olsen, GW; Burris, JM; Mandel, JH; Zobel, LR. 1999. "Serum perfluorooctane sulfonate and hepatic and lipid clinical chemistry tests in fluorochemical production employees." J. Occup. Environ. Med. 41(9):799-806. Olsen, GW; Ehresman, DJ; Buehrer, BD; Gibson, BA; Butenhoff, JL; Zobel, LR. 2012."Longitudinal assessment of lipid and hepatic clinical parameters in workers involved with the demolition of perfluoroalkyl manufacturing facilities." J. Occup. Environ. Med. 54(8):974-983. Olsen, GW; Logan, PW; Hansen, KJ; Simpson, CA; Burris, JM; Burlew, MM; Vorarath, PP; Venkateswarlu, P; Schumpert, JC; Mandel, JH. 2003b. "An occupational exposure assessment of a perfluorooctanesulfonyl fluoride production site: Biomonitoring." AIHA J. 64:651-659. Olsen, GW; Mandel, JH. 2003. "Descriptive Analysis of Serum Fluorochemical Concentrations from Cottage Grove Employee Participants of the 2002 Medical Surveillance Program (Final Report)." 3M Company (St. Paul, MN) Medical Department, Epidemiology, 220-3W-05. June 17. Olsen, GW; Zobel, LR. [3M]. 2006. "An Analysis of the 2000 Flurorochemical (Perfluorooctanoate, PFOA) Medical Surveillance Program at 3M Company's Antwerp (Belgium), Cottage Grove (Minnesota), and Decatur (Alabama) Facilities (Final)." Submitted to US EPA Docket No. AR-226. 38p., May 16. Olsen, GW; Zobel, LR. 2007. "Assessment of lipid, hepatic, and thyroid parameters with serum perfluorooctanoate (PFOA) concentrations in fluorochemical production workers." Int. Arch. Occup. Environ. Health 81(2):231-246. Palazzolo, MJ. 1993. "13-Week Dietary Toxicity Study with T-5180, Ammonium Perfluorooctanoate (CAS No. 3825-26-1) in Male Rats (Final)." Report to 3M (St. Paul, MN), HWI 6329-100. 657p., May 17. Perkins, RG; Butenhoff, JL; Kennedy, GL; Palazzolo, MJ. 2004. "13-week dietary toxicity study of ammonium perfluorooctanoate (APFO) in male rats." Drug Chem. Toxicol. 27(4):361-378. Pohl, HR; Chou, CH. 2005. "Health effects classification and its role in the derivation of minimal risk levels: Hepatic effects." Regul. Toxicol. Pharmacol. 42:161-171. Sakr, CJ; Kreckmann, KH; Green, JW; Gillies, PJ; Reynolds, JL; Leonard, RC. 2007a. "Cross-sectional study of lipids and liver enzymes related to a serum biomarker of exposure (ammonium perfluorooctanoate or APFO) as part of a general health survey in a cohort of occupationally exposed workers." J. Occup. Environ. Med. 49(10):1086-1096. 15 G:\Projects\214032_PFCs\TextProc\r042914r.docx Sakr, CJ; Leonard, RC; Kreckmann, KH; Slade, MD; Cullen, MR. 2007b. "Longitudinal study of serum lipids and liver enzymes in workers with occupational exposure to ammonium perfluorooctanoate." J. Occup. Environ. Med. 49(8):872-879. Sakr, CJ; Symons, JM; Kreckmann, KH; Leonard, RC. 2009. "Ischaemic heart disease mortality study among workers with occupational exposure to ammonium perfluorooctanoate." Occup. Environ. Med. 66(10):699-703. Schuman, LM; Mandel, JS. 1980. ""An Epidemiologic Mortality Study of Employees at the Chemolite Plant." February. Seacat, AM; Thomford, PJ; Hansen, KJ; Clemen, LA; Eldridge, SR; Elcombe, CR; Butenhoff, JL. 2003. "Sub-chronic dietary toxicity of potassium perfluorooctanesulfonate in rats." Toxicology183:117131. Seacat, AM; Thomford, PJ; Jansen, KJ; Olsen, GW; Case, MT; Butenhoff, JL. 2002. "Subchronic toxicity studies on perfluorooctanesulfonate potassium salt in Cynomolgus monkeys." Toxicol. Sci. 68:249-264. Shen, DD. 2013. "Toxicokinetics." In Casarett and Doull's Toxicology: The Basic Science of Poisons (Eighth Edition). (Ed: Klaassen, CD), McGraw-Hill Education, New York, p367-390. Steenland, K; Woskie, S. 2012. "Cohort mortality study of workers exposed to perfluorooctanoic acid." Am. J. Epidemiol. 176(10):909-917. Stein, CR; Savitz, DA. 2011. "Serum perfluorinated compound concentration and Attention Deficit/Hyperactivity Disorder in children aged 5 to 18 years." Environ. Health Perspect. 119:1466-1471 Toxicology Science Advisory Council (TOXSAC). 2002. "Hepatocellular Hypertrophy." Office of Pesticides Programs, Health Effects Division (HED) Guidance Document # G0201. 24p., October 21. US EPA. 1991. "Guidelines for developmental toxicity risk assessment." Fed. Reg. 56(234):6379863826, December 5. US EPA. 1998. "Guidelines for neurotoxicity risk assessment." Fed. Reg. 63(93):26926-26954, May 14. US EPA. 2014a. "Health Effects Document for Perfluorooctanoic Acid (PFOA) (Draft)." Office of Water, Health and Ecological Criteria Division, EPA 822-R-14-001. February. US EPA. 2014b. 2014. "Health Effects Document for Perfluorooctane Sulfonate (PFOS) (Draft)." Office of Water, Health and Ecological Criteria Division, EPA 822-R-14-002. 208p., February. Williams, GM; Iatropoulos, MJ. 2002. "Alteration of liver cell function and proliferation: Differentiation between adaptation and toxicity." Toxicol. Pathol. 30(1):41-53. Wolf, CJ; Takacs, ML; Schmid, JE; Lau, C; Abbott, BD. 2008. "Activation of mouse and human peroxisome proliferator-activated receptor alpha (PPAR{alpha}) by perfluoroalkyl acids (PFAAs) of different functional groups and chain lengths." Toxicol. Sci. 106(1):162-171. Yang, Q; Nagano, T; Shah, Y; Cheung, C; Ito, S; Gonzalez, FJ. 2008. "The PPAR alpha-humanized mouse: A model to investigate species differences in liver toxicity mediated by PPAR alpha." Toxicol. Sci. 101(1):132-139. 16 G:\Projects\214032_PFCs\TextProc\r042914r.docx York, RG; Kennedy, GL Jr.; Olsen, GW; Butenhoff, JL. 2010. "Male reproductive system parameters in a two-generation reproduction study of ammonium perfluorooctanoate in rats and human relevance." Toxicology 271(1-2):64-72. Zhao, Y; Tan, YS; Haslam, SZ; Yang, C. 2010. "Perfluorooctanoic acid effects on steroid hormone and growth factor levels mediate stimulation of peripubertal mammary gland development in C57Bl/6 mice." Toxicol. Sci. 115(1):214-224. 17 G:\Projects\214032_PFCs\TextProc\r042914r.docx Tables G:\Projects\214032_PFCs\TextProc\r042914r.docx Table 1 Liver, Cardiovascular, and Metabolic Disease in Occupational Cohorts and Associated Serum Levels Location/ Chemical/N Liver Disease Results Cardiovascular Disease Results Metabolic Syndrome Results Study-specific Cohort Serum Levels (ng/mL unless otherwise noted) Cottage Grove N = 3,688 1947-1978 No increased liver cancer or cirrhosis- related mortality No increased mortality from heart disease PFOA Not available Not measured Cottage Grove N = 4,085 1947-1983 Cottage Grove N = 3,537 1947-1983 No increased liver cancer or cirrhosis- related mortality No increased liver cancer or cirrhosis- related mortality No increased mortality from heart disease No increased mortality from heart disease Not available No mortality from diabetes Not measured Not measured Cottage Grove N = 3,992 1947-1997 No increased liver cancer or cirrhosis- related mortality Cottage Grove N = 3,993 1947-2002 No increased liver cancer or cirrhosis- related mortality Parkersburg N = 5,523 1959-2001 No increased liver cancer or cirrhosis- related mortality No increased mortality from heart disease No increased mortality from "all" heart disease, including ischemic heart disease (IHD) Increased incidence of rheumatic heart disease, acute myocardial infarction, and atherosclerosis and aneurysm No deaths from diabetes mellitus in "definite" exposure group, but statistical increase in lower group with lower serum levels No deaths from diabetes mellitus in "definite" exposure group, but statistical increase in "probable" exposure group No significant increases in mortality Not measured 2,600-5,200 (Definite) 300-1,500 (Probable) Not measured General Facility Serum Levelsa (ng/mL) Reference Means:b 4,325-6,800 Range:b ND-114,100 Means:b 4,325-6,800 Range:b ND-114,100 Means:b 4,325-6,800 Range:b ND-114,100 Means:b 4,325-6,800 Range:b ND-114,100 Schuman and Mandel, 1980 Mandel and Schuman, 1989 Gilliland and Mandel, 1993 Alexander, 2001 Lundin et al., 2009 Leonard, 2003 G:\Projects\214032_PFCs\TextProc\r042914r.docx Page 1 of 2 Location/ Chemical/N Parkersburg N = 6,027 1948-2002 Parkersburg N = 4,747 1948-2002 Parkersburg N = 5,791 1952-2008 Decatur N = 2,166 1961-1991 Liver Disease Results Cardiovascular Disease Results Metabolic Syndrome Results No increased liver- related or cirrhosis- related mortality Not available No increased mortality or risk from "all" heart disease, including ischemic heart disease No increased mortality from IHD Increase in diabetes- related mortality Not available No increased liver cancer or liver disease-related mortality No increased liver cancer or cirrhosis- related mortality No increased mortality from ischemic heart disease No increased mortality from heart disease Increase in diabetes- related mortality PFOS No mortality from diabetes Study-specific Cohort Serum Levels (ng/mL unless otherwise noted) Not measured Range of cumulative exposure means: < 3,820 to > 6,780 ng/mL-years Range of cumulative exposure: 0 to > 2,700 ng/mL-years Not measured General Facility Serum Levelsa (ng/mL) Means: 1,290-2,440 Range: ND-12,830c Reference Leonard et al., 2008 Sakr et al., 2009 Steenland and Woskie, 2012 Mandel and Johnson, 1995 Decatur No increased liver No increased Not available Mean (geo.) for highest Alexander N = 2,083 cancer or cirrhosis- mortality from "all" exposed group = 2,000 et al., 2003 1961-1997 related mortality heart disease Decatur No increased liver Not available Not available 110-1,970 Grice et al., N = 1,400 cancer or liver 2007 1961-not disease reported Notes: (a) Excludes PFOA serum measurements based on organic or inorganic fluorine. (b) PFOA arithmetic means and ranges based on information in the following studies: Olsen et al., 2000; Gilliland and Mandel, 1996; Olsen and Zobel, 2007; Olsen and Mandel, 2003. (c) PFOS arithmetic means and ranges based on information in the following studies: Olsen et al., 1999; Olsen and Zobel, 2007; Olsen et al., 2003a,b. G:\Projects\214032_PFCs\TextProc\r042914r.docx Page 2 of 2 Table 2 Summary of Serum Lipid and Clinical Chemistry Results in Workers Location/ Chemical/N Serum Lipid Results Clinical Chemistry Results Cottage Grove N = 111 for 1993 N = 80 for 1995 N = 74 for 1997 Antwerp, Cottage Grove, Decatur Total N = 552 Antwerp (N = 196) Cottage Grove (N = 122) Decatur (N = 188) 2000 Antwerp, Cottage Grove, Decatur Total N = 506: Antwerp N = 196 Cottage Grove N = 122 Decatur N = 188 2000 Cottage Grove, Decatur N = 179 2008-2010 Parkersburg N = 454 1979-2004 No association with abnormal liver function tests, hypolipidemia, or cholestasis. PFOA Negative association with cholecystokinin. No association with total cholesterol or low-density lipoprotein (LDL). Inconsistent associations with high-density lipoprotein (HDL) and triglycerides by location. No association with total cholesterol or LDL. Inconsistent negative association with HDL and positive association with triglycerides by location. Small positive association with HDL. Negative association with ratio of total cholesterol/HDL. No associations with total or non-HDL cholesterol. Association with total cholesterol. No association with LDL, HDL, or triglycerides. Not available No associations with hepatic enzymes, TSH, or T4. Negative association with free T4 and positive association with T3, but values were within normal ranges. Negative association with bilirubin. No association with alkaline phosphatase or aspartate aminotransferase (AST). Negative association with bilirubin, positive association with AST. No association with gamma- glutamyl transferase (GGT), alanine aminotransferase (ALT), alkaline phosphatase. Cohort Serum Levels (ng/mL unless otherwise noted) 0-1,141,000 Not available 7-92,030 Mean = 881 (3M employees); 28.9 (contract workers) 0-22,660 Mean = 1,130 Reference Olsen et al., 2000 Olsen and Zobel, 2006 Olsen and Zobel, 2007 Olsen et al., 2012 Sakr et al., 2007b G:\Projects\214032_PFCs\TextProc\r042914r.docx Page 1 of 2 Location/ Chemical/N Parkersburg N = 1,025 Not specified - between 2004 and 2007 Trissino, Italy N = 53 Yearly 2000-2007 Decatur and Antwerp 1995 Antwerp N = 88 Decatur N = 90 1997 Antwerp N = 65 Decatur N = 84 Antwerp and Decatur Antwerp N = 255 Decatur N = 263 1998 Cottage Grove Decatur N = 179 (N = 14 3M employees; 165 contract workers) 2008-2010 Serum Lipid Results Positive association with total cholesterol, LDL, and very low-density lipoprotein (VLDL). No association with HDL. Slight but significant positive association with total cholesterol. No association with HDL or triglycerides. No association with cholesterol or lipoproteins at serum concentrations < 6,000 ng/mL. Clinical Chemistry Results Positive association with GGT. No association with bilirubin, AST, or ALT. Positive association with uric acid and alpha-2u-globulins. Negative association with bilirubin. PFOS No associations Cohort Serum Levels (ng/mL unless otherwise noted) 4.6-9,550 200-47,040 0 to > 6000 Reference Sakr et al., 2007a Costa et al., 2009 Olsen et al., 1999 Positive association with total cholesterol and triglycerides using multivariate regression but not by odds ratio. No association with HDL. No associations with total, HDL, non-HDL cholesterol, or ratio of total cholesterol/HDL. No associations No associations 10-7,040 Olsen et al., 2003a Mean = 404 (3M employees); Olsen et al., 2012 36.6 (contract workers) G:\Projects\214032_PFCs\TextProc\r042914r.docx Page 2 of 2