SAR Rpyjpw Draft; Hn Not Citf nr
              DRAFT RISK ASSESSMENT

     OF THE POTENTIAL HUMAN HEALTH EFFECTS

           ASSOCIATED WITH EXPOSURE TO

             PERFLUOROOCTANOIC ACID

                    AND ITS SALTS
             U.S. Environmental Protection Agency
             Office of Pollution Prevention and Toxics
                  Risk Assessment Division
                     January 4, 2005

-------
SAR RPVJPW Draft; Hn Nnf Cite nr Diinfp
                                   Table of Contents

Executive Summary                                                 6
1.0 Scope of the Assessment                                          11
2.0 Chemical Identity                                                12
2.1 Physicochemical Properties                                       12
3.0 Hazard Characterization                                          13
3.1 Epidemiology Studies                                            13
3.1.1 Mortality and Cancer Incidence Studies in Workers                 13
3.1.2 Hormone Study in Male Workers                                17
3.1.3 Occupational Study on Episodes of Care                          18
3.1.4 3M Medical Surveillance Studies                                20
3.1.4.1 Antwerp and Decatur Plants-Cross-Sectional
       and Longitudinal Studies                                      20
3.1.4.2 Cottage Grove Plant-Cross-sectional Studies of
       Clinical Chemistries and CCK                                 21
3.2 Metabolism and Pharmacokinetics                                 24
3.2.1 Metabolism and Pharmacokinetic Studies in Humans               24
3.2.2 Metabolism and Pharmacokinetic Studies in Non-Human Primates   24
3.2.3 Metabolism and Pharmacokinetic Studies in Adult Rats            26
3.2.3.1 Absorption Studies                                            26
3.2.3.1.1 Oral Exposure                                              26
3.2.3.1.2 Inhalation Exposure                                         27
3.2.3.1.3 Dermal Exposure                                           27
3.2.3.2 Serum Pharmacokinetic Parameters in Adult P.ats                 28
3.2.3.2.1 Oral and Intravenous Exposure in Sprague-Dawley Rats         28
3.2.3.2.2 Intravenous Exposure in Wistar Rats                          30
3.2.3.3 Distribution Studies in Adult Rats                              31
3.2.3.3.1 Oral Exposure                                              31
3.2.3.3.2 Intravenous Exposure                                       34
3.2.3.3.3 Intraperitoneal Exposure                                     34
3.2.3.4 Metabolism Studies in Adult Rats                               36
3.2.3.5 Elimination Studies in Adult Rats                               37
3.2.3.5.1 Enterohepatic Circulation                                    37
3.2.3.5.2 General Elimination Studies                                  37
3.2.3.5.2.1 Oral Exposure                                            37
3.2.3.5.2.2 Intravenous Exposure                                      37
3.2.3.5.3  Elimination Studies in the Pregnant Rat                       38
3.2.3.5.4 Studies on the Mechanism of the Gender Difference
       in Elimination in Adult Rats                                   38
3.2.4 Metabolism and Pharmacokinetic Studies in Immature Rats         40
3.2.4.1 PFOA Levels During Pregnancy and Lactation                       40
3.2.4.2 PFOA Levels in the Postweaning Rat                           42
3.2.4.3 Serum and Tissue Distribution in Immature
       Wistar Rats Following Oral Exposure                           43
3.2.5 Comparative Studies of Protein Binding in Humans,
       Non-Human Primates, and Rats                                44
3.2.6 Metabolism and Pharmacokinetic Studies in Other Test Species     45
3.3 Acute Toxicity Studies in Animals                                46
3.4 Mutagenicity Studies                                            47
3.5 Repeat Dose Studies in Animals                                   47
3.5.1 Subchronic Studies in Non-Human Primates                       47

-------
SAR RPVJPW Draft? Fin Nnf Pifp nt-
3.5.2 Subchronic Studies in Rodents                                  5 2
3.5.3 Chronic Toxicity and Carcinogenicity Studies in Rats              55
3.6 Immunotoxicity Studies in Mice                                  59
3.7 Prenatal Developmental Toxicily Studies in Animals                60
3.8 Reproductive Toxicity Studies in Animals                         64
3.8.1 FO Generation                                                66
3.8.1.1FO Males                                                  66
3.8.1.2 FO Females                                                67
3.8.2 Fl Generation                                                68
3.8.2.1 Fl Males                                                  68
3.8.2.2 Fl Females                                                69
3.8.3 F2 Generation                                                71
3.8.4 Conclusions                                                  71
3.9 Mode of Action and Summary of Weight of Evidence               73
3.9.1 Epidemiology Studies                                         73
3.9.2MetabotismandPharmacokinetics                              74
3.9.3 Mode of Action Analyses and Cancer Descriptor                 75
3.9.3.1 Mode of Action Analysis of Liver Toxicity
       and Liver Adenomas in Rats                                  75
3.9.3.2 Human Relevance of the Rat  PPARcc-agonist Induced
       Liver Toxicity and Liver Adenomas                           80
3.9.3.2 Leydig Cell Adenomas in Rats                                81
3.9.3.3 Pancreatic Acinar Cell Tumors in Rats                         82
3.9.3.4 Cancer Descriptor                                          83
3.9.4 Toxicity in Adult Repeat-Dose  Animal Studies                   84
3.9.4.1 Non-Human Primates                                        84
3.9.4.2 Adult Male Rats                                            85
3.9.4.3 Adult Female Rats                                          87
3.9.4.4 Adult Mice                                                87
3.9.5 Developmental and Reproductive Toxicity in Animal Studies       88
4.0 Biomonitoring Data                                           90
4.1 Occupational Exposures                                        90
4.1.1 3M Occupational Data                                        90
4.1.2 DuPont Occupational Data                                    92
4.2 General Population Exposures                                   92
5.0 Risk Assessment                                               95
5.1 Selection of Endpoints                                          95
5.2 Use of Serum Levels as a Measure of Internal Dose in Humans       97
5.2.1 General Population                                           97
5.2.2 Workers                                                     97
5.3 Calculation of MOEs Based  on Non-Human Primate Studies         98
5.4 Calculation of MOEs Based  on Adult Rat Studies                   98
5.5 Calculation of MOEs Based  on Rat Developmental Toxicity Studies  99
5.6 Uncertainties in the Risk Characterization                         101
6.0 Overall Conclusions                                            103
7.0 References                                                    105
Appendix A                                                      A-l

-------
SAB Review Draft! Tin Nnf fitp nr
ALT
AP
APFO
AST

AUC

BMI
BUN

CCG
CCK
CI
Cl
CLR
ConA

DEEP
DHEAS

EOF

FMPP
FID
FSH

GD
GOT
           Glossaiy of Abbreviations

alanine aminotransferase, serum glutamyl pyruvic transaminase, SGPT
alkaline phosphatase
ammonium salt of perfluorooctanoic acid
aspartate aminotransferase, senim glutamyl oxaloacetic transaminase,
SGOT
area under the curve

body mass index
blood urea nitrogen test

Clinical Care Groups software
cholecystokinin-33 (human); cholecystokinin (rat)
confidence interval
total body clearance
renal clearance
concanavalin A

di(2-ethylhexyl)phthalate
dehydroepiandrosterone sulfate

epidermal growth factor

Familial Male-limited Precocious Puberty
flame ionization detector
follicle stimulating hormone

gestation day
gamma glutamyl transferase
HDL               high-density lipoprotein
17-HP              17 gamma-hy droxy progesteron e
HPLC/ESMS       high performance liquid chromatography/electrospray tandem mass
                   spectrometry
HRBC             horse red blood cell
HSA               human serum albumin

i.p.                intraperitoneal
i.v.                intravenous

Kj                 dissociation  constant

LD                lactation day
LDL               low-density lipoprotein
LH                luteinizing hormone
LLOQ             lower limit of quantitation
LOAEL            lowest-observed-adverse-effect level
LD50              lethal dose 50%
LPS               lipopolysaccharide

-------
SAR RPVJPW Draft; f>n Not Citf nr
M570              N-methyl perfluorooctanesulfonamidoacetate
MOA              mode of action

NMR              nuclear magnetic resonance
NOAEL            no-observed-adverse-effect level

OAT               organic anion transporter
OPPT              Office of Pollution Prevention and Toxics

PACT              pancreatic acinar cell tumors
PFHS              perfluorohexanesulfonate
PFOA              perfluorooctanoic acid
PFOS              perfluorooctane sulfpnate
PFC               plaque forming cell immune function test
PPARa            peroxisome proliferator-activated receptor a
ppb                parts per billion
ppm               parts per million

RREpC            relative risk ratio for each episode of care
RSA               rodent serum albumin

SEM               standard error of the mean
SCOT              serum glutamyl oxaloacetic transaminase, AST
SGPT              serum glutamyl pyruvic transaminase, ALT
SHBG             sex hormone-binding globulin
SIR               standardized incidence ratio
SMR               standardized mortality ratio

tyorTi/2           half life
TSH               thyroid-stimulating hormone

Vd                volume of distribution

-------
SAR RPVJPW Draft; Dn Nnt rite nr Oiintp
                                 Executive Summary

As part of the effort by the Office of Pollution Prevention and Toxics (OPPT) to understand
health and environmental issues presented by fluorochemicals in the wake of unexpected
toxicological and bioaccumulation discoveries with respect to perfluorooctane sulfonates
(PFOS), OPPT has been investigating perfluorooctanoic acid (PFOA) and its salts.  PFOA and
its salts are fully fluorinated organic compounds that can be produced synthetically or through
the degradation or metabolism of other fluorochemical products. PFOA is primarily used as a
reactive intermediate, while its salts are used as processing aids in the production of
fluoropolymers and fluoroelastomers and in other surfactant uses.  PFOA and its salts are
persistent in the environment. Most of the toxicology studies have been conducted with the
ammonium salt of perfluorooctanoic acid, which is referred to as APFO in this report.

Human Health Effects

Epidemiological studies on the effects of PFOA in humans have been conducted on workers.
Most of the studies were cross-sectional and focused primarily on males.  Developmental
outcomes have not been studied. A retrospective cohort mortality study demonstrated a
statistically significant association between prostate cancer mortality and employment duration
in the chemical facility of a plant that manufactures PFOA. However, in an update to this study
in which more specific exposure measures were used, a significant association for prostate
cancer was not observed. Other mortality studies lacked  adequate exposure data which could be
linked to health outcomes. A study which examined hormone levels in workers reported an
increase in estradiol levels in workers with the highest PFOA serum levels; however, these
results may have been confounded by body mass index. Cholesterol and triglyceride levels  in
workers were positively associated with PFOA exposures, which is inconsistent with the
hypolipidemic effects observed in rat studies.  A statistically significant positive association was
reported for PFOA and T3 levels in workers but not for any other thyroid hormones.

Little information is available concerning the  pharmacokinetics of PFOA and its salts in humans.
An ongoing 5-year, half-life study in 7 male and 2 female retired workers has suggested a mean
serum PFOA half-life of 4.37 years (range, 1.50 - 13.49 years). Gender differences in
elimination of PFOA have not been observed in humans based on the limited data available in
the half-life study in retired workers. Metabolism and pharmacokinetic studies in non-human
primates are limited to a study of 3 male and 3 female cynomolgus monkeys administered a
single i.v. dose of 10 mg/kg potassium PFOA. In male monkeys, the average serum half-life was
20.9 days. In female monkeys, the average serum half-life was 32.6 days.

Studies in adult rats have shown that the ammonium salt  of PFOA (APFO) is absorbed following
oral, inhalation and dermal exposure.  Serum  pharmacokinetic parameters and the distribution of
PFOA has been examined in the tissues of adult rats following administration by gavage and by
i.v. and i.p. injection.  PFOA distributes primarily to the liver, serum, and kidney, and to a lesser
extent, other tissues of the body. It does not partition to the lipid fraction or adipose tissue.
PFOA is not metabolized and there is evidence of enterohepatic circulation of the compound.
The urine is the major route of excretion of PFOA in the  female rat, while the urine and feces are
both main routes  of excretion in male rats.

There are gender differences in the elimination of PFOA  in adult rats following administration
by gavage and by i.v.  and i.p.injection. In female rats, following oral administration, estimates
of the serum half-life were dependent on dose and ranged from approximately 2.8 to 16 hours,

-------
SAR Rpvifw Draft: Do Nnf fitf nr Onnte
while in male rats estimates of the serum half-life following oral administration were
independent of dose and ranged from approximately 138 to 202 hours. In female rats,
elimination of PFOA appears to be biphasic with a fast phase and a slow phase. The rapid
excretion of PFOA by female rats is believed to be due to active renal tubular secretion (organic
acid transport system); this renal tubular secretion is believed to be hormonally controlled.
Hormonal changes  during pregnancy  do not appear to cause a change in the rate of elimination in
rats.

Several recent studies have been conducted to examine the kinetics of PFOA in the developing
Sprague-Dawley rat. These studies have shown that PFOA readily crosses the placenta and is
present in the breast milk of rats. During lactation and for the first several weeks after weaning,
the elimination of PFOA is similar in males and females pups. Between  4-5 weeks of age, the
elimination in male rats assumes the adult pattern and the gender difference becomes readily
apparent. Distribution studies in the postweaning rat have shown that PFOA is distributed
primarily to the serum, liver, and kidney.

In acute toxicity studies in animals, the oral LD50 values for CD rats were >500 mg/kg for males
and 250-500 mg/kg for females, and <1000 mg/kg for male and female Wistar rats.  There was
no mortality following inhalation exposure of 18.6 mg/L for one hour in  rats. The dermal LD50
in rabbits was determined to be greater than 2000 mg/kg. APFO is a primary ocular irritant in
rabbits, while the data regarding potential skin irritancy are conflicting.

APFO  is not mutagenic. APFO did not induce mutation in either S. typhimurium or E.  coli when
tested either with or without mammalian activation.  APFO did not induce gene mutation when
tested with or without metabolic activation in the K-l line of Chinese hamster ovary (CHO) cells
in culture. APFO did not induce chromosomal aberrations in human  lymphocytes when tested
with and without metabolic activation up to cytotoxic concentrations. APFO was tested twice
for its ability to induce chromosomal aberrations in CHO cells. In the first assay, APFO induced
both chromosomal  aberrations and pplyploidy in both the presence and absence of metabolic
activation.  In the second assay, no significant increases in chromosomal aberrations were
observed without activation. However, when tested with metabolic activation, APFO induced
significant increases in chromosomal aberrations and in polyploidy. APFO was negative in a
cell transformation assay in C3H 10T1/4 mouse embryo fibroblasts and in the mouse micronucleus
assay.

Repeat-dose studies have been conducted in non-human primates. In a 13-week study with
Rhesus monkeys, exposure to doses of 30 mg/kg-day or higher resulted in death. Clinical signs
of toxicity were noted at doses as low as 3 mg/kg-day. Unlike rodent studies, analyses  of the
serum and liver levels did not reveal a gender difference in monkeys, but the sample size was
very small. In a 6-month study of male cynomolgus monkeys, there was a steep dose response
curve for mortality. Increases in liver weight were noted at doses  as low as 3 mg/kg-day, but
there was no evidence of peroxisome proliferator-activated receptor alpha activity (PPARcc).

Repeat-dose studies in rats and mice demonstrated that the liver is the primary target organ. Due
to gender differences in elimination, adult male rats exhibit effects at lower administered doses
than adult female rats.  Dietary exposure to APFO for 90 days resulted in significant increases in
liver weight and hepatocellular hypertrophy in female rats at 1000 ppm (76.5 mg/kg-day) and in
male rats at doses as low as 100 ppm (5 mg/kg-day). Chronic dietary exposure of rats to 300
ppm(males, 14.2 mg/kg-day; females, 16.1 mg/kg-day) APFO for 2 years resulted in increased
liver weight, hepatocellular hypertrophy, hematological effects, and testicular masses in males;

-------
SAR RPVJPW Draft; Dn Not rite nr
and reductions in body weight and hematological effects in females.

The carcinogenic potential of PFOA has been investigated in two dietary carcinogen:city studies
in rats. Under the conditions of these studies, there is some evidence that PFOA is carcinogenic,
inducing liver tumors, Ley dig cell tumors (LCT), and pancreatic acinar cell tumors (PACT) in
male Sprague-Dawley rats.  The evidence for mammary fibroadenomas in the female rats is
equivocal since the incidences were comparable to some historical background incidences.
There is sufficient evidence to indicate that PFOA is a PPARa-agonist and that the liver
carcinogenicity (and toxicity) of PFOA is mediated by binding to the PPARcc in the liver.  A
mode of action analysis has demonstrated that the hepatic effects are due to PPARa-agonism,
and that this mode of action is unlikely to occur in humans. The mode(s) of action for the
Leydig cell and pancreatic acinar cell tumors have been investigated, but there is insufficient
evidence to link these modes of action to PPARa.  The LCT and PACT induced in the rat by
PFOA probably do not represent a significant cancer hazard for humans because of quantitative
differences in the expressions of LH and  CCKA receptors and of other toxicodynamic differences
between the rat and the human. Based on no adequate human studies and uncertain relevance of
the tumors from the rat studies, PFOA may be best described as "suggestive evidence of
carcinogenicty, but not sufficient to assess human carcinogenic potential" under the draft 1999
EPA Guidelines for Carcinogen Risk Assessment.

PFOA appears to be immunotoxic in mice.  Feeding C57B1/6 mice a diet containing 0.02%
PFOA resulted in adverse effects to both  the thymus and spleen. In addition, this feeding
regimen resulted in suppression of the specific humoral immune response to horse red blood
cells, and suppression of splenic lymphocyte proliferation. The suppressed mice recovered their
ability to generate a humoral immune response when they were fed a diet devoid of PFOA.
Studies using transgenic mice showed that the PPARa was involved in causing the adverse
effects to the immune system.

There was no evidence of prenatal developmental toxicity in rats after oral exposure to doses as
high as 150 mg/kg-day.  Maternal toxicity was seen at 100 mg/kg-day. In a rabbit oral prenatal
developmental toxicity study there was a significant increase in skeletal variations after exposure
to 50 mg/kg-day APFO. There was no evidence of maternal toxicity at 50 mg/kg-day, the
highest dose tested.

A variety of endpoints were evaluated throughout different lifestages in a two-generation
reproductive toxicity study in rats exposed to 0,1, 3, 10, and 30 mg/kg/day APFO. In that study,
a reduction in Fl  pup mean body weight  on a litter basis was observed during lactation (sexes
combined) in the 30 mg/kg-day group. Fl male pups in the 10 and 30 mg/kg-day groups
exhibited a significant reduction in body  weight gain during days 8-50 postweaning, and body
weights were significantly reduced in the 10 mg/kg-day group beginning on postweaning day 36,
and in the 30 mg/kg-day group beginning on postweaning day 8. Fl female pups in the 30
mg/kg-day group exhibited a significant reduction in body weight gain on days 1-15
postweaning, and in body weights beginning on day 8 postweaning.  Reproductive indices were
not affected in the Fl animals. There was a significant increase in mortality mainly during the
first few days after weaning, and a significant delay in the timing of sexual maturation for Fl
male and female pups in the  30 mg/kg-day group.  No effects were observed in the F2 pups.
However, it should be noted that the F2 pups were sacrificed at weaning, and thus it was not
possible to ascertain the potential post-weaning effects that were noted in the Fl generation.
Adult systemic toxicity consisted of reductions in body weight in both the FO and Fl animals.

-------
SAR Review Draft; F>n Nnf fite nr
Human Biomonitoring

While the environmental concentrations and pathways of human exposure to PFOA and its salts
are unknown, there are data on PFOA serum levels in workers and the general population.
PFOA has been measured in the serum of workers occupationally  exposed to perfluorinated
chemicals for many years.  PFOA has also been detected recently  in the serum of the general
U.S. population, but at much lower levels than those reported in occupational biomonitoring
studies.  Individual blood serum samples from 3 separate non-occupational cohorts have been
analyzed for PFOA. Cohorts of adults (n= 645) and children (n=598) from various geographic
areas of the U.S. and an elderly cohort from Seattle (n = 238) have indicated consistent mean
serum levels of PFOA (approximately 5 ng/ml or 5 ppb). The reports indicate that serum levels
for most of the individuals in these samples are below 10 ng/ml; however, some of the levels are
as high as 56 ng/ml, indicating that a small number of individuals  are being exposed at higher
concentrations than the rest of the general population.

Risk Assessment and Uncertainties

A margin of exposure (MOE) approach can be used to describe the potential for human health
effects associated  with exposure to a chemical. The MOE is calculated as the ratio of the
NOAEL or LOAEL for a specific endpoint to the estimated human exposure level. The specific
endpoint may be from an epidemiology study or an animal toxicology study. The MOE does not
provide an estimate of population risk, but simply describes the relative "distance" between the
exposure level and the NOAEL or LOAEL.  In this risk assessment there is no information on
the sources or pathways of human exposure. However, serum levels of PFOA, which are a
measure of cumulative exposure, were available from human biomonitoring studies. In addition,
serum levels of PFOA were available for many of the animal toxicology studies or there was
sufficient pharmacokinetic information to estimate serum levels. Thus, in this assessment
internal doses from animal and human studies were compared; this is somewhat analogous to a
MOE approach which uses external exposure estimates.                      *

The results of existing epidemiology studies are not adequate for use in quantitative risk
assessment, and therefore the analysis was restricted to endpoints  in the animal toxicology
studies.  MOEs were calculated for the general U.S. population. Although some serum level
data were available for workers, the data were not adequate to calculate MOEs for occupational
exposures. In general, the mean serum levels following occupational exposures appear to be
orders of magnitude higher than observed in the general population. Thus, MOEs for workers
would be expected to be  much less than for the general population.

A variety of endpoints from the animal toxicology studies were used to calculate MOEs for this
draft risk assessment. The endpoints encompassed different species, gender and life stages. For
this draft assessment, specific recommendations on the most appropriate
endpoint/lifestage/species/gender have not been made; rather,  all have been presented to provide
transparency.

For adults, two sets of MOEs were calculated based on the toxicology studies in non-human
primates and rats.  First,  MOEs were calculated from the cynomolgus monkey study and are
based on increased liver weight and possible mortality.  The MOE using the geometric mean for
the human serum level is 16,739 (8,191 for the 90th percentile). Second, MOEs were calculated
from the adult rat  studies and are based on reductions in body weight. MOEs were calculated
separately for the  female and male rat due to the gender differences in pharmacokinetics in this

-------
SAR RPVJPW Draff; Dn Nnf fite nr
species. MOEs were calculated by dividing the AUC in the adult female rat by the AUC for the
adult humans which is 398 (195 for the 9(r percentile) and by dividing the AUC for the adult
male rat by the AUC for the adult humans which is 9158 (4481 for the 90th percentile).

MOEs were calculated for the developmental effects in the two-generation reproductive toxicity
study in rats.  These effects were observed at various times during the maturation of the Fl pups.
For both Fl males and females there were reductions in body weight during lactation; significant
increases in mortality during the first few days after weaning; and significant delays in the
timing of sexual maturation. Mean body weights were also significantly reduced in the time
period prior to sexual maturation in both the Fl males  and females.  The critical period of
exposure for each of the effects is not known. For example, it is not known whether prenatal
and/or lactational exposure is important for the reduced body weight that was observed during
lactation. Similarly, it is not known whether the reduced body weight, mortality, or delayed
sexual maturation that occurred during the postweanirig period are due to prenatal, lactational,
and/or postweaning exposures. Ideally, MOEs would be calculated for each of these exposure
periods; however MOEs were not calculated for the lactation period due to uncertainties in
pharmacokinetics.

For the prenatal period, MOEs were calculated for the  pregnant human female. MOEs were not
calculated for the fetus since there is no information on human serum levels in fetuses. MOEs
were calculated using both Cmax and AUC;  the MOE based on C   is 3,095 (1548 for the 90th
percentile) and the MOE based on the AUC is 823 (412 for the 9
-------
SAR RPVJPW Draft; fin Not Cite nr Qnnfp
1.0 Scope of the Assessment

As part of the effort by the Office of Pollution Prevention and Toxics (OPPT) to understand
health and environmental issues presented by fluorochemicals in the wake of unexpected
toxicological and bioaccumulation discoveries with respect to perfluorooctane sulfonates
(PFOS), OPPT has been investigating perfluorooctanoic acid and its salts (PFOA).  PFOA and
its salts are fully fluorinated organic compounds that can be produced synthetically or through
the degradation or metabolism of other fluorochemical products. PFOA is primarily used as a
reactive intermediate, while its salts are used as processing aids in the production of
fluoropolymers and fluoroelastomers and in other surfactant uses.

OPPT released a preliminary Draft Hazard Assessment of Perfluorooctanoic Acid and Its Salts,
dated February 20, 2002, on March 28, 2002, and issued a minor correction to that document on
April 15, 2002.  That draft assessment indicated that PFOA and its salts are persistent in the
environment and in humans with a half life of years. The assessment noted the potential
systemic toxicity and carcinogenicity associated with the ammonium salt of PFOA (APFO),
which has been the focus of the animal toxicology studies, and observed that blood monitoring
data suggested widespread exposure to the general population, albeit at low levels.  The Agency
has since received considerable additional animal toxicology data on APFO that suggest a
potential for developmental/reproductive toxicity and immunotoxicity, and additional human
biomonitoring data that indicate low level exposures to the general population that cannot be
explained at mis time.

On September 27, 2002, the Director of OPPT issued a memorandum announcing that OPPT
would initiate a priority review to determine whether PFOA and its salts meets the criteria for
action under section 4(f) of the Toxic Substances Control Act. As  part of the priority review, the
hazard assessment was revised and released on September 30, 2002. Another revision was then
released November 4, 2002. In addition, OPPT released a Preliminary Risk Assessmment of the
Developmental Toxicity of PFOA and Its Salts, dated March 4, 2003. Although there is a wide
range of toxicological endpoints associated with exposure to APFO, the initial scope of the
preliminary risk assessment focused only on the endpoints that are included in section 4(f); these
include cancer,  mutations, and birth defects. OPPT did not include gene mutations in the
preliminary risk assessment since APFO is not known to be mutagenic. In addition, APFO is a
peroxisome proliferator-activated receptor a (PPARa) agonist and through this mode of action
could lead to the formation of liver tumors in rodents.  The relevance of this mode of action for
humans is currently under scientific debate, and the Agency was engaged in activities to resolve
this issue. Therefore, the preliminary risk assessment was narrowly restricted to examine only
the potential risks of developmental toxicity.

Since the release of the preliminary risk assessment, several pieces of information have become
available. Several uncertainties regarding the pharmacokinetics of PFOA during development
were discussed in the preliminary risk assessment, and studies have now been conducted to
address these uncertainties. In addition, activities aimed  at addressing the uncertainties
associated with the PPARa-agonist mode of action have made significant progress, and a draft
Proposed OPPTS Science Policy paper was presented to the FIFRA Science Advisory Panel on
December 9, 2003. This new information has enabled OPPT to extend the assessment activities
such that this draft Risk Assessment of the Potential Human Health Effects Associated With
Exposure to PFOA and Its Salts considers all toxicological endpoints to the extent possible.

The  relevant information pertaining to chemical properties, epidemiology, pharmacokinetics and

                                          11

-------
SAB Review Draft; Dn Nnf Tifp nr Quote
metabolism, animal toxicology studies, and human exposure have been included in this draft risk
assessment.  Other information pertaining to ecotoxicity, production and uses, fate and transport,
and environmental monitoring are avilable in the Draft Hazard Assessment ofPerfluorooctanoic
Acid and Its Salts, dated November 4, 2002.

2.0 Chemical Identity

Chemical Name: Perfluorooctanoic Acid
Molecular formula: C8 H F15 02

Structural formula: F-CF2-CF2-CF2-CF2-CF2-CF2-CF2-C(=0)-X,

The free acid and some common derivatives have the following CAS numbers:
The perfluorooctanoate anion does not have a specific CAS number.

Free Acid            (X = OM+; M = H)         [335-67-1]
Ammonium Salt     (X = OM+; M = NH4)       [3825-26-1]
Sodium Salt         (X = OM+; M = Na)        [335-95-5]
Potassium Salt       (X = OM+; M = K)         [2395-00-8]
Silver Salt           (X = OM+; M = Ag)        [335-93-3]
Acid Fluoride        (X = F)                    [335-66-0]
Methyl Ester        (X = CH3)                 [376-27-2]
Ethyl Ester          (X = CH2-CH3)            [3108-24-5]

Synonyms:   1-Octanoic  acid, 2,2,3,3,4,4,5,5,6,6,7,7,8,8,8-pentadecafluoro-PFOA

2.1 Physicochemical Properties

PFOA is a completely fluorinated organic acid.  The typical structure has a linear chain of eight
carbon atoms. The physical chemical properties noted below are for the free acid, unless
otherwise stated. The data for the free acid, pentadecafluorooctanoic acid [335-67-1], is the
most complete.  The reported vapor pressure of 10 mm Hg appears high for a low melting solid
when compared to other low melting solids (chloroacetic acid: solid; MP = 61 to 63 °C; BP =
189 °C; VP = 0.1 kPa [0.75 mmHg] @ 20 °C; NIOSH, 1994), but is consistent with other
perfluorinated compounds with similar boiling points (perfluorobutanoic acid BP = 120 "C, VP
10 mm Hg @ 20 °C; Boit,  1975).  Another explanation may be that the 10 mm vapor pressure
was measured at an elevated temperature (but the temperature inadvertently omitted), as
perfluorooctanoic acid is typically handled as a liquid at 65 °C (3M, 2001). The free acid is
expected to completely dissociate in water,  leaving the anionic carboxylate in the water and the
perfluoroalkyl chain on the surface. In aqueous solutions, individual molecules of PFOA anion
loosely associate on the water surface and partition between the air / water interface. Several
reports note that PFOA salts self-associate at the surface, but with agitation they disperse and
micelles form at higher concentrations (Simister et al., 1992; Calfours and Stilbs, 1985; Edwards
et al., 1997). Water solubility has been reported for PFOA, but it is unclear whether these values
are for a microdispersion of micelles, rather than true solubility.   Due to these same surface-
active properties of PFOA, and the test protocol for the OECD shake flask method, PFOA is
anticipated to form multiple layers in octanol/water, much like those observed for PFOS.
Therefore, an n-octanol/water partition coefficient cannot be determined.

The available physicochemical properties for the PFOA free acid are:

                                          12

-------
SAR RPVJPW Draft; f>n Nnf fite nt-
Molecular weight: 414 (Boit, 1975)
Melting point: 45 - 50 °C (Boit, 1975)
Boiling point: 189 - 192 °C / 736 mmHg (Boit, 1975)
Vapor pressure: 10 mmHg @ 25 'C (approx.) (Exfluor, 1998)
Water solubility: 3.4 g/L (telomeric [MP - 34 °C ref. 0.01 - 0.02 mol/L ~4 - 8 g/L) (MSDS from
Merck, and Fischer, 2003)
pKa: 2.5 (Gilliland, 1992)
pH (Ig/L): 2.6 (MSDS Merck)

The PFOA derivative  of greatest concern and most wide spread use is the ammonium salt
(APFO; CAS No. 3825-26-1).  The water solubility of APFO has been inconsistently reported.
One 3M study reported the water solubility of APFO to be > 10%.  It was noted in an earlier
study that at concentrations of 20 g/L, the solution "gelled" (Welter, 1979).  These numbers
seem surprisingly low for a salt in light of Apollo Scientific selling a 31% aqueous solution of
APFO. Reported values for estimated partition coefficient (log Pow) of APFO do not agree.
The anticipated formation of an emulsified layer between the octanol and water surface interface
would make determination of log Kow impossible.

Determination of the vapor pressure of APFO is complicated.  A vapor pressure of 7 x 10"5 mm
Hg at 20 °C has been reported for APFO; however, this appears to be too low for a material that
sublimates as the ammonium salt (Wolter, 1993).  The ammonium salt begins to sublimate at
130 °C. As the temperature increases, 20% of the sample weight of APFO is lost by 169 °C.
Other salts (Cs, K, Ag, Pb, Li) do not demonstrate similar weight loss until 237 °C or higher
(Lines and Sutcliffe, 1984). Decomposition of different salts produces perfluoroheptene (loss of
metal fluoride and carbon dioxide). This occurs at 320 °C for the sodium salt and at 250-290 °C
for the silver salt (Boit, 1975).

3.0 Hazard Characterization

3.1 Epidemiology Studies

3M and Duponthave  conducted several epidemiology and medical surveillance studies of the
workers at their plants in various cities of the U.S. No remarkable health effects that can be
directly attributed to PFOA exposure have been reported in fluorochemical production workers
described in the studies below. (Serum PFOA concentrations in workers volunteering in these
biomonitoring programs are presented in Table 24, Section 4.1.1 and Table 25, Section 4.1.2).

3.1.1  Mortality and Cancer Incidence Studies in Workers

A retrospective cohort mortality study was performed on employees at the 3M Cottage Grove,
MN plant which produces APFO (Gilliland and Mandel, 1993). At this plant, APFO production
was limited to the Chemical Division. The cohort consisted of workers who had been employed
at the plant for at least 6 months between January 1947 and December 1983. Death certificates
of all of the workers were obtained to determine cause of death. There was almost complete
follow-up (99.5%) of all of the study participants. The exposure status of the workers was
categorized based on their job histories.  If they had been employed for at least 1 month in the
Chemical Division, they were considered exposed. All others were considered to be not exposed
to PFOA.  The number of months employed in the Chemical Division provided the cumulative
exposure measurements.  Of the 3537 (2788 men and 749 women) employees who participated
in this study, 398 (348 men and 50 women) were deceased. Eleven of the 50 women and 148 of

                                          13

-------
SAR RPVIPW Draft; Fin Nnt fite nr
the 348 men worked in the Chemical Division, and therefore, were considered exposed to PFOA.

Standardized Mortality Ratios (SMRs), adjusted for age, sex, and race were calculated and
compared to U.S. and Minnesota white death rates for men.  For women, only state rates were
available. The SMRs for males were stratified for 3 latency periods (10,15, and 20 years) and 3
periods of duration of employment (5,10, and 20 years).

For all female employees, the SMRs for all causes and for all cancers were less than 1. The only
elevated (although not significant) SMR was for lymphopoietic cancer, and was based on only 3
deaths. When exposure status was considered, SMRs for all causes of death and for all cancers
were significantly lower than expected, based on the U.S. rates, for both the Chemical Division
workers and the other employees of the plant.

In all male workers at the plant, the SMRs were close to 1 for most of the causes of death when
compared to both the U.S. and the Minnesota death rates. When latency and duration of
employment were considered, there were no elevated SMRs. When employee deaths in the
Chemical Division were compared to Minnesota death rates, the SMR for prostate cancer for
workers in the Chemical Division was  2.03 (95% CI .55 - 4.59). This was based on 4 deaths
(1.97 expected). There was also a statistically significant (p = 0.03) association with length of
employment in the Chemical Division  and prostate cancer mortality.  Based on the results of
proportional hazard models, the relative risk for a 1-year increase in employment in the
Chemical Division was  1.13 (95% CI 1.01 to 1.27). It rose to 3.3 (95% CI  1.02 -10.6) for
workers employed in the Chemical Division for 10 years when compared to the other employees
in the plant.  The SMR for workers not employed in the Chemical Division was less than
expected for prostate cancer (.58).

An update of this study was conducted to include the death experience of employees through
1997 (Alexander, 2001a).  The cohort consisted of 3992 workers. The eligibility requirement
was increased to 1 year of employment at the Cottage Grove plant, and the exposure categories
were changed to be more specific. Workers were placed into 3 exposure groups based on job
history information:  definite PFOA exposure (n = 492, jobs where cell generation, drying,
shipping and packaging of PFOA occurred throughout the history of the plant); probable PFOA
exposure (n = 1685, other chemical division jobs where exposure to PFOA was possible  but with
lower or transient exposures); and not exposed to fluorochemicals (n = 1815, primarily non-
chemical division jobs).

In this new cohort, 607 deaths were identified: 46 of these deaths were in the PFOA exposure
group, 267 in the probable exposure group, and 294 in the non-exposed group. When all
employees were compared to the state  mortality rates, SMRs were less than 1 or only slightly
higher for all of the causes of death analyzed. None of the SMRs were statistically significant at
p = .05. The highest SMR reported was for bladder cancer (SMR =1.31, 95% CI = 0.42 - 3.05).
Five deaths were observed (3.83 expected).

A few SMRs were elevated for employees in the definite PFOA exposure group:  2 deaths from
cancer of the large intestine (SMR =1.67, 95% CI = 0.02 - 6.02), 1 from pancreatic cancer
(SMR = 1.34, 95% CI = 0.03 - 7.42), and 1 from prostate cancer (SMR = 1.30, 95% CI = 0.03 -
7.20). In addition, employees in the definite PFOA exposure group were 2.5 times more likely
to experience cerebrovascular disease mortality (5 deaths observed, 1.94 expected; 95%  CI =
0.84-6.03).
                                          14

-------
SAR RPVJPW Draft; Tin Not Cite nr Dimfp
In the probable exposure group, 3 SMRs were elevated:  cancer of the testis and other male
genital organs (SMR = 2.75, 95% CI = 0.07 - 15.3); pancreatic cancer (SMR = 1.24, 95% CI =
0.45 - 2.70); and malignant melanoma of the skin (SMR= 1.42,  95% CI = 0.17 - 5.11). Only 1,
6, and 2 cases were observed, respectively. The SMR for prostate cancer in this group was 0.86
(95%CI = 0.28-2.02)(n = 5).

There were no notable excesses in SMRs in the non-exposed group, except for cancer of the
bladder and other urinary organs.  Four cases were observed and only 1.89 were expected (95%
CI = 0.58-5.40).

It is difficult to interpret the results of the prostate cancer deaths  between the first study and the
update because the exposure categories were modified in the update. Only 1 death was reported
in the definite exposure group and 5 were observed in the probable exposure group. All of these
deaths would have been placed in the chemical plant employees  exposure group in the first
study. The number of years that these employees worked at the plant and/or were exposed to
PFOA was  not reported. This is important because even 1 prostate cancer death in the definite
PFOA exposure group resulted in an elevated SMR for the group. Therefore, if any of the
employees' exposures were misclassified, the results of the analysis could be altered
significantly. This issue has become more apparent, given the results of a biomonitoring study
that took place at the Cottage Grove plant in 2000 in which PFOA concentrations were not
correlated with years worked in the Chemical Division but instead were associated with the
specific area of the plant where APFO was produced (Olsen, et al., 2003f).

The excess mortality in cerebrovascular disease noted in employees in the definite exposure
group was further analyzed based on number of years of employment at the plant.  Three of the 5
deaths occurred in workers who were employed in jobs with definite PFOA exposure for more
than 5 years but less than 10 years (SMR = 15.03/95% CI = 3.02 - 43.91).  The other 2 occurred
in employees with less than 1 year of definite exposure.  The SMR was 6.9 (95% CI = 1.39 -
20.24) for employees with greater than 5 years of definite PFOA exposure.  In order to confirm
that the results regarding cerebrovascular disease were not an artifact of death certificate coding,
regional mortality rates were used for the reference population. The results did not change.
When these deaths were further analyzed by cumulative exposure (time-weighted according to
exposure category),  workers with 27 years of exposure in probable PFOA exposed jobs or those
with 9 years of definite PFOA exposure were 3.3 times more likely to die of cerebrovascular
disease than the general population. A dose-response relationship was not observed with years
of exposure.

It is difficult to compare the results of the first and second mortality studies at the Cottage Grove
plant since the exposure categories were modified. Although the potential for exposure
misclassification was certainly more likely in the first study, it may still have  occurred in the
update as well.  It is difficult to judge the reliability of the exposure categories that were defined
without measured exposures. Although serum PFOA measurements were considered in the
exposure matrix developed for the update, they were not directly used. In the second study, the
chemical plant employees were sub-divided into PFOA-exposed groups, and the film plant
employees essentially remained in the "non-exposed" group. This was an effort to more
accurately classify exposures; however, these new categories do not take into account duration
of exposure or length of employment.  Another limitation to this study is that  17 death
certificates  were not located for deceased employees and therefore were not included in the
study.  The inclusion or exclusion of these deaths could change the analyses for the causes of
death that had a small number of cases. Follow-up of worker mortality at Cottage Grove (and

                                           15

-------
SAR RPVJPW Draft; Dn Not Cite* nr Oiinfp
Decatur) needs to continue. Although there were more than 200 additional deaths included in
this analysis, it is a small number and the cohort is still relatively young. Given the results of
studies on fluorochemicals in both animals and humans, further analysis is warranted.

Limited data are available on mortality and cancer incidence at Dupont's Washington Works
Plant in Parkersburg, WV.  These studies were periodically undertaken as part of a medical
surveillance program at the plant. The most recent report is summarized here. Cancer incidence
for active employees was reported for the 1959-2001 time period and mortality data were
reported for active and retired employees for 1957 through 2000 (Dupont, 2003). No other data,
such as employee exposure information, lifestyle factors, employee demographics, or other
chemicals used at the plant, are available in this report.

In the cancer incidence report, cancer cases were identified through a combination of company
health and life insurance claims and company cancer and mortality registries. Standardized
incidence ratios (SIRs) were only calculated for those cancers for which 5 or more cases were
observed, which included 14 types of cancer.  Two of those cancer types were elevated and
statistically significant (p = 0.05): bladder [SIR = 1.9; 95% CI (1.15-3.07)]  and kidney and
urinary organs [SIR = 2.3 (95% CI = 1.36-3.65)]. All of the reported cases were male. Some
other cancer types with elevated SIRs but which were not statistically significant at p = 0.05
included myeloid leukemia (2.02), cancer of the larynx (1.77), multiple myeloma and
immunoproliferative (1.72), malignant melanoma of skin (1.3), testicular cancer (1.46), and
brain cancer (1.2).

In the mortality report, when all causes of death were reported, SMRs, adjusted for age and
gender, were statistically significant (p = 0.05) for rheumatic heart disease (SMR = 3.55; 95%
CI, 1.14-8.30) and atherosclerosis and aneurysm (SMR = 1.98; 95% CI, 1.17-3.14).

Two separate analyses of leukemia incidence were conducted prior to the above studies at
Dupont's Washington Works plant (Walrath and Burke, 1989; Karns and Fayerweather, 1991).
The initial study reported a statistically significant (p = 0.10) elevated odds ratio (OR) of 2.1  for
leukemia incidence for male employees working at the plant from  1956-1989.  Eight cases (all
male) of different types of leukemia were identified.  The OR remained  elevated when the
workers were divided into wage and salaried employees (2.2 and 2.0, respectively). In the
follow-up case-control study, four controls were selected from the plant for each case, matched
on gender, age and payroll  status. Matched odds ratios were significantly elevated (p = 0.10) for
employees who had previously worked as custodians and engineers, 8.0 (90% CI, 1.1-60.0) and
7.9 (90% CI, 1.0 - 76.0), respectively and remained elevated (although not statistically
significant) for these same job categories within the plant (OR= 4.0 and 5.1, respectively).
Matched OR were also reported based on the area of the plant where the cases worked; however,
no statistically significant (p = 0.10) OR were reported.

The mortality data reported above do not show any statistically significant (p = 0.10) elevations
in leukemia deaths (all of the cases in the case-control study were  dead at the time of the
mortality report), possibly because the number of cases was very small and divided among
different types of leukemias.  The Washington Works data provide some insight as to where
more medical surveillance  should be concentrated at this plant but provide little information
about the relationship of PFOA to mortality or cancer incidence since no exposure information,
use of other chemicals, or lifestyle information was collected on these employees.
                                           16

-------
SAR RPVJAW Draft! Dn Nnt fite nr Onnte
3.1.2 Hormone Study in Male Workers

Endocrine effects have been associated with PFOA exposure in animals; therefore, medical
surveillance data, including hormone testing, from male employees only of the Cottage Grove,
Minnesota plant were analyzed (Olsen et al.,  1998a). PFOA serum levels were obtained for
volunteer workers in 1993 (n= 111) and 1995 (n = 80). Sixty-eight employees were common to
both sampling periods. In 1993, the range of PFOA was 0-80 ppm and 0-115 ppm in 1995 using
thermospray mass spectrophotometry assay.  Eleven hormones were assayed from the serum
samples.  They were: cortisol, dehydroepiandrosterone sulfate (DHEAS), estradiol, follicle
stimulating hormone (FSH), 17 gamma-hydroxy progesterone (17-HP), free testosterone, total
testosterone, luteinizing hormone (LH), prolactin, thyroid-stimulating hormone (TSH) and sex
hormone-binding globulin (SHBG). Employees were placed into 4 exposure categories based on
their serum PFOA levels: 0-1 ppm, 1- < 10 ppm, 10- < 30 ppm, and >30 ppm. Statistical
methods used to compare PFOA levels and hormone values included: multivariable regression
analysis, ANOVA, and Pearson correlation coefficients.

PFOA was not highly correlated with any of the hormones or with the following covariates: age,
alcohol consumption, body  mass index (BMI), or cigarettes. Most of the employees had PFOA
serum levels less than 10 ppm. In 1993, only 12 employees had serum levels > 10 ppm, and 15
in 1995. However, these levels ranged from  approximately 10 ppm to over 114 ppm. There
were only 4 employees in the >30 ppm PFOA group in 1993 and only 5 in 1995. Therefore, it is
likely that there was not enough power to detect differences in either of the highest categories.
The mean age of the employees in the highest exposure category was the lowest in both 1993
and 1995 (33.3 years and 38.2 years, respectively).  Although not significantly different from the
other categories, BMI was slightly higher in the highest PFOA category.

When the mean values of the various hormones were compared by exposure categories, there
was a statistically significant (p = .01) elevation in prolactin in 1993 only for the 10 workers
whose serum levels were between 10 and 30  ppm compared to the lower 2 exposure categories.
In addition, TSH was significantly (p = .002) elevated in the same exposure category for 1995
only (mean blood serum level was 2.9 ppm).  However, mean TSH levels for the other exposure
categories, including the ^ 30 ppm category,  were all the same (1.7 ppm). In 1993, TSH was
elevated only in this same exposure category, as well; however did not reach statistical
significance (p = .09).

Estradiol levels in the >30 ppm group in both years were 10% higher than the other PFOA
groups; however, the difference was not statistically significant (p <0.05). These results were
confounded by estradiol being highly correlated with BMI (r = .41, p < .001 in 1993, and r = .30,
p < .01 in  1995).  In 1995, all 5 employees with PFOA levels > 30 ppm had BMIs > 28, although
this effect was not observed in 1993. The authors postulate that the study may not have been
sensitive enough to detect an association between PFOA and estradiol because measured serum
PFOA levels were likely below the observable effect levels suggested in animal studies (55 ppm
PFOA in the CD rat). Only 3  employees in this study had PFOA serum levels this high. They
also suggest that the higher  estradiol levels in the highest exposure category could suggest a
threshold relationship between PFOA and estradiol.

The authors did not report a negative association between PFOA serum levels and testosterone.
There were no  statistically significant trends  (p < 0.05) noted for PFOA and either bound or free
testosterone. However, 17-HP, a precursor of testosterone, was highest in the >30 ppm PFOA
group in both 1993 and 1995.  In 1995, PFOA was significantly associated with 17-HP in

                                          17

-------
SAR Rpyjpw Draft; lln Not fite nr Onnfp
regression models adjusted for possible confounders. However, the authors state that this
association was based on the results of one employee (data were not provided in the report).
Free testosterone was highly correlated with age in both 1993 and 1995 (r = -.48, p < .001; r = -
.40, p < .001, respectively).

There are several design issues that should be noted when evaluating the results of this study.
First, although there were 2  study years (1993 and 1995), the populations were not independent.
Sixty-eight employees participated in both years. Second, there were 31 fewer employees who
participated in the study in 1995, thus reducing the power of the study.  There were also very few
employees in either year with serum PFOA levels greater than 10 ppm. Third, the cross-
sectional design of the study does not allow for analysis of temporality of an association. Since
the half-life of PFOA is at least 1 year, the authors suggest that it is possible that there may  be
some biological accommodation to the effects of PFOA. Fourth, only one sample  was taken for
each hormone for each of the study years.  In order to get more accurate measurements for some
of the hormones, pooled blood taken in a short time period should have been used  for each
participant. Fifth, some of the associations that were measured in this study were done based on
the results of an earlier paper that linked PFOA with increased estradiol and decreased
testosterone levels. However, total serum organic fluorine was measured in that study instead of
PFOA, making it difficult to compare the results. Finally, there may have been some
measurement error of some of the confounding variables.

3.1.3 Occupational Study on Episodes of Care

In order to gain additional insight into the effects of fluorochemical exposure on workers' health,
an "episode of care" analysis was undertaken at the Decatur plant to screen for morbidity
outcomes that may be associated with long-term, high exposure to fluorochemicals (Olsen et al.,
2001g). An "episode of care" is a series of health care services provided from the  start of a
particular disease or condition until solution or resolution of that problem. Episodes of care
were identified in employees' health claims records using Clinical Care Groups (CCG) software.
All inpatient and outpatient visits to health care providers, procedures, ancillary services and
prescription drugs used in the diagnosis, treatment, and management of over 400 diseases or
conditions were tracked.

Episodes of care were analyzed for 652 chemical employees and 659 film plant employees who
worked at the Decatur plant for at least 1 year between January 1,1993 and December 31,1998.
Based on work history records,  employees were  placed into different comparison groups: Group
A consisted of all film and chemical plant workers; Group B had employees who only worked in
either the film or chemical plant; Group C consisted of employees who worked in jobs with high
POSF exposures; and Group D had employees who worked in high exposures in the chemical
plant for 10 years or more prior to the onset of the study. Film plant employees were considered
to have little or no fluorochemical exposure, while chemical plant employees were assumed to
have the highest exposures.

Ratios of observed to expected episodes of care were calculated for each plant. Expected
numbers were based on 3M's employee population experience using indirect standardization
techniques. A ratio of the chemical plant's observed to expected experience divided by the film
plant's observed to expected experience was calculated to provide a relative risk ratio for each
episode of care (RREpC). For each RREpC,  95% confidence intervals were calculated.
Episodes of care that were of greatest interest were those which had been reported in animal or
epidemiologic literature on PFOS and PFOA: liver and bladder cancer, endocrine disorders

                                          18

-------
SAB RPVIPW Draff; Dn Not rite nr Qnnfp
involving the thyroid gland and lipid metabolism, disorders of the liver and biliary tract, and
reproductive disorders.

The only increased risk of episodes for these conditions of a priori interest were for neoplasms of
the male reproductive system and for the overall category of cancers and benign growths (which
included cancer of the male reproductive system). There was an increased risk of episodes for
the overall cancer category for all 4 comparison groups.  The risk ratio was greatest in the group
of employees with the highest and longest exposures to fluorochemicals (RREpC = 1.6, 95% CI
= 1.2- 2.1). Increased risk of episodes in long-time, high-exposure employees also was reported
for male reproductive cancers (RREpC = 9.7, 95% CI = 1.1 - 458).  It should be noted that the
confidence interval is very wide for male reproductive cancers and the sub-category of prostate
cancer.  Five episodes of care were observed for reproductive cancers in chemical plant
employees (1.8 expected), of which 4 were prostate cancers (RREpC = 8.2, 95% CI = 0.8 - 399).
One episode of prostate cancer was observed in film plant employees (3.4 expected).  This
finding should be noted because an excess in prostate cancer mortality was observed in the
Cottage Grove plant mortality study when there were only 2 exposure categories (chemical
division employees and non-chemical  division employees). The update of the study sub-divided
the chemical plant employees and did not corroborate this finding when exposures were divided
into definitely exposed and probably exposed employees.

There was an increased risk of episodes for neoplasms of the gastrointestinal tract in the high
exposure group (RREpC = 1.8,  95% CI = 1.2- 3.0) and the long-term employment, high
exposure group (RREpC = 2.9,  95% CI = 1.7 - 5.2). Most of the episodes were attributable to
benign colonic polyps.  Similar numbers of episodes were reported in film and chemical plant
employees.

In the entire cohort, only 1 episode of care was  reported for liver cancer (0.6 expected) and 1 for
bladder cancer (1.5 expected). Both occurred in film plant employees. Only 2 cases of cirrhosis
of the liver were observed (0.9 expected), both  in the chemical plant.  There was a greater risk of
lower urinary tract infections in chemical plant employees, but they were mostly due to recurring
episodes of care by the same  employees. It is difficult to draw any conclusions about these
observations, given the small number of episodes reported.

Chemical plant employees in the high exposure, long-term employment group were 2 V* times
more likely to seek care for disorders of the biliary tract than their counterparts in the film plant
(RREpC = 2.6, 95% CI = 1.2 - 5.5).  Eighteen episodes of care were observed in chemical plant
employees and 14 in film plant workers. The sub-categories that influenced this observation
were episodes of cholelithiasis with acute cholecystitis and cholelithiasis with chronic or
unspecified cholecystitis.  Most of the observed cases occurred in chemical plant employees.

Risk ratios of episodes of care for endocrine disorders, which included sub-categories of thyroid
disease, diabetes, hyperlipidemia, and other endocrine or nutritional disorders, were not elevated
in the comparison groups. Conditions which were not identified a priori but which excluded the
null hypothesis in the 95% confidence interval for the high exposure, long-term employment
group included:  disorders of the pancreas, cystitis, and lower urinary tract infections.

The results of this study should only be used for hypothesis generation.  Although the episode of
care design allowed for a direct comparison of workers with similar demographics but different
exposures, there are many limitations to this design. The limitations include:  1) episodes of care
are reported, not disease incidence, 2) the data are difficult to interpret because a large RREpC

                                           19

-------
SAB RPVJPW Draft; T>n Nnt fite nr
may not necessarily indicate high risk of incidence of disease, 3) many of the risk ratios for
episodes of care had very wide confidence intervals, thereby providing unstable results, 4) the
analysis was limited to 6 years, 5) the utilization of health care services may reflect local
medical practice patterns, 6) individuals may be counted more than once in the database because
they can be categorized under larger or smaller disease classifications, 7) episodes of care may
include the same individual several times, 8) not all employees were included in the database,
such as those on long-term disability, 9) the analysis may be limited by the software used, which
may misclassify episodes of care, 10) the software may assign 2 different diagnoses to the same
episode, and 11) certain services, such as lab procedures may not have been reported in the
database.

3.1.4 3M Medical Surveillance  Studies

3.1.4.1 Antwerp and Decatur Plants-Cross-Sectional and Longitudinal Studies

A cross-sectional analysis of the  data from the 2000 medical surveillance program at the Decatur
and Antwerp plants was undertaken to determine if there were any associations between PFOA
and hematology, clinical chemistries, and hormonal parameters of volunteer employees (Olsen et
al., 2001e). The data were analyzed for all employees from both plant locations. Mean PFOA
serum levels were 1.03 ppm for all male employees at the Antwerp plant and 1.90 ppm for all
male employees at the Decatur plant. Male production employees at the Decatur plant had
significantly higher (p < .05) mean serum levels (2.34 ppm) than those at the Antwerp plant
(1.28 ppm). Non-production employees at both plants had mean levels below 1  ppm.  PFOA
serum levels were higher than the PFOS serum values at both plants, especially the Decatur plant
where serum levels are higher overall. In addition, values for total organic fluorine were even
higher than the PFOA levels.

Multivariable regression analyses were conducted to adjust for possible confounders that may
affect the results of the clinical chemistry tests. The following variables were included:
production job (yes or no), plant, age, BMI, cigarettes/day, drinks/day and years worked at the
plant. A positive significant association was reported between PFOA and cholesterol (p = .05)
and PFOA and triglycerides (p = .002). Age was also significant in both analyses.  Alcohol
consumed per day was significant in the cholesterol model, while BMI and cigarettes smoked
per day was significant for triglycerides. When both PFOA and PFOS were included in the
analyses, neither reached statistical significance in the cholesterol model, while PFOA remained
significant (p = .02) in the triglycerides model.  High-density lipoprotein (HDL) was negatively
associated with PFOA (p = .04) and remained significant (p = .04) when both PFOA and PFOS
were included in the model. A positive association (p = .01) between T3 and PFOA was also
observed and remained statistically significant (p = .05) when PFOS was included in the model.
BMI, cigarettes/day, alcohol/day were also significant in the model.  None of the other clinical
chemistry, thyroid or hematology measures were significantly associated with PFOA in the
regression model.

A longitudinal analysis of the above data and previous medical surveillance results was
performed to determine whether  occupational exposure to fluorochemicals over time is related to
changes in clinical chemistry and lipid results in employees of the Antwerp and Decatur
facilities (Olsen et al., 2001f). The clinical chemistries included: cholesterol, HDL,
triglycerides, alkaline phosphatase, gamma glutamyl transferase (GGT), aspartate
aminotransferase (AST), alanine aminotransferase (ALT), total and direct bilirubin. Medical
surveillance data from 1995,1997, and 2000 were analyzed using multivariable regression

                                           20

-------
SAB RPVJPW Draft; F>n Nnt fitf nt- Oiinfp
analysis.  The plants were analyzed using 3 subcohorts that included those who participated in 2
or more medical exams between 1995 and 2000. A total of 175 male employees voluntarily
participated in the 2000 surveillance and at least one other. Only 41 employees were
participants in all 3 surveillance periods.

When mean serum PFOA levels were compared by surveillance year, PFOA levels in the
employees participating in medical surveillance at the Antwerp plant increased between 1994/95
and 1997 and then decreased slightly between 1997 and 2000. At the Decatur plant, PFOA
serum levels decreased between 1994/95 and 1997 and then increased between 1997 and 2000.
When analyzed using mixed model multivariable regression and combining Antwerp and
Decatur employees, there was a statistically significant positive association between PFOA and
serum cholesterol (p = .0008) and triglycerides (p = .0002) over time. When analyzed by plant
and also by subcohort, these associations were limited to the Antwerp employees (p = .005) and,
in particular, the 21 Antwerp employees who participated in all 3 surveillance years (p = .001).
However, the association between PFOA and triglycerides was also statistically significant
(p = .02)  for the subgroup in which employees participated in biomonitoring in 1994/95 and
2000. There was not a significant association between PFOA and triglycerides among Decatur
workers.  There were no significant associations between PFOA and changes over time in HDL,
alkaline phosphatase, GGT, AST, ALT, total bilirubin, and direct bilirubin.

There are several limitations to the 2000 cross-sectional and longitudinal studies including:
1) serum PFOA levels were significantly higher at the Decatur plant than at the Antwerp plant,
2) all participants were volunteers, 3) there were several consistent differences in clinical
chemistry profiles and demographics between employees of the Decatur and Antwerp  plants
(Antwerp employees as compared to Decatur employees had lower PFOA serum levels, were
younger,  had lower BMIs, worked fewer years, had higher alcohol consumption, higher mean
HDL and bilirubin values, lower mean triglyceride, alkaline phosphatase, GGT, AST,  and ALT
values, and mean thyroid hormone values tended to be higher), 4) PFOS and other perfluorinated
chemicals are also present in these plants, 5) in the cross-sectional study, plant populations
cannot be compared because they were placed into quartiles based on PFOS serum distributions
only which were different for each subgroup and not applicable to PFOA, 6) only one
measurement at a certain point in time was collected for each clinical chemistry test, and 7)
PFOA serum levels overall have been increasing over time in these employees. In addition, in
the longitudinal study only a small number of employees participated in all 3 sampling periods
(24%), different labs and analytical techniques for PFOA were used each year, and female
employees could not be analyzed because of the small number of participants.

3.1.4.2 Cottage Grove Plant-Cross-sectional Studies of Clinical Chemistries and CCK

A voluntary medical surveillance program was offered to employees of the Cottage Grove,
Minnesota plant in  1993,1995, and 1997 (n = 111, 80 and 74 employees, respectively) (Olsen et
al., 1998b, 2000). The clinical chemistry parameters (cholesterol, hepatic enzymes, and
lipoprotein levels) used in the longitudinal and cross-sectional studies of the Antwerp and
Decatur plants were also  used in this study. In addition, in 1997 only, cholecystokinin-33 (CCK)
was also measured at the Cottage Grove plant. CCK levels were observed because certain
research has  suggested that pancreas acinar cell adenomas seen in rats exposed to PFOA may be
the result of increased CCK levels (Obourn et al., 1997).

Only male employees involved in PFOA production were included in the study.  Sixty-eight
employees were common to the 1993 and 1995 sampling periods, 21 were common between

                                         21

-------
SAR RPVJPW Draff; Dn Nnf Cite' nr
1995 and 1997, and 17 participated in all three surveillance years.  Cottage Grove has the highest
serum PFOA levels of the 3 plants studied.

Employees' serum PFOA levels were stratified into 3 categories (<1,1- <10, and ^ 10 ppm),
chosen to provide a greater number of employees in the 2 10 ppm category. As employees'
mean serum PFOA levels increased, no statistically significant (p < 0.05) abnormal liver
function tests, hypolipidemia, or cholestasis were observed in any of the sampling years.
Multivariable regression analyses controlling for potential confounders (age, alcohol
consumption, BMI, and cigarettes smoked) yielded similar results. The authors also reported
that renal function, blood glucose,  and hematology measures were not associated with serum
PFOA levels; however, these data were not provided in the paper.

The mean CCK value reported for the 1997 sample was 28.5 pg/ml (range 8.8 - 86.7 pg/ml).
The means in the 2 serum categories < 10 ppm were at least 50% higher than in the ^ 10 ppm
category. A statistically significant (p = .03) negative association between mean CCK levels and
the 3 PFOA serum categories was observed.  A scatter plot of the natural log of CCK and PFOA
shows that all but 2 CCK values are within the assay's reference range of 0 - 80 pg/ml.  Both of
these employees (CCK values of 80.5 and 86.7 pg/ml) had serum PFOA levels less than 10 ppm
(0.6 and 5.6 ppm, respectively).  A multiple regression model of the natural log of CCK and
serum PFOA levels continued to display a negative association after adjusting for potential
confounders.

The cross-sectional design is a limitation of this study.  Only 17 subjects were common to all 3
sampling years.  In addition, the medical surveillance program is a voluntary one. The
participation rate of eligible production employees decreased from approximately 70% in 1993
to 50% in 1997.  Also, the laboratory reference range changed substantially for ALT in 1997.
Finally, different analytical methods were used to measure serum PFOA.  Serum PFOA was
determined by electrospray high-performance liquid chromatography/mass spectrometry in
1997, but by thermospray in 1993 and 1995.

3M used data collected in their 2000 medical surveillance program to determine whether serum
PFOA levels greater than 5 ppm were associated with changes in hepatic, lipid and thyroid
function in workers (Olsen, et al., 2003f). Clinical chemistries, including thyroid function tests,
were performed for 131 male and 17 female workers at the plant. Serum samples were extracted
and quantitatively analyzed for PFOA using HPLC/EMSS. All of the samples were above the
lower limit of quantitation (LLOQ) (see Table 24 in Section 4.1.1 for biomonitoring data).

Fifteen percent (n = 20) of male employees had serum concentrations that exceeded 5 ppm, and
none of the female employees were above 5 ppm. Number of years worked in the Chemical
Division of the plant were not correlated with PFOA serum measurements, but were correlated
with specific production areas. When the male employees were separated into 3 groups based on
serum PFOA levels (< 1 ppm, 1-4.9 ppm, and >5 ppm), there were no statistically significant (p
< .05) differences in mean lipid  and hepatic test result:? or in thyroid hormone levels, between the
3 groups both before and after adjusting for potential confounders (eg. BMI, smoking status, and
alcohol consumption). The elimination of employees receiving cholesterol-reducing drugs (n =
9) from the analysis did not alter these findings. Similar to findings at the Decatur and Antwerp
plants, triglyceride levels were higher in employees with the highest PFOA serum
concentrations, although not statistically significant (p  < .05). In simple linear regression
analyses, a weak negative association between T4 and serum PFOA concentrations was reported
(p = .07). However, all of the serum PFOA concentrations were within the T4 reference range,

                                          22

-------
SAR RPVJPW Draft; Tin Nnt Cite nr
the statistical association explained minimal variation in the model (r2 = .03), there was no
increase in serum TSH or T3 levels, and no negative association between free T4 levels and
PFOA. No statistically significant (p < 0.05) associations between PFOA and clinical
chemistries or thyroid test results for the small group of female employees was reported.

An earlier medical surveillance study on workers who were employed in the 1980's was
conducted at the Cottage Grove plant; however, total serum fluorine was measured instead of
PFOA (Gilliland and Mandel, 1996). Based on animal studies that reported that animals exposed
to PFOA develop hepatomegaly and alterations in lipid metabolism, a cross-sectional,
occupational study was performed to determine if similar effects are present in workers exposed
to PFOA.  In a PFOA production facility, 115 workers were studied to determine whether serum
PFOA affected their cholesterol, lipoproteins, and hepatic enzymes.  Forty-eight workers who
were exposed to PFOA from 1985-1989 were included in the study (96% participation rate).
Sixty-five employees who either volunteered or were asked to participate, were included in the
unexposed group.  These employees were assumed to have little or no PFOA exposure based on
their job description. However, when serum levels were analyzed, it was noted that this group of
workers had PFOA levels much greater than the general population. Therefore, instead of job
categories, total  serum fluorine was used to classify workers into exposure groups.

Total serum fluorine was used as a surrogate measure for PFOA.  Serum PFOA was not
measured, due to the cost of analyzing the samples.  Blood samples were analyzed for total
serum fluorine, serum glutamyl oxaloacetic transaminase (SCOT or AST), serum glutamyl
pyruvic transaminase (SGPT or ALT), GGT, cholesterol, low-density lipoproteins (LDL), and
HDL. All of the participants were placed into five categories of total serum fluorine levels: <1
ppm, 1-3 ppm, >3 -10 ppm, >10 -15 ppm, and > 15 ppm.  The range of the serum fluorine
values was 0 to 26 ppm (mean 3.3 ppm).  Approximately half of the workers fell into the > 1 - 3
ppm category, while 23 had serum levels < 1 ppm and 11 had levels > 10 ppm.

There were no significant differences between exposure categories when analyzed using
univariate analyses for cholesterol, LDL, and HDL. In the multivariate analysis, there was not a
significant association between total serum fluorine and cholesterol or LDL after adjusting for
alcohol consumption, age, BMI, and cigarette smoking.  There were no statistically significant
differences among the exposure categories of total serum fluorine for AST, ALT and GGT.
However, increases in AST and ALT occurred with increasing total serum fluorine levels in
obese workers (BMI = 35 kg/m2).  This result was not observed when PFOA was measured
directly in serum of workers in 1993,1995, or 1997 surveillance data of employees of the
Cottage Grove plant (Olsen et al., 2000).

Since PFOA was not measured directly and there is no exposure information provided on the
employees (e.g. length of employment/exposure), the results of the study provide limited
information.  The authors state that no adverse clinical outcomes related to PFOA exposure have
been observed in these employees; however, it is not clear that there has been follow-up of
former employees.  In addition, the range of results reported for the liver enzymes were fairly
wide for many of the exposure categories, indicating variability in the results.  Given that only
one sample was taken from each employee, this is not surprising.  It would be much more
helpful to have several samples taken over time to ensure their reliability.  It also would have
been interesting to compare the results of the workers who were known to be exposed to PFOA
to those who were originally thought not to be exposed to see if there were any differences
among the employees in these groups.  There were more of the "unexposed" employees (n = 65)
participating in the study than those who worked in PFOA production (n = 48).

                                         23

-------
SAB RPVJPW Draft; f>n Nnt fite nr Quote
3.2 Metabolism and Pharmacokinetics

3.2.1 Metabolism and Pharmacokinetic Studies in Humans

Little is known about the metabolism and pharmacokinetics of PFOA in humans.  One report
notes the presence of PFOA in the cord blood of some pregnant workers suggesting that PFOA
can cross the placenta (U.S. EPA, 2003). In addition, there are limited data on the half-life of
PFOA in humans. With the exception of a 1980 study in which total organic fluorine in blood
serum was measured in one worker, no other data were available until June 2000 (Ubel et al.,
1980).   A half-life study on 27 retirees from the Decatur and Cottage Grove 3M plants was
undertaken, in which serum samples were drawn every 6 months over a 5-year period. Two
interim reports describing the results thus far have been submitted (Burns et al., 2000, 2002).
The first interim report suggested a median serum half-life of PFOA of 344 days, with a range of
109 to 1308 days. The two highest half-life calculations were for the 2 female retirees who
participated in this study (654 and 1308 days).

There were several limitations to this first analysis including: 1) the limited data available and
the range of serum PFOA levels measured;  2) serum was analyzed after each collection period
with only one measurement per time period on different days using slightly different analytical
techniques; and  3) the reference material purity was not determined until after the first 3 samples
had been analyzed.

An effort was made to minimize experimental error, including systematic and random error in
the analytical method.  Serum samples were collected from 9 of the original 27 subjects over 4
time periods spanning 180 days, measured in triplicate with all time points from each subject
analyzed in the same analytical run. This would allow for statistical evaluation of the precision
of the measurement and assure that all systematic error inherent in the assay equally affected
each sample used for half-life determination. Single serum measurements were made on
samples of the remaining 18 retirees, but were not included by the investigators in the analysis
because triplicate analyses of all time points were not conducted.

Of the 9 retirees included in this analysis, there were 7 males and 2 females, all from the Decatur
plant. The average age of the retirees was 61 years, the mean number of years worked at
Decatur was 27.7 years, and the average number of months retired from the plant at study
initiation was  18.9. The average BMI of this group was 27.9 (range 22.5-33, SD = 3.6). The
mean PFOA value at study initiation was 0.72 ppm (range 0.06 - 1.84 ppm, SD = 0.64).

The mean serum half-life for PFOA was 4.37 years (range 1.50 - 13.49 years, SD = 3.53).  Only
1 employee had a half-life value that exceeded 4.3 years. The 2 females had values of 3.1 and
3.9 years. Age,  BMI, number of years worked or years since retirement were not significant
predictors of serum half-lives in multivariable regression analyses.

This analysis has attempted to reduce experimental error in the determination of a half-life for
PFOA. However, two issues should be noted.  First, the effect of continual non-occupational,
low-level exposure on the half-life is unknown. Second, systematic error of the analytical
method could be as high as +/- 20% and still satisfy the data quality criteria.

3.2.2 Metabolism and Pharmacokinetic Studies  in Non-Human Primates

Noker and Gorman (2003) administered a single intravenous (i.v.) dose of 10 mg/kg potassium

                                          24

-------
SAB RPVJPW Draff; Tin Nnt fite nr Onntp
PFOA to 3 male and 3 female cynomolgus monkeys.  The animals were approximately 3-4 years
of age at the start of the study.  Each monkey was examined shortly after dosing for clinical
signs of toxicity; all animals were observed twice daily for signs of mortality/moribundity.
Additional clinical observations were performed on the days blood was collected. Each animal
was weighed on days 1,4, 7,14, 21 and 28.  Urine and feces were collected on days 0,1, 2, 7,
14, 21, and 28. Blood was collected at 0, 0.5, 2, 4, 8 and 24 hours and on days 2,4, 7,11,14, 21,
28, 57, 79, 87, and 123. Serum and urine samples were analyzed by HPLC/MS/MS.

No deaths that could be attributed to administration of the test article occurred during the study.
One male was euthanized on day 79 because he had developed repeated episodes of self-
mutilation. These episodes do not appear to be related to PFOA administration. No adverse
clinical signs were noted in any of the other monkeys during the course of the study.  Body
weights of the treated animals did not change between days 1 and 28.

At 0.5 hours after dosing, serum concentration in males and females were similar, and ranged
from 91,130 to 96,660 ng/ml (ppb) in males and from 88,940 to 96,400 ng/ml in females. Serum
concentrations declined slowly, and the levels decreased faster in two of the male monkeys than
in the third male monkey and all three female monkeys.  By day 123, PFOA concentrations were
at or slightly above the limit of quantitation (20 ng/ml) in the two surviving males and between
885 and 4701 ng/ml in the three females.

The pharmacokinetic parameters that were calculated from the serum concentrations of PFOA
are presented in Table 1.  The estimated values for half-life and total body clearance indicated
that two of the three male monkeys eliminated PFOA at a faster rate than did the female
monkeys. The volume of distribution of PFOA at steady  state was similar for both sexes.  Male
#2052 appeared to behave more like the females in his pharmacokinetic parameters than either
of the other two males in the study. One explanation is the stress of the experiment caused this
monkey to release high levels of cortisol that could have affected carbohydrate, protein, and/or
lipid metabolism, caused a shift in electrolyte and water balance, or increased plasma proteins.
This animal was also observed to have periods of severe  self-mutilation and euthanized for this
reason on day 79.
                                          25

-------
SAR RPVIPW Uraff; Dn Not fife nr Oiintp
                                       Table 1

       Pharmacokinetic Parameters Calculated from Serum Concentrations of PFOA in
                                 Cynomolgus Monkeys
Parameter
Cmax a
(fig/mL)
ll/2 (day)
AUC0_last c
(|ig»day/mL)
AUCo.infinity d
(jig»day/mL)
Cl (mL/day/kg)e
Vdss(mL/kg)f
Males
2052
101
35.3
1999
2501
4.0
192
2054
98.4
13.7
569
571
17.5
168
2211
91.6
13.6
633
634
15.8
184
Mean
97.0
20.9
1067
1235
12.4
181
S.D.
4.9
12.5
808
1097
7.7
12
Females
2058
105
26.8
3083
3224
3.1
133
2059
93.0
41.7
2314
2560
3.9
190
2061
103
29.3
1056
1094
9.1
270
Mean
100
32.6
2151
2293
5.4
198
S.D.
6.0
8.0
1023
1090
3.3
69
a Maximum concentration in serum
b Half-life of the terminal elimination phase
c Area under the serum concentration versus time curve calculated from 0 to the last time point
d Area under the serum concentration versus time curve calculated from 0 to infinity
e Total body clearance
f Volume of distribution at steady state

There were no obvious sex differences in the urinary excretion of PFOA. The rate of urinary
excretion by the monkeys was slow.  Less than 20% of the administered dose was excreted in the
urine of the male and female monkeys within the first 48 hours after dosing, and as much as 0.1 -
1% of the administered dose was excreted in the urine on day 28.

3.2.3  Metabolism and Pharmacokinetic Studies in Adult Rats

3.2.3.1 Absorption Studies

Studies in rats have shown that PFOA is absorbed following oral, inhalation, and dermal
exposures.
3.2.3.1.1 Oral Exposure
                                                             14,
Gibson and Johnson (1979) administered a single dose of 11.0 mg/kg  C-PFOA by gavage to
groups of 3 male 10-week old CD rats. Twenty-four hours after administration, at least 93% of
the total carbon-14 was absorbed; the elimination half-life of carbon-14 from the plasma was 4.8
days.
                                          26

-------
SAR RPVJPW Draft; Dn Not Pifp nr Oiinfp
Ophaug and Singer (1980) administered 2 ml of an aqueous solution of 2 mg PFOA to female
Holtzman rats. Ionic fluoride, nonionic fluorine, and total fluorine were measured.  Within 4.5
hr, 37% of the total fluorine in the administered dose was recovered in the urine.  The quantity of
nonionic fluorine recovered in the urine increased to 61%, 76% and 89% at 8, 24 and 96 hr,
respectively, after administration.  Within 4.5 hr, serum from treated rats had a nonionic fluorine
level of 13.6 ppm. The nonionic fluorine level in the serum decreased to 11.2 ppm at 8 hr, 0.35
ppm at 24 hr, and 0.08 ppm at 96 hr. Despite the large decrease in nonionic fluorine in the
serum, the ionic fluoride level was only 0.03 ppm and remained at that level throughout the
experiment.  Prior to administration of PFOA, the ionic and nonionic fluorine background levels
in serum were 0.032 and 0.07 ppm, respectively. The authors concluded that PFOA is rapidly
absorbed from the gastrointestinal tract and then rapidly cleared from the serum.

3.2.3.1.2 Inhalation Exposure

Hinderliter (2003) measured the serum concentrations of PFOA following single and repeated
inhalation exposures in Sprague-Dawley rats. For the single exposure study, male and female
rats (3/sex/group) were exposed to a single nose-only exposure of an aerosol of 0,1,10, or 25
mg/m3 PFOA. Preliminary range-finding studies demonstrated that aerosol sizes were 1.8 - 2.0
nmmass median aerodynamic diameter (MMAD) with geometric standard  deviations ranging
from 1.9 - 2.1 \im. Blood samples were collected pre-exposure, at 0.5,1, 3  and 6 hours during
exposure, and at 1, 3,  6,12,18 and 24 hours after exposure.  Plasma was analyzed by LC-MS.
PFOA plasma concentrations were proportional to the inhalation exposure concentrations. The
male Cmax values were approximately 2-3 times higher than the female Cmax. The female Cmax
occurred approximately one  hour after the exposure period, while the male  Cma3! occurred from
the end of the exposure period up to  six hours after exposure. In females, the elimination of
PFOA was rapid at all exposure levels, and by  12 hours after exposure the plasma levels had
dropped below the analytical limit of quantitation (0.1 ng/ml). In males, the plasma elimination
was much slower, and at 24 hours after exposure, the plasma concentrations were approximately
90% of the peak concentrations at all exposure levels.

In the repeated dose study, Hinderliter (2003) exposed male and female rats (5/sex/group) to the
same aerosol concentrations of PFOA for 6 hrs/day, 5 days/week for 3 weeks. Blood was
collected immediately before and after the daily exposure period three days per week. The
aerosol sizes were 1.3 -1.9 nm MMAD with geometric standard deviations of 1.5 - 2.1.  PFOA
plasma concentrations were proportional to the inhalation exposure concentrations,  and repeated
exposures produced little plasma carryover in females, but significant carryover in males. Male
rats reached steady  state plasma levels by three weeks with plasma concentrations of 8, 21, and
36 jig/ml for the 1,10 and 25 mg/m3 groups, respectively.  In females, the post-exposure plasma
levels were 1, 2, and 4 ng/ml for the 1,10, and 25 mg/m  groups, respectively. When measured
immediately before the daily exposure, plasma levels had returned to baseline in females.

3.2.3.1.3 Dermal Exposure

No specific dermal  absorption studies have been conducted in rats.  However, Kennedy (1985)
treated rats dermally with a total of 10 applications of APFO at doses of 0, 20, 200 or 2,000
mg/kg.  Treatment resulted in elevated blood organofluorine levels that increased in a dose-
related manner.
                                          27

-------
SAB RPVJPW Draft! Hn Nnt fite nr Oiinfp
3.2.3.2 Serum Pharmacokinetic Parameters in Adult Rats

Serum pharmacokinetic parameters of PFOA have been evaluated in adult Sprague-Dawley rats
following gavage administration, and in Wistar rats following i.v. administration.

3.2.3.2.1 Oral and Intravenous Exposure in Sprague-Dawley Rats

Kemper (2003) examined the plasma concentration profile of PFOA following gavage and
intravenous administration in sexually mature Sprague-Dawley rats. Male and female rats
(4/sex/group) were administered single doses of PFOA by gavage at dose rates of 0.1,1, 5, and
25 mg/kg PFOA, and intravenously at a dose rate of 1 mg/kg PFOA. After dosing, plasma was
collected for 22 days in males and 5 days in females.  Plasma concentration vs. time data were
then analyzed by non-compartmental pharmacokinetic methods (Tables 2 and 3). Comparison of
AUC for the oral and intravenous 1 mg/kg doses indicated that oral bioavailability of PFOA was
approximately 100%. Plasma elimination curves were linear with respect to time in male rats at
all dose levels, while elimination kinetics were biphasic in females at the 5 mg/kg and 25 mg/kg
dose levels.  In males, plasma elimination half-lives were independent of dose level and ranged
from approximately 138 hours to 202 hours.  In females, terminal elimination half-lives ranged
from approximately 2.8 hours at the lowest dose to approximately 16 hours at the high dose.  To
further characterize plasma elimination kinetics, particularly in male rats, animals were given
oral PFOA at a rate of 0.1 mg/kg, and plasma samples were collected until PFOA concentrations
fell below quantitatipn limits (24 hours and 2016 hours in females and males, respectively).
Estimated plasma elimination half-lives in this experiment were approximately 277 hours in
males and 3.4 hours in females (Tables 2 and 3).
                                         28

-------
SAR RPVJPVV Fit-aft; Hn Nnt Cite nr Oiintp
                                      Table 2
                Pharmacokinetic Parameters in Male Sprague-Dawley Rats
                Following Administration of PFOA by Gavage (Mean (SD))
Parameter
Tmax (hi)
Cmax (ng/mL)
Lambda z (l/hr)
Tin(hr)
AUCiNF (hr-ng/mL)
AUCiNF/D (hr-ng/mL/mg/kg)
Clp(mL/kg-hr)
0.1 mg/kg
10.25
(6.45)
0.598
(0.127)
0.004
(0.001)
201.774
(37.489)
123.224
(35.476)
1096.811
(310.491)
0.962
C0.240Nl
1 mg/kg
9.00
(3.83)
8.431
(1.161)
0.005
(0.001)
138.343
(31.972)
1194.463
(215.578)
1176.009
(206.316)
0.871
CO.ISS')
5 mg/kg
15.0
(10.5)
44.75
(6.14)
0.0041
(0.0007)
174.19
(28.92)
6733.70
(1392.83)
1221.89
(250.28)
0.85
(o.2n
25 mg/kg
7.5
(6.2)
160.0
(12.0)
0.0046
(0.0012)
157.47
(38.39)
25155.61
(7276.96)
942.65
(284.67)
1.13
(0.3 n
1 mg/kg
i.v.
NA
NA
0.004
(0.000)
185.584
(19.558)
1249.817
(113.167)
1123.384
(100.488)
0.896
(0.082^
0.1 mg/kg
extended
time
5.5
(7.0)
1.08
(0.42)
0.0026
(0.0007)
277.10
(56.62)
206.38
(59.03)
2111.28
(586.77)
0.51
CO. 17^
                                        29

-------
SAB Review Draft! F>n Not Cite nr Oiinfp
                                       Table 3
               Pharmacokinetic Parameters in Female Sprague-Dawley Rats
                Following Administration of PFOAby Gavage (Mean (SD))
Parameter
Tmax(hr)
Cmax(ng/mL)
Lambda z (i/hr)
T:/2
-------
SAR RPVJPW Draft; Dn Nnf Cite nr Onnfp
male and female rats (*, P<0.01: **, PO.001).

3.2.3.3 Distribution Studies in Adult Rats

The distribution of PFOA has been examined in tissues of adult rats following administration by
gavage and by i.v. and intraperitoneal (i.p.) injection. PFOA distributes primarily to the liver,
plasma, and kidney, and to a lesser extent, other tissues of the body. It does not partition to the
lipid fraction or adipose tissue.

3.2.3.3.1 Oral Exposure

Kemper (2003) examined the distribution and clearance of PFOA in tissues of male and female
Sprague-Dawley rats following administration by gavage. Rats were administered 1, 5, and 25
mg/kg 14C-PFOA by oral gavage. Tissue concentrations were determined at the time of
maximum plasma concentration (Traax) and at the time plasma concentration had fallen to one
half the maximum (Tmax/2). Values for Tmax and Tmax/2 for male and female rats were determined
from pharmacokinetic experiments. Tmax/2 was calculated as Tmax + T1/2. For cases in which
biphasic elimination was evident, the rapid phase T1/2 was used for calculation of Tmax/2. Tissues
from male rats were collected at 10.5 hours (Tmax) and 171 hours (Tmax/2) after dosing.  Tissues
from female rats were collected at 1.25 hours (Tmax) and 4 hours (Tmax/2) after dosing. The results
are summarized in Tables 5 and 6 for males and females, respectively. Liver, kidney and blood
were the primary tissues for distribution of 14C-PFOA.  In males, the fraction of the dose found
in liver increased from Tmax to Tmax/2, but remained constant or decreased in other tissues. In
females, the fraction of the dose present in all tissues remained constant or decreased between
Traax and Tmax/2;  Liver-to-blood  concentration ratios for 14C at Tmax in males were greater than 1,
and increased between Tmax and Tmax/2.  Kidney-to-blood concentration ratios at Tmax in females
were approximately 2  at all dose levels and remained relatively  constant between Tmax and Tmax/2.
                                           31

-------
SAB Rpyjpw Draft; Dn Nnf Tite ni- Oiintp
                                       Table 5
           Distribution of PFOA in Male Sprague-Dawley Rats after Oral Exposure
                   (Percent of dose recovered at Tmax and Tmax/2 in tissues)


prostate
skin"
blood8
brain
faf
heart
lungs
spleen
liver
kidney
G.I.
tract
G.I.
contents
thyroid
thymus
testes
adrenals
muscle0
bonea
Total"
1 mg/kg
T
^max
0.083±0.039
14.77±2.135
22.148±0.692
0.071±0.018
2.281±0.467
0.451±0.119
0.74±0.147
0.086±0.011
21.708±5.627
1.949±0.402
2.930±0.929
2.083±0.625
0.008±0.005
0.085±0.008
0.755±0.079
0.019±0.004
12.025±0.648
3.273±0.538
85.465±6.426
1 mg/kg
Tmax/2
0.030±0.002
6.06H0.274
8.232±1.218
0.022±0.002
0.593±0.136
0.195±0.024
0.341±0.043
0.045±0.006
32.627±3.601
1.14±0.215
0.980±0.300
0.239±0.025
0.004±0.003
0.051±0.018
0.356±0.037
O.OlOiO.OOl
4.984±0.745
1.120±0.094
57.026±3.379
5 mg/kg
T
^max
0.071±0.045
15.565±0.899
24.919±1.942
0.051±0.021
2.815±0.225
0.443±0.037
0.593±0.376
0.096±0.017
18.750±2.434
2.170±0.354
2.508±0.713
2.632±0.934
0.011±0.006
0.085±0.012
0.693±0.180
0.022±0.004
13.565±0.576
3.047±0.544
88.033±1.420
5 mg/kg
T
1 max/2
0.057±0.020
7.233±0.430
11.140±0.624
0.023±0.008
0.916±0.205
0.252±0.030
0.344±0.194
0.060±0.007
25.231±1.289
1.212±0.115
1.052±0.202
0.270±0.028
0.004±0.002
0.053±0.003
0.372±0.062
0.009±0.001
6.429±0.648
1.375±0.169
56.031±1.025
25 mg/kg
T
^max
0.067±0.018
13.836±0.969
22.905±1.177
0.063±0.007
2.153±0.430
0.461±0.053
0.863±0.103
0.106±0.015
17.528±0.900
2.293±0.286
2.784±0.608
4.186±1.349
0.009±0.002
0.120±0.025
0.623±0.098
0.026±0.004
12.855±0.841
3.062±0.438
83.937±3.680
25 mg/kg
Tmax/2
0.028±0.012
5.419±0.237
7.904±1.032
0.019±0.002
0.628±0.110
0.164±0.032
0.303±0.057
0.042±0.005
20.145±3.098
1.003±0.122
0.808±0.189
0.210±0.084
0.005±0.001
0.045±0.010
0.224±0.031
0.009±0.003
4.253±0.358
0.906±0.100
42.112±4.740
1 Percent recovery scaled to whole animal assuming the following: skin=19%,
 whole blood=7.4%, fat=7%, muscle=40.4%, bone=7.3% of body weight.
b Totals are calculated from individual animal data.
                                         32

-------
SAR RPVJPW Draff; Tin Nnt Pifp nr
                                       Table 6
          Distribution of PFOA in Female Sprague-Dawley Rats after Oral Exposure
                   (Percent of dose recovered at Tmax and Tmax/2 in tissues)
Tissue
skin8
blood3
brain
faf
heart
lungs
spleen
liver
kidney
G.I.
tract
G.I.
contents
thyroid
thymus
ovaries
adrenals
muscle3
uterus
bone3
Total"
1 mg/kg
T
'max
0.434±0.162
5.740±1.507
0.037±0.009
0.134±0.032
0.198±0.079
0.454±0.148
0.063±0.027
7.060±1.266
3.288±0.948
10.699±9.066
21.956±13.48
0.010±0.003
0.052±0.017
0.047±0.019
0.014±0.005
0.170±0.051
0.243±0.091
0.101±0.017
50.698±16.48
T
1 max/2
0.403±0.096
4.438±1.625
0.047±0.008
0.164±0.079
0.253±0.055
0.546±0.082
0.058±0.006
6.817±1.537
2.769±0.784
8.462±6.519
3.89U2.395
0.016±0.021
0.058±0.024
0.048±0.006
0.018±0.004
0.258±0.089
0.374±0.247
0.153±0.052
28.772±10.98
5 mg/kg
T
*max
0.624±0.142
8.089±2.080
0.066±0.019
0.220±0.111
0.388±0.057
0.827±0.102
0.10H0.021
11.190±2.192
4.293±0.771
7.142±2.594
2.896±2.305
0.008±0.002
0.105±0.030
0.07U0.012
0.026±0.005
0.325±0.010
0.354±0.046
0.174±0.057
36.897±3.187
T
1 max/2
0.307±0.121
5.41 1±1. 466
0.045±0.010
0.110±0.069
0.236±0.051
0.570±0.179
0.060±0.012
7.176±0.982
2.685±0.736
8.255±8.967
5.60U6.165
0.006±0.002
0.068±0.021
0.041±0.012
0.015±0.004
0.229±0.031
0.247±0.068
0.142±0.078
31.201±12.63
25 mg/kg
T
'max
0.380±0.166
7.158±2.232
0.058±0.008
0.147±0.053
0.317±0.035
0.678±0.067
0.091±0.007
10.538±1.723
5.867±0.946
6.923±1.846
2.491±1.548
0.009±0.003
0.091±0.032
0.071±0.012
0.031±0.005
0.441±0.116
0.358±0.124
0.157±0.072
35.803±2.554
T
1 max/2
0.415±0.175
6.407±1.406
0.058±0.018
0.148±0.065
0.287±0.069
0.775±0.204
0.070±0.002
9.080±0.895
4.749±0.393
3.547±1.306
1.121±1.010
0.007±0.002
0.077±0.020
0.070±0.012
0.02U0.001
0.304±0.099
0.365±0.029
0.181±0.090
27.680±2.569
 Percent recovery scaled to whole animal assuming the following: skin=19%,  whole
blood=7.4%, fat=7%, muscle=40.4%, bone=7.3% of body weight.
b Totals are calculated from individual animal data.
                                         33

-------
SAR RPVJPW Draft; f)n Nnt rite ni-
3.2.3.3.2 Intravenous Exposure

Gibson and Johnson (1980) examined the tissue distribution in 10-week old male CD rats
following a single i.v. dose of 13.1 mg/kg 14C-PFOA.  The animals were sacrificed 36 days after
dosing.  The results are presented in Table 7. PFOA distributed mainly to the liver, plasma and
kidney.

                                       Table 7
                 Distribution of PFOA in Male CD Rats After i.v. Exposure
Tissue
Liver
Plasma
Kidney
Lung
Red Blood Cells
Skin
Spleen
Bone marrow
Subcutaneous Fat
Muscle
Brain
Abdominal Fat
Mean ± SDa
7.97 + 4.02
3.19 + 1.72
2.33 ±1.13
0.79 + 0.51
0.77 + 0.45
0.40 + 0.28
0.33 ±0.1 9
0.33 ± 0.26
0.28 ±0.1 8
0.20 ±0.09
0.11 ±0.04
<0.12
                    a - Total 14C concentration expressed as \ig equivalents of 14C-PFOA.

3.2.3.3.3 Intraperitoneal Exposure

Ylinen et al. (1990) studied the distribution and elimination of PFOA in a limited sample of
tissues following a single i.p. dose of 50 mg/kg in 10 week old Wistar rats (20/sex). Samples
were collected from 2 animals per sex for analysis of PFOA at 12, 24 -168 (in 24 hour
intervals), 224 and 336 hours after dosing. The concentration of PFOA in the serum and tissues
was determined with capillary gas chromatography equipped with a flame ionization detector
(FID). A mass spectrometer was used in the selected ion monitoring mode when the PFOA
concentration was below the quantitation limit of the FID (1 ^.g/ml).  The half-life of PFOA was
estimated from the linear regression of time and concentration of PFOA in a semilogarithmic
plot. No PFOA was detected in the adipose tissue. In both sexes at 12 hours after
administration, the highest concentration of PFOA was found in the serum, followed by the liver,
kidney, spleen and brain.  In the females, the concentration of PFOA in the serum, liver and
kidney decreased in a discontinuous fashion indicating distinct phases. The half-life of PFOA in
                                          34

-------
SAR RPVIPW Draft; Tin Nnf Pi to nr Onntp
the serum, liver, kidney and spleen is presented in Table 8; the values are generally much lower
in females than in males.

                                        Table 8
              Mean Half-life (hours) of PFOA in Wistar Rats After i.p. Exposure
Tissue
Serum
Liver
Kidney
Spleen
Males
105
210
130
170
Females
24
60
145
73
Vanden Heuvel et al. (1991b) examined the distribution of PFOA in the serum and a limited
number of tissues following administration of 9.4 umol/kg 14C-PFOA by i.p. injection to 6-week
old male and female Harlan Sprague-Dawley rats. At various time-points for 28 days following
treatment, four rats per sex were sacrificed; blood was collected by cardiac puncture, and tissues
were removed and frozen. 14C-PFOA-derived radioactivity was quantitated using a Liquid
Scintillation Analyzer. Assuming first order kinetics, tissue elimination rates were calculated as
T,A = In2/ke. The distribution and elimination of PFOA-derived 14C in selected tissues is
summarized in Table 9. In the male rats, 21% of the administered dose was present in the liver,
and the next highest concentrations were found in the plasma and kidney. Far lower PFOA
concentrations were found in the heart, testis, fat, and gastrocnemius muscle. In females, the
highest concentrations of PFOA were found in the plasma followed by the kidney,  liver and
ovaries.  In males, PFOA was eliminated from the liver at a slower rate than the other tissues;
the T,/2 for liver was 11 days compared to 8-9 days in most extrahepatic tissues.  The rates of
elimination were much faster in the female rats than in the male rats.

The high concentration of PFOA in the male liver was further examined using a liver perfusion
technique.  Liver was infused with 0.08 umol  C-PFOA/min over a 48 min period for a total of
3.84 umol 14C.  Approximately 11% of the cumulative dose of 14C-PFOA infused was extracted
by the liver in a first pass. At 2 min, the cumulative percent of PFOA extracted by the liver was
33%; that was substantially greater than the 11% cumulative dose of  C that was extracted after
48 min indicating that first-pass hepatic uptake of PFOA may be saturable.
                                          35

-------
SAR Review Draft! Do Nnt Tiff nr
                                       Table 9
   Distribution and Elimination of PFOA in Selected Tissues in Harlan Sprague-Dawley Rat
                                  after i.p. Exposure8
Tissue
Liver
Plasma
Kidney
Heart
Gastrocnemius
Fatc
Testis/
Ovary
% Dose8
Male
(2hrs)
2.03
(0.10)
1.99
(0.03)
0.95
(0.08)
0.42
(0.02)
0.26
(0.01)
0.32
(0.02)
0.33
(0.01)
Female
(2hrs)
1.53
(0.25)
2.39
(0.44)
2.00
(0.38)
ND
ND
ND
0.53
(0.10)
Male
(24 hrs)
2.08
(0.14)
1.63
(0.11)
0.74
(0.03)
0.39
(0.03)
0.21
(0.01)
0.27
(0.04)
0.27
(0.03)
Female
(24 hrs)
0.06
(0.02)
0.02
0.06
(0.02)
ND
ND
ND
0.05
(0.01)
Elimination Rateb
Male
(day1)
0.062
(0.005)
0.077
(0.006)
0.074
(0.005)
0.066
(0.006)
0.076
(0.007)
0.087
(0.008)
0.078
(0.006)
Female
(hour •')
0.185
(0.011)
0.242
(0.007)
0.214
(0.012)
ND
ND
ND
ND
Half-life
Male
(day)
11.3
9.0
9.4
10.4
9.2
8.0
9.0
Female
(hour)
3.8
2.9
3.2
ND
ND
ND
ND
^D - Not determined
a- Values are percent dose of PFOA-derived 14C per gram tissue; mean (SEM); n=4; values with
no SEM indicate that SEM < 0.01
b- Mean (SEM); n=32 for males and n=16 for females
c-Epididymal fat pad

3.2.3.4 Metabolism Studies in Adult Rats

Several studies have examined metabolism of PFOA. However, no studies show clear evidence
of metabolism. Ophaug and Singer (1980) found no change in ionic fluoride level in the serum
or urine following oral administration of PFOA to female Holtzman rats.  Ylinen et al. (1989)
found no evidence of phase II metabolism of PFOA following a single intraperitoneal PFOA
dose (50 mg/kg) in male and female Wistar rats.

Vanden Heuvel et al. (1991b) investigated the metabolism of PFOA in Harlan Sprague-Dawley
rats administered 14C-PFOA (9.4 ^mol/kg, i.p.). Pooled daily urine samples (0-4 days post-
treatment) and bile extracts analyzed  by HPLC contained a single radioactive peak eluting
identically to the parent compound. Tissues were taken from rats treated 4,14, and 28 days after
treatment to determine the presence of PFOA-containing lipid conjugates. Only the parent
compound was present in rat tissues; no PFOA-containing hybrid lipids were detected.  Fluoride
concentrations in plasma and urine before and after PFOA treatment were unchanged, indicating
that PFOA does not undergo defluorination.
                                         36

-------
SAB RPVJPW Draft; Tin Not Cite nr Quote
3.2.3.5 Elimination Studies in Adult Rats

In adult rats, there is evidence of enterohepatic circulation of PFOA.  The urine is the major
route of excretion of PFOA in the female rat, while the urine and feces are both main routes of
excretion in male rats. There are gender differences in the elimination of PFOA in rats.  The
rapid excretion of PFOA by female rats is due to active renal tubular secretion (organic acid
transport system); this renal tubular secretion is believed to be hormonaUy controlled. Hormonal
changes during pregnancy do not appear to change the rate of elimination in rats.

3.2.3.5.1 Enterohepatic Circulation

Johnson et al. (1984)  investigated the effect of feeding cholestyramine to rats on the elimination
of APFO.  Since APFO exists as an anion at physiologic pH, it would be expected to complex
with cholestyramine.  Ten male Charles River CD rats, 12 weeks of age, were given a single i.v.
injection of 13 mg/kg 14C-APFO.  Five rats were given 4% cholestyramine in feed.  Urine and
feces samples were collected at intervals for 14 days, at which time the animals were sacrificed
and liver samples were collected.  At 14 days post dose, the mean percentage of PFOA
eliminated in the feces of cholestyramine-treated rats was 9.8-fold the mean percentage
eliminated in the feces of rats that did not receive cholestyramine.  Excretion in urine was 41%
of the administered dose for cholestyramine treated rats and 67% for rats that did not receive
cholestyramine. 14C in the liver equaled 4% or 12.1  ± 2.1 \ig eq/g in cholestyramine treated rats
and 8 % or 22.3 ± 6.2 \ig eq/g in rats that did not receive cholestryamine. In plasma, the levels
were 5.1 ± 1.7 |ig eq/ml in cholestyramine treated rats and 14.7±6.8 \Lg eq/ml in rats that did not
receive cholestyramine.  In red blood cells, the levels were 1.8 ± 0.7 |ig eq/ml in cholestryamine
treated rats and 4.2 ±  2.4 jig eq/ml in rats that did receive cholestryamine. The high
concentration of 14C-APFO in the liver at 2 weeks after dosing and the fact that cholestyramine
treatment enhances fecal elimination of 14C nearly 10-fold suggests that there is enterohepatic
circulation of PFOA.

3.2.3.5.2 General Elimination Studies

3.2.3.5.2.1 Oral Exposure

Kemper (2003) investigated the elimination of PFOA in male and female Sprague-Dawley rats
(4 rats/sex/group) administered a single dose of 14C-PFOA by oral gavage at dose levels of 1, 5,
and 25 mg/kg. Urine and feces were collected for 28 days in males and 7 days in females. Urine
was the primary route of excretion of 14C in both sexes, accounting for 43-62% of the
administered dose in males and 76-84% of the administered dose in females. Cumulative
recovery of 14C in the urine increased gradually over the 28 days in male rats, but was essentially
complete in female rats within the first 72 hours. Fecal excretion of 14C accounted for 6-14% of
the dose in males and 2-6% of the dose in females. Pilot experiments demonstrated that 14C was
not eliminated as either 14C02 or volatile organic compounds in 14C-PFOA-treated rats.
Pretreatment of rats with 1 mg/kg-day PFOA for 14 days had little or no effect on the excretion
of a challenge dose of 1 mg/kg 14C-PFOA.

3.2.3.5.2.2 Intravenous Exposure

Gibson and Johnson (1980) examined the excretion of total 14C in male and female CD rats after
a single i.v. dose of 14C-PFOA.  The mean dose for females was 16.7 mg/kg while that for males
was 13.1 mg/kg. Female rats excreted essentially all of the administered dose via the urine in the

                                          37

-------
SAR RPVJPW Draft; Hn Not Pitt* nr Ounte
24 hours after treatment.  During the same time period, male rats excreted only 20% of the total
dose. Male rats excreted 83% of the total dose via the urine and 5.4% via the feces by 36 days
post dose.  No radioactivity was detected in tissues of female rats at 17 days post dose; 2.8% of
the total dose was detected in the liver of male rats and 1.1% in the plasma at 36 days post dose
with lower levels equaling < 0.5% of the total dose in other organs.

3.2.3.5.3 Elimination Studies in the Pregnant Rat

Hormonal  changes during pregnancy do not appear to cause a change in the rate of elimination
of 14C after oral administration of a single dose of 14C-APFO (Gibson and Johnson, 1983). At 8
or 9 days after conception, four pregnant CD rats and two nonpregnant female CD rats were
given a mean dose of 15 mg/kg l4C-APFO.  Individual urine samples were collected at 12,24,
36, and 48 hours post dose and analyzed for 14C content. Essentially all of the 14C was
eliminated via the urine within 24 hours for both groups of rats.

3.2.3.5.4 Studies on the Mechanism of the Gender Difference in Elimination in Adult Rats

Several studies have been conducted to elucidate the cause of the gender difference in rats in the
elimination of PFOA. Hanhijarvi et al. (1982) conducted a series of studies to examine the
effect of probenecid, which inhibits the renal active secretion system for organic acids, on the
elimination of PFOA in male and female Holtzman rats. In the first study, 4 male and 6 female
Holtzman rats were administered 2 mg of nonionic fluorine as PFOA by gavage. Seven female
rats were administered 2 ml distilled water as controls.  The animals were then placed in
metabolism cages and urine was collected until the animals were sacrificed at 24 hr by cardiac
puncture.  Serum was collected. Ionic fluoride and total fluorine content of serum and urine
were determined, and nonionic fluorine was calculated  as the difference. Twenty-four hours
after oral administration of PFOA, female rats had excreted 76 ± 2.7% of the dose in the urine
and had a mean serum nonionic fluorine level of 0.35 ± 0.11 [ig/ml, while male rats had excreted
only 9.2 ±  3.5% of the dose and had a mean serum nonionic fluorine level of 44.0 ±1.7 jig/ml.
A mean of 97.5±0.25% of the PFOA was bound in the plasma of both male and female rats.

In the second study, Hanhijarvi et al. (1982) examined the effect of probenecid on the clearance
of PFOA and inulin. Holtzman rats  were anesthetized  and the femoral artery was  cannulated  for
continuous infusion of 5% mannitol in isotonic saline., while the femoral artery was cannulated
for drawing blood samples. The urinary bladder was also cannulated for serial collections of
urine. When the urine and serum collections for the clearance study were complete, 65-68
mg/kg probenecid was administered by i.p. injection and after 20 - 30 minutes, additional 10
minute clearance tests were performed. Administration of probenecid reduced the PFOA/inulin
clearance ratio in females from 14.5 to 0.46. PFOA clearance was reduced from 5.8 to 0.11
ml/min/lOOg. Net PFOA excretion was reduced from 4.6 jig/min/lOGg to 0.13 iig/min/lOOg.  In
male rats, however, the PFOA/inulin clearance ratio and the net excretion of PFOA were
virtually unaffected by probenecid. In the males, PFOA clearance was 0.17 ml/min/lOOg, the
PFOA/inulin clearance ratio was 0.22, and net PFOA excretion was 0.17 jig/min/ing.

Finally, Hanhijarvi et al. (1982) examined the cumulative excretion of PFOA over a 7-hour
period.  Holtzman rats were dosed i.v. with a mixture of 10%-20% radiolabeled-PFOA  and 80-
90% unlabeled PFOA. Mannitol (5%) was infused and  urine specimens were collected over 30-
min intervals. The effect of probenecid was assessed by administering 65-68 mg/kg by i.p.
injection at least 30 min prior to the administration of PFOA.  Female rats excreted 76% of the
administered dose of PFOA, while males excreted only 7.8% of the administered dose over a 7-

                                         38

-------
SAR RPVJPW Draft! Do Nnf rifp or Oiinfp
hr period.  Probenecid administration modified the cumulative excretion curve for males only
slightly. However, in females probenecid markedly reduced PFOA elimination to 1 1.8%.  The
authors concluded that the female rat possesses an active secretory mechanism which rapidly
eliminates PFOA from the body.

Ylinen et al. (1989) studied the urinary excretion of PFOA in male Wistar rats after castration
and estradiol administration. Twenty male rats were castrated at 28 days of age and were used in
tests of PFOA excretion 5 weeks later. Ten castrated and 10 intact males were given 500 jig/kg
estradiol valerate by s.c. injection every second day for 14 days before administration of PFOA.
PFOA was administered as a single i.p. injection at 50 mg/kg. Urine was collected in metabolism
cages for 96 hr after PFOA administration.  Blood samples were collected by  cardiac puncture.
Six female rats were also included in the experiment. Castration and administration of estradiol
to the male rats had a significant stimulatory effect on the urinary excretion of PFOA. During the
first 24 hours, female rats excreted 72 ± 5% of the administered dose of PFOA, whereas the
intact males  excreted only 9 ± 4%.  After the estradiol treatment, both the intact and castrated
males excreted PFOA in amounts similar to females, 61 ± 19% and 68 ± 14%, respectively. The
castrated males without estradiol treatment excreted 50 ± 13% of the administered dose of PFOA
in the urine.  This was faster than the intact males but less than the females and the estrogen
treated males. At the end of the test, the concentration of PFOA in the serum of intact males was
17- 40 times higher than the concentration PFOA in the serum of other groups. There were no
statistically significant differences in the serum concentrations between the other groups.  PFOA
was similarly bound by the proteins in the serum of males and females.

Vanden Heuvel et al. (1992a) investigated whether androgens or estrogens are involved in the
marked sex-differences in the urinary excretion of PFOA.  Castrated Harlan Sprague-Dawley
male rats were given 9.4 iimol/kg 14C-PFOA by i.p. injection. Castration increased the
elimination of PFOA in the urine (36% of the dose was eliminated in 4 days versus 16% in
controls), suggesting that a factor produced by the testis is responsible for the slow elimination
of PFOA in male rats. Castration plus 17B-estradiol had no further effect on PFOA elimination
whereas castration plus testosterone replacement at the physiological level reduced PFOA
elimination to the same level as rats with intact testis. Thus, in male rats, testosterone exerts an
inhibitory effect on renal excretion of PFOA. In female rats, neither ovariectomy or
ovariectomy plus testosterone affected the urinary excretion of PFOA, demonstrating that the
inhibitory effect of testosterone on PFOA renal excretion is a male-specific response.
Probenecid,  which inhibits the renal transport system, decreased the high rate of PFOA renal
excretion in  castrated males but had no effect on male rats with intact testis.
Kudo et al. (2002) examined the role of sex hormones on the renal clearance (CLjJ of PFOA and
the renal mRNA levels of specific organic anion transporters in male and female Wistar rats.
Castration of male rats caused a 14-fold increase in CLR of PFOA. The elevated PFOA CLR in
castrated males was reduced by treating them with testosterone. Treatment of male rats with
estradiol increased the CLR of PFOA. In female rats, ovariectomy caused a significant increase
in CLR of PFOA, which was reduced by estradiol treatment.  Treatments of female rats with
testosterone reduced the CLR of PFOA. Treatment with probenecid, a known inhibitor of
organic anion transporters, markedly reduced the CLR of PFOA in male rats, castrated male rats,
and female rats. To identify the transporter molecules that are responsible for PFOA transport in
the rat kidney, renal mRNA levels of specific organic anion transporters were determined in
male and female rats under various hormonal states and compared with the CLR of PFOA. The
level of OAT2 mRNA in male rats was only 13% that in female rats. Castration or estradiol
treatment increased the level of OAT2 mRNA whereas treatment of castrated male rats with

                                          39

-------
SAR RPVJPW Draff! T)n Nnt Titf nr Oimfp
testosterone reduced it.  Ovariectomy of female rats significantly increased the level of OATS
mRNA. Multiple regression analysis of the data suggested that organic anion transporter 2
(OAT2) and OATS are responsible for urinary elimination of PFOA in the rat.

3.2.4 Metabolism and Pharmacokinetic Studies in Immature Rats

No studies have been conducted to specifically examine the absorption, metabolism or
elimination of PFOA in the developing rat.  However, recent studies have been conducted to
examine the concentrations of PFOA in the developing Sprague-Dawley rat, and to determine
when the gender difference in elimination of PFOA becomes apparent.  In addition, several
studies have examined the serum and tissue distribution of PFOA in newly weaned Wistar rats.
These studies have shown that PFOA readily crosses the placenta and is present in the breast
milk of rats. During lactation and immediately after weaning, the elimination of PFOA is similar
in males and females.  In the male rat between 4-5 weeks of age, the factor(s) responsible for the
gender difference develop, and the rats assume the adult male elimination profile.  In addition,
distribution studies in the postweaning rat have shown, that PFOA is distributed primarily to the
serum, liver, and kidney.

3.2.4.1 PFOA Levels During Pregnancy and Lactation

Mylchreest (2003) examined PFOA levels during gestation and lactation.  Pregnant Sprague-
Dawley rats were dosed with  0, 3,10 or 30  mg/kg-day APFO during days 4-10, 4-15, or 4-21  of
gestation, or from gestation day 4 to  lactation day  21. Clinical observations and body weights
were recorded daily. On gestation days 10,15, and 21, 5 rats/group/time point were sacrificed
and the number, location and type of implantation sites were recorded. Embryos were collected
on day 10, and placentas, amniotic fluid, and embryos/fetuses were collected on days 15 and 21.
Maternal blood samples were collected at 2 hours + 30 minutes post-dose.  The remaining 5 rats
per group were allowed to deliver. On lactation days 0, 3,7,14, and 21, the pups were counted,
weighed (sexes separate), and examined for abnormal appearance and behavior. Randomly
selected pups  were sacrificed  and blood samples were collected.  On lactation days 3, 7,14, and
21, the dams were anesthetized and milk and blood samples were collected; dams were removed
from their litters 1-2 hours prior to collection. Plasma, milk, amniotic fluid extract, and tissue
homogenates  (placenta, embryo, and fetus)  supernatants were analyzed for PFOA concentrations
by HPLC-MS.

All dams survived and there were no clinical signs of toxicity.  In the 30 mgykg-day group, mean
body weight gain was approximately 10% lower than the control group during gestation, and
mean body weights were approximately 4% lower than controls throughout gestation and
lactation. The number of implantation sites, resorptions, and live fetuses were comparable
among groups on days 10,15, and 21 of gestation.  One dam in the 3 mg/kg-day group and two
dams in the 30 mg/kg-day group delivered small litters (litter size of 3-6 pups as compared to 12-
19 pups/litter  in the control group); however, given the small sample size the biological
significance of this finding is  unclear.  There were no clinical signs of toxicity in the pups, and
pup survival and pup body weights were comparable among groups.

Maternal PFOA levels during gestation and lactation are presented in Table 10.  Maternal plasma
levels at 2 hrs post-dosing (approximately the time of peak blood levels following a gavage
dose) were fairly similar during the course of the study with a mean level of 11.2, 26.8, and 66.6
jig/ml in the 3,10, and 30 mg/kg-day groups, respectively; PFOA levels in the control group
were below the limit of quantitation (0.05 jig/ml).  The concentration of PFOA in the milk was

                                          40

-------
SAR RPVJPW Draft; Tin Nnt Cite nr Oiinfp
also fairly similar throughout lactation and was approximately 1/1 Oth of the PFOA levels in the
plasma; the mean values were 1.1, 2.8, and 6.2 fig/ml in the 3,10, and 30 mg/kg-day groups,
respectively.

                                       Table 10
               Maternal PFOA Levels (ng/ml) During Gestation and Lactation3
Exposure Period
GD 4 - GD 10
GD 4 - GD 15
GD 4 - GD 21
GD 4 - LD 3
GD 4 - LD 7
GD 4 - LD 14
GD 4 - LD 21
NA
Sample Time
GD 10 plasma
GDI 5 plasma
GD 21 plasma
LD 3 - plasma
-milk
LD 7 - plasma
-milk
LD 14 -plasma
- milk
LD 21 - plasma
- milk
Average plasma
Average milk
3 mg/kg-day
8.53 + 1.06
15.92 + 12.96
14.04 + 2.27
11.01+2.11
1.07 ±0.26
10.09 + 2.90
0.94 ± 0.22
9.69 + 0.92
1.15 + 0.06
9.04 + 1.01
1.13 + 0.08
11.19 + 2.76
1.07 + 0.09
10 mg/kg-day
23.32 + 2.15
29.40+14.19
34.20 + 6.68
22.47 + 2.74
2.03 ±0.33
25.83 + 2.07
2. 74 ±0.91
23.79 + 2,81
3.45 ±1.18
28.84 + 5.15
3.07 + 0.51
26.84 + 4.21
2. 82 ±0.60
30 mg/kg-day
70.49 + 8.94
79.55 ±3.11
76.36 ±14.76
54.39 + 17.86
4.97 ±1.20
66.91 + 11.82
5.76 ±1.26
54.65 + 11.63
6.45 ±1.38
64.13 + 1.45
7.48 ±1.63
66.64 + 9.80
6.16 ±1.06
a- mean ± SD; samples were from 5 dams/group/time point and were collected 2 hrs post-dosing

PFOA levels in the placenta, amniotic fluid, embryo, fetus, and pup plasma are presented in
Table 11.  The levels of PFOA in the placenta on gestation day 21 were approximately twice the
levels observed on gestation day 15, and the levels of PFOA in the amniotic fluid were
approximately four times higher on day 21 than on day 15. The concentration of PFOA in the
embryo/fetus was highest in the day 10 embryo and lowest in the day 15 embryo; PFOA levels
in the day 21 fetus were intermediate. The concentration of PFOA in the plasma of the day 21
fetus were approximately half the levels observed in the maternal plasma; the mean values were
5.9,14.5, and 33.1 jig/ml in the 3,10, and 30 mg/kg-day groups, respectively.  Pup plasma levels
decreased until lactation day 7, and were thereafter similar to the levels observed in the milk.
                                          41

-------
SAB Review Draft; fin Not fife nt- Oiintp
                                      Table 11
    PFOA Concentrations (fig/ml) During Gestation and Lactation in Sprague-Dawley Ratsa
Exposure Period
GD 4 - GD 10
GD 4 - GD 15
GD 4 - GD 21
GD 4 - LD 3
GD 4 - LD 7
GD 4 - LD 14
GD 4 - LD 21
Tissue
GD 10 -embryo
GD 15 - placenta
- amniotic fluid
- embryo
GD 21 - placenta
- amniotic fluid
-fetus
- fetal plasma
LD 3 - pup plasma
LD 7 - pup plasma
LD 14 -pup plasma
LD 21 - pupjplasma
3 mg/kg-day
1.40 + 0.30
2.22+1.79
0.60 + 0.69
0.24 + 0.19
3.55 + 0.57
1.50 + 0.32
1.27 + 0.26
5.88 + 0.69
2.89±0'.70
0.65 ± 0.20
0.77 ±0.10
1.28 + 0.72
10 mg/kg-day
3.33 + 0.81
5.10 + 1.70
0.70 + 0.15
0.53 + 0.18
9.37 + 1.76
3.76 + 0.81
2.61+0.37
14.48 ±1.51
5.94 ±1.44
2.77 + 0.58
2.22 ±0.38
3.25 + 0.52
30 mg/kg-day
12.49 ±3.50
13.22 + 1.03
1.70 + 0.91
1. 24 ±0.22
24.37 + 4.13
8.13 + 0.86
8.77 + 2.36
33.11 ±4. 64
11.96 + 1.66
4.92 + 1.28
4.91 ±1.12
7.36±2.17
a - mean ± SD; samples were pooled by litter and were collected 2 hrs post-dosing
3.2.4.2 PFOA Levels in the Postweaning Rat

Han (2003) examined the relationship between age and plasma PFOA concentrations in the
postweaning Sprague-Dawley rat. Four to eight week old rats (10/sex/time period) were
administered a single dose of 10 mg/kg-day APFO by gavage. Blood samples were collected 24
hours after dosing and the plasma concentration of PFOA was measured by HPLC-MS. At four
weeks of age, the concentration of plasma PFOA was approximately 2.7 times higher in males
than in the females (Table 12). In females, the concentration of plasma PFOA decreased by 2.7
fold between 4 and 5 weeks of age, and thereafter remained fairly steady. In males, the
concentration of plasma PFOA increased by 5.4 fold between 4 and 5 weeks of age, and
thereafter remained fairly steady. Between 5 and 8 weeks of age, the PFOA plasma
concentrations were 34.7 - 65.1 fold higher in males than in females of the same age.  Thus, it
appears that the elimination of PFOA is similar among males and females until week 4, and
between weeks 4 and 5 the maturation of the factor(s) responsible for the gender difference in
elimination of PFOA occurs in the male rat.
                                         42

-------
SAR Rpyjpw Draft; Dn Nnt fite nr Oiinfp
                                       Table 12
         Plasma PFOA Concentrations (tig/ml) in Postweaning Sprague-Dawley Ratsa
Age (weeks)
4
5
6
7
8
Males
7.32 + 1.01
39.24 + 3.89
43.19 + 3.79
37.12 ±4.07
38.55 + 5.44
Females
2. 68 ±0.64
1.13 ±0.46
1.18 + 0.52
0.57 ±0.29
0.81 + 0.27
a - mean + SD; samples from 10 animals/sex/group

3.2.4.3 Serum and Tissue Distribution in Immature Wistar Rats Following Oral Exposure

Ylinen et al. (1990) administered newly weaned Wistar rats (18/sex/group) doses of 3,10, and
30 mg/kg-day PFOA by gavage for 28 days.  At necropsy, serum was collected as well as the
brain, liver, kidney, lung, spleen, ovary, testis, and adipose tissue.  The concentration of PFOA
in the serum and tissues was determined with capillary gas chromatography equipped with a
flame ionization detector (FID). A mass spectrometer was used in the selected ion monitoring
mode when the PFOA concentration was below the quantitation limit of the FID (1  jig/ml). The
concentration of PFOA in the serum and tissues following 28 days of administration is presented
in Table 13. PFOA was not detected in the adipose tissue. The concentrations of PFOA in the
serum and tissues were much higher in males than in females. In the males, the levels of PFOA
in the serum and tissues were generally lower in the 30 mg/kg-day group than in the 10 mg/kg-
day group due to increased urinary elimination in the 30 mg/kg-day group.
                                          43

-------
SAR RPVJPW Draff; F>n Nnt fife nr
                                      Table 13
           Tissue Distribution of PFOA in Wistar Rats after 28 Days of Treatment
Tissue
Serum
Liver
Kidney
Spleen
Lung
Brain
Ovary
Testis
Males3
3
mg/kg-day
48.6±10.3
39.9+7.25
1.55+0.71
4.75+1.66
2.95+0.54
0.398±0.144

6.24 + 2.04
10
mg/kg-day
87.27±20.09
51.71+11.18
40.56+14.94
7.59+3.5
22.58+4.59
1.464±0.211

9.35 + 4.02
30
mg/kg-day
51.65±1.47
49.77+10.76
39.81+17.67
4.1+1.57
23.71+5.42
0.71+0.32

7.22 + 3.17
Females3
3
mg/kg-day
2.4b
1.81+0.49
0.06+0.02
0.15+0.04
0.24b

-------
SAR RPVJPW Draft; fin Nnf Pitt* nr Quote
                                      Table 14
                  % Protein Binding to Rat, Human and Monkey Plasma
Nominal Concentration
(ppm)
1
10
100
250
500
Rat
-100
99.5
98.6
97.6
97.3
Monkey
-100
99.8
99.8
99.8
99.5
Human
-100
99.9
99.9
99.6
99.4
% Binding values reported as "-100" reflect a nonquantifiable amount of test article in the
plasma water BQL<6.25 ng/ml

Han et al. (2003) investigated the binding of PFOA to rat and human plasma proteins in vitro.
Rats treated in vivo and then sacrificed showed no gender difference in the binding of PFOA to
serum., though the persistence of PFOA in vivo is much greater in male than female rats.  The
authors conclude that there is no correlation between the PFOA persistence and binding of the
PFOA to rat serum. The primary PFOA binding protein in plasma was serum albumin.
However, the method used (ligand blotting) would not theoretically allow the identification of
low abundance proteins with high affinity for PFOA. Further investigation of purified rodent
and human serum albumin binding using labeled 19F NMR allowed the calculation of
disassociation constants for PFOA binding to rodent and human serum albumin. No significant
differences between binding to the two proteins was detected (Table 15).

                                      Table 15
 Dissociation Constants (Kd) of Binding between PFOA and Rodent Serum Albumin (RSA) and
  Human Serum Albumin (HSA) and the Number of PFOA Binding Sites (ri) on RSA and HSA
Parameter
Kd(mM)
Kd(mM)
n
Method
NMRa
micro-SECb
micro-SECb
RSA
0.29±0.10C
0.36 + 0.08°
7.8 + 1.5
HSA

0.38 + 0.04
7.2+1.3
"Average of the two Rvalues (0.31 + 0.15 and 0.27 + 0.05 mM) obtained by NMR. "Values
were obtained from three independent experiments and their standard deviations are shown. °0n
the basis of the result of unpaired t-test at 95% confidence interval, the difference of A^ values
determined by NMR and micro-SEC is statistically insignificant.

3.2.6  Metabolism and Pharmacokinetic Studies in Other Test Species

There is limited information on the metabolism and pharmacokinetics of PFOA in mice, rabbits
and dogs.  No specific pharmacokinetic studies have been conducted in mice.  However,
toxicology studies in mice indicate that PFOA is absorbed,  and furthermore, there does not
                                         45

-------
SAB Review Draff? fin Not fite nr Quote
appear to be a gender difference in elimination. For example, Sohlenius et al. (1992) exposed
male and female C57B1/6 mice to dietary levels of 0.02% PFOA for one week. There was a
significant decrease in mean body weight and a significant increase in absolute liver weight; the
response was similar in male and female mice.

In rabbits, there is no information available on the metabolism or elimination of PFOA by any
route of exposure. No specific studies of the absorption of PFOA have been conducted
following oral or inhalation exposure. However, there is evidence that PFOA is absorbed
following dermal exposure.  O'Malley and Ebbins (1981) treated male and female New Zealand
White rabbits dermally with doses of 100,1,000 and 2,000 mg/kg APFO for 14 days. Mortality
was 100% (4/4) in the 2,000 mg/kg group, 75% (3/4) in the 1,000 mg/kg group and 0% (0/4) in
the 100 mg/kg group.  Similarly, Kennedy (1985) treated rabbits dermally with a total of 10
applications of APFO at doses of 0, 20, 200 or 2,000 mg/kg.  Treatment resulted in elevated
blood organofluorine levels that increased in a dose-related manner.

In addition, limited information is available on the serum and liver distribution of PFOA in
rabbits following i.v. administration. Johnson (1995a) administered individual female rabbits
intravenous doses of 0, 4,16, 24 and 40 mg/kg tetrabutyl ammonium salt of PFOA. The animal
given 40 mg/kg died within 5 minutes of treatment.  All other animals appeared normal
throughout the study.  Serum samples were analyzed for total organic fluorine at 2, 4, 6, 8,12,
24, and 48 hours post dose. At 2 hrs, serum organic fluorine levels in the rabbits that received 0,
4,16, and 24 mg/kg were 1.25 ng/ml, 4.09 ng/ml, 14.9 \ig/rd, and 41.0 ng/ml, respectively.
There was a rapid decrease of total organic fluorine in the serum with tune; it was non-detectable
at 48 hr. The biological half-life was on the order of 4 hours. The total organic fluorine levels in
whole liver at 48 hr post dose for the rabbits that received 0 mg/kg, 4 mg/kg, 16 mg/kg, and 24
mg/kg were 20 |ig, 43 p,g, 66 \ig, and 54 \ig, respectively.

There is no  information on the specific absorption, metabolism or distribution of PFOA in dogs
by any route of exposure.  One study examined the elimination of PFOA in dogs following i.v.
administration.  Hanhijarvi et al. (1988) administered beagle dogs (3/sex) an i.v. injection of 30
mg/kg of PFOA followed by continuous infusion with 5% mannitol.  Urine and blood were
collected at 10 minute intervals for 60 min. Probenecid was then administered by i.v. injection,
and urine and blood samples were collected as before.  Renal clearance of PFOA was calculated
for the before and after probenecid injection periods. Four additional dogs (2/sex) were given 30
mg/kg of PFOA by i.v. injection. These dogs were kept in metabolism cages, .and blood samples
were collected intermittently for 30 days. The renal clearance rate was approximately  0.03
ml/min/kg.  Probenecid significantly reduced the PFOA clearance rate in both sexes, indicating
an active secretion mechanism for PFOA. The plasma half-life of PFOA was 473 hr before
probenecid  administration and 541 hr after in male dogs, and 202 hr before probenecid and 305
hr after in the female dogs.

3.3 Acute Toxicity Studies in Animals

Dean and Jessup (1978) reported an oral LD50 of 680 mg/kg and 430 mg/kg for male and female
CD rats, respectively. Glaza (1997) reported an oral LD50 of greater than 500 mg/kg in male
Sprague-Dawley rats and between 250 and 500 mg/kg in females. Gabriel (1976d) reported an
oral LD50 of less than 1000 mg/kg for male and female Sherman-Wistar rats.  Rusch (1979)
reported no mortality  in male or female Sprague-Dawley rats following inhalation exposure to
18.6 mg/L for one hour.  The dermal LD50 in New Zealand White rabbits was determined to be
greater than 2000 mg/kg (Glaza, 1995).

                                          46

-------
SAR RPVIPW Draff' Fin Nnf dtp nr Onnfp
APFO is an ocular irritant in rabbits when the compound is not washed from the eyes (Gabriel,
1976b, 1976e), but is not an irritant in rabbits when washed from the eye (Gabriel, 1976a).
Markoe (1983) found APFO to be a skin irritant in rabbits, while Gabriel (1976c) did not.

3.4 Mutagenicity Studies

APFO was tested twice (Lawlor, 1995; 1996) for its ability to induce mutation in the Salmonella
- E. co///mammalian-microsome reverse mutation assay. The tests were performed both with
and without metabolic activation. A single positive response seen at one dose level in S.
typhimurium TA1537 when tested without metabolic activation was not reproducible. APFO
did not induce mutation in either S. typhimurium or E. coli when tested either with or without
mammalian activation.  APFO did not induce chromosomal aberrations in human lymphocytes
when tested with and without metabolic activation up to cytotoxic concentrations (Murli, 1996c;
NOTOX, 2000).  Sadhu (2002) recently reported that APFO did not induce gene mutation when
tested with or without metabolic activation in the K-l line of Chinese hamster ovary (CHO) cells
in  culture.

Murli (1996b) tested APFO twice for its ability to induce chromosomal aberrations in CHO
cells.  In the first assay, APFO induced both chromosomal aberrations and polyploidy in both the
presence and absence of metabolic activation. In the second assay, no significant increases in
chromosomal aberrations were observed without activation. However, when tested with
metabolic activation, APFO induced significant increases in chromosomal aberrations and in
polyploidy (Murli, 1996b).

APFO was tested in a cell transformation and cytotoxicity assay conducted in C3H 10T,^ mouse
embryo fibroblasts. The cell transformation was determined as both colony transformation and
foci transformation potential. There was no evidence of transformation at any of the dose levels
tested in either the colony or foci assay methods (Garry & Nelson, 1981).

APFO was tested twice in the mouse micronucleus assay. APFO did not induce any significant
increases in micronuclei and was considered negative under the conditions of this  assay (Murli,
1996a).

3.5 Repeat Dose Studies in Animals

3.5.1 Subchronic Studies in Non-Human Primates

Goldenthal (1978b) administered rhesus monkeys (2/sex/group) doses of 0, 3,10, 30 or 100
mg/kg-day APFO by gavage for 90 days. Animals were observed twice daily and body weights
were recorded weekly.  Blood and urine samples were collected once during the control period,
and at 1 and 3 months of the study for hematology, clinical chemistry and urinary sis. Organs
and tissues from animals that were sacrificed at the end of the study and from animals that died
during the treatment period were weighed, examined for gross pathology and processed for
histopathology.

All monkeys in the 100 mg/kg-day group died during the study. The first death occurred during
week 2; all animals were dead by week 5. Signs and symptoms which first appeared during
week 1 included anorexia, frothy  emesis which was sometimes brown in color, pale face and
gums, swollen face and eyes, slight to severe decreased activity, prostration and body trembling.
Three monkeys from the 30 mg/kg-day group died during the study; one male died during week

                                         47

-------
SAR RAVIPW Draft; Hn Nnf Pitp nr Oiinfp
7 and the two females died during weeks 12 and 13. Beginning in week 4, all four animals
showed slight to moderate and sometimes-severe decreased activity.  One monkey had emesis
and ataxia, swollen face, eyes and vulva, as well as pallor of the face and gums. Beginning in
week 6, two monkeys had black stools and one monkey had slight to moderate dehydration and
ptosis of the eyelids.  No monkeys in the 3 or 10 mg/kg-day groups died during the study.  One
monkey in the 10 mg/kg-day group was anorexic during week 4, had a pale and swollen face in
week 7 and had black stools for several days in week 12.  Animals in the 3 mg/kg-day group
occasionally had soft stools or moderate to marked diarrhea; frothy emesis was also occasionally
noted in this group.

Changes in body weight were similar to the controls for animals from the 3 and 10 mg/kg-day
groups.  Monkeys from the 30 and 100 mg/kg-day groups lost body weight after week 1. At the
end of the study, this loss was statistically significant for the one surviving male in the 30
mg/kg-day group (2.30  kg vs 3.78 kg for the control).  The results of the urinalysis, and
hematological and clinical chemistry analyses were comparable for the control and the 3 and 10
mg/kg-day groups at one and three months.

At necropsy, the following changes in absolute and relative organ weight changes were noted:
absolute and relative weight of the hearts in females from the 10 mg/kg-day group were
significantly decreased; absolute brain weight of females from this same group were also
significantly decreased  and relative group mean weight of the pituitary in males from the 3
mg/kg-day group was significantly increased. The biological significance of these weight
changes is difficult to assess, as they were not accompanied by morphologic changes.

In animals that died before the end of the study, one male and two females from the 30 mg/kg-
day group and all animals from the 100 mg/kg-day group had marked diffuse lipid depletion in
the adrenal glands. All  males and females  from the 30 and 100 mg/kg-day groups also had slight
to moderate hypocellularity of the bone marrow and moderate atrophy of lymphoid follicles in
the spleen.  One female from the 30 mg/kg-day group and all animals in the 100 mg/kg-day
group had moderate atrophy of the lymphoid follicles in the lymph nodes.

Only one male in the 30 mg/kg-day group survived until terminal sacrifice, and this male also
had slight to moderate hypocellularity of the bone marrow and moderate atrophy  of lymphoid
follicles in the spleen. No treatment related lesions were seen in the organs of animals from the
3 and 10 mg/kg-day groups.

The levels of PFOA in the serum and liver of these animals are presented below in Table 16.
Individual values are presented so there are double entries for most dose levels.
                                          48

-------
SAR
            Draft; F>n Nnf Pifp nr Onntp
                                       Table 16
            Levels of PFOA in the Serum and Liver of Surviving Rhesus Monkeys
Dose
0
3
3
10
10
30
Serum (ng/ml)
Male
ND
53
48
45
71
145
Female
1
65
50
79
71
Dead
Liver (u£/ml)
Male
0.05
3
ND
9
ND
61
Female
0.07
7
ND
ND
10
Dead
Liver total (ng)
Male
3
250
ND
600
ND
4000
Female
5
350
ND
ND
750
Dead
ND - Not Determined

Under the conditions of this study the LOAEL is 3 mg/kg-day and no NOAEL was established
because effects were observed at the lowest dose.

Thomford (2001 a,b) conducted a range-finding and a 6-month toxicity study in male
cynomolgus monkeys. In the range-finding study, Thomford (2001a) administered male
cynomolgus monkeys an oral capsule containing 0, 2, or 20 mg/kg-day APFO for 4 weeks.
There were 3 monkeys in the 2 and 20 mg/kg-day groups and one monkey in the control group.
The monkeys weighed 2.1 to 3.6 kg at the start of treatment. Animals were observed twice daily
for mortality and moribundity and were examined at least once daily for signs of poor health or
abnormal behavior. Body weights were recorded weekly and food consumption was assessed
qualitatively. The monkeys were fasted overnight and blood samples were collected one week
prior to the start of the study and on day 30 for clinical hematology and clinical chemistry
analyses, and hormone and PFOA level.  Blood for clinical chemistry was also collected from
each animal on day 2 (approximately 24 hours after the first dose).  Samples were analyzed for
estradiol, estrone, estriol, thyroid stimulating hormone, total and free triiodothyronine, and total
and free thyroxin.

At scheduled necropsy, samples of the right lateral lobe of the liver were collected from each
animal and analyzed for palmitoyl CoA oxidase activity. Representative samples of liver, right
and left testes, and pancreas were collected from each animal for cell proliferation evaluation
using proliferation cell nuclear antigen. Bile was collected from each animal for bile acid
determination.  A sample of liver was collected from each animal for PFOA concentration
analysis. The adrenals, liver, pancreas, spleen, and testes from each animal were examined
microscopically, and the remaining tissues were preserved for possible future examination.

All animals survived to scheduled sacrifice. There were no clinical signs  of toxicity in the
treated groups and there was no effect on body weight.  Low or no food consumption was
observed for one animal given 20 mg/kg-day.  There were no effects on estradiol, estriol, thyroid
stimulating hormone, total and free triiodothyronine, and total and free thyroxin. Estrone levels
were notably lower for males given 2 and 20 mg/kg-day APFO.  There was no evidence of
                                          49

-------
SAR RPVIPW Draft; Dn Nnf Cite nr Onntf
peroxisome proliferation or cell proliferation in the liver, testes or pancreas of treated monkeys.
No adverse effects were noted in either gross or clinical pathology studies.

In the 26-week study, male cynomolgus monkeys were administered APFO by oral capsule at
doses of 0, 3,10 or 30 mg/kg-day for 26 weeks (Thomford 2001b; Butenhoff et al., 2002). At
study initiation the monkeys weighed 3.2 to 4.5 kg. There were 4 monkeys in the 3 mg/kg-day
group and 6 monkeys in each of the other groups.  Dosing of animals in the 30 mg/kg-day dose
group was stopped from days 11-21 because of toxicity. When dosing was resumed on day 22,
animals received 20 mg/kg-day and this group was designated the 30/20 mg/kg-day group. At
the end of the 26-week treatment period, 2  animals in the control  and 10 mg/kg-day groups were
observed for a 13-week recovery period.

Animals were observed twice daily for mortality and moribundity and were examined at least
once daily for signs of poor health or abnormal behavior. Ophthalmic examinations were done
before initiation of treatment and during weeks 26 and 40. Body  weights were recorded weekly
and food consumption was assessed qualitatively.  Blood and urine samples were collected for
clinical hematology, clinical chemistry, and urinalysis before the  start of treatment and on days
11, 31, 63, 91,182, 217, 245 and 275. Blood samples were also taken for hormone
determinations;  samples were analyzed for estradiol, estrone, estriol, thyroid stimulating
hormone, total and free triiodothyrpnine, total and free thyroxin, testosterone, and
cholecystokinin (CCK). Blood, urine and feces were collected during week 2 and every 2 weeks
thereafter during treatment and recovery for PFOA concentration analyses.

At scheduled necropsy, liver samples were taken for determination of PFOA levels. The right
lateral lobe of the liver was collected from each animal for palmitoyl CoA oxidase activity
analyses, and representative samples of liver, right and left testes, and pancreas were collected
from each animal for cell proliferation evaluation using proliferation cell nuclear antigen. All
available bile was collected for bile acid determination. Weights  of the adrenal glands,  brain,
epididymis, kidney, liver, pancreas, testis, and thyroid with parathyroid were recorded.  The
following tissues were collected for histopathology:  adrenal (2),  aorta, brain, cecum,  colon,
duodenum, epididymis (2), esophagus, eyes [preserved in Davidson's fixative (2)], femur with
bone marrow (articular surface of the distal end), gallbladder, heart, ileurn, jejunum, kidneys  (2),
lesions, liver, lung, mesenteric lymph node, mammary gland, pancreas, pituitary, prostate,
rectum, salivary gland [mandibular (2)], sciatic nerve, seminal vesicle (2), skeletal muscle
(thigh), skin, spinal cord (cervical, thoracic, and lumbar), spleen,  sternum with bone marrow,
stomach, testis [(2) preserved in Bourn's solution], thymus, thyroid (2) with parathyroid, trachea
and urinary bladder.

Two animals, one male from the 30/20 mg/kg-day dose group and one male from the  3 mg/kg-
day dose group, were sacrificed in moribund condition during the study. The male in the 30/20
mg/kg-day dose group was sacrificed on day 29. This animal exhibited signs of hypoactivity,
weight loss, few or no feces, low or no food consumption and the entire body was cold to the
touch before death. Necropsy revealed esophageal and gastric lesions that were indicative of an
injury that occurred during dosing and liver lesions that were presumed to be treatment related.
The animal from the 3 mg/kg-day dose group was sacrificed on day 137.  This animal showed
clinical signs of limited use and paralysis of the hind limbs, ataxia and hypoactive behavior, few
feces and no food consumption. The cause of death was not determined, but APFO treatment
could not be ruled out.
                                          50

-------
SAB RPVJPW riraft; Hn Nnf fitt* nt-
Males given 30 mg/kg-day from days 1-11 had clinical signs of few feces and low food
consumption and they lost weight during week 1 of treatment.  Based on these signs, treatment
was stopped on day 11 and was not resumed until day 22.  When treatment was resumed, the
dose was lowered to 20 mg/kg-day; this group was then designated the 30/20 mg/kg-day group.
Of the remaining animals in this group, only 2 tolerated this dose level for the remaining 23
weeks of treatment.  Treatment of three males given 30/20 mg/kg-day was halted on days 43
(week 7), 66 (week 10), and 81 (week 12) respectively. Clinical signs in these animals included
thin appearance, few or no feces, low or no food consumption, and weight loss. The animals
appeared to recover from compound-related effects within 3 weeks after cessation of treatment.

Mean body weight changes were notably lower during weeks 1 and 2 for males receiving 30
mg/kg-day.  During week 2, this change was statistically significant.  Treatment was stopped on
day 11  and when it was resumed at 20 mg/kg-day on day 21, mean body weight changes were
significantly lower than controls during weeks 7, 9 and 24. Overall mean body weight changes
through week 27 were notably lower for the males in the 30/20 mg/kg-day group.  There was an
increased incidence of low or no food consumption for animals in  the 30/20 mg/kg-day group
that was considered to be treatment related.

There were no consistent or clearly dose-related effects on estron,  estradiol, estriol, testosterone,
or CCK levels in treated groups over time. Similarly, thyroid stimulating hormone and free and
total thyroxine levels remained relatively constant and were not significantly  altered throughout
the study.  Mean individual values for free and total triiodothyronine levels were statistically
significantly decreased during the dosing period at the 30/20 mg/kg/day dose group compared to
controls, although the biological significance of this decrease is unclear.

At terminal sacrifice  at 26 weeks, there were statistically significant increases in mean absolute
liver weights and mean liver-to-body weight percentages in all dose groups. In addition, there
was a positye dose-response trend towards an increased relative liver-to-body weight,  with
statistical significance at the high dose group of 30/20 mg/kg/day.   The increased liver weights
were considered to be treatment-related. The increased liver weight was thought to be due, in
part, to hepatocellular hypertrophy (as demonstrated by decreased hepatic DNA content) which
in turn may be due to mitochodrial proliferation (as demonstrated by increased succinate
dehydrogenase activity).

Since administration of APFO to rats results in liver, Ley dig cell and  pancreatic acinar cell
tumors, Butenhoff et al  (2002) specifically looked for markers of tumor formation in the
monkeys.  In the liver, there was only a two-fold increase in hepatic palmitoyl CoA oxidase
activity in the 30/20 mg/kg-day group, which is consistent with reports for other species that are
not particularly responsive to PPARa-agonists. Replicative DNA  synthesis in the liver, an
indication of cell proliferation, was not altered in the treated animals.  Similarly, it has been
proposed that changes associated with the pancreatic acinar cell tumors in rats include increased
serum CCK concentrations and indications of cholestasis, including alkaline phosphatase,
bilirubin, and bile acids; none of these changes were noted in the cynomolgus monkeys.  Finally,
in the rat, it has been proposed that the Ley dig cell tumors are due to a sustained increase in
estradiol resulting from aromatase induction. In the treated cynomolgus monkeys there were no
significant changes in estradiol, estriol, or testosterone. In addition, mere was no change in
replicative DNA synthesis in the pancreas or testes.

At the recovery sacrifice, there were no treatment-related effects on terminal body weights or on
absolute or relative organ weights indicating that the liver weight changes seen at terminal

                                           51

-------
SAR Rpvifw Draft? F>n Nnt Tito nr
sacrifice were reversible over time. There were no treatment-related macroscopic or
microscopic changes at the recovery sacrifice.

Serum and liver concentrations of PFOA did not increase in a dose-related manner. This may
have been due to saturation or attaining steady state levels in the first several weeks of the study.
In addition, there was a great deal of variability in the PFOA levels.  This may have been due to
the method of dosing (by capsule), the timing of dosing relative to blood sample collection and
gall bladder emptying, or the analytical method (+ 30% for interassay, intra-assay and system).
Since steady state appeared to have been reached by 4-6 weeks of dosing, the study authors
calculated the mean serum levels for the period following 6 weeks; steady state serum levels
were 77 + 39, 86 + 33 and 158 + 100 jig/ml for the 3,10 and 30/20 mg/kg-day groups,
respectively. In the control animals, 2/3 of the serum samples contained PFOA and averaged a
level of 0.203 + 0.154 jig/ml. There was not a statistically significant difference in the mean
serum levels of the 3 and 10 mg/kg-day groups, but the mean serum level in the 30/20 mg/kg-
day group was significantly higher than the 3 and 10 rng/kg-day groups. At terminal sacrifice,
the levels of PFOA in the liver were similar in the 3 and 10 mg/kg-day groups, and ranged from
6.29 -21.9 ng/g. The two monkeys in the 30/20 mg/kg-day group had liver concentrations of 16
and 83.3 u.g/g.  After the recovery period, the serum and liver PFOA levels in the 10 mg/kg-day
group had returned to baseline.

Under the conditions of the study, the LOAEL was 3 mg/kg-day (increased liver weight and
possibly mortality) and a NOAEL was not established.

3.5.2 Subchronic Studies in Rodents

Christopher and Marisa (1977) administered ChR-CD mice (5/sex/group) dietary concentrations
of 0, 30,100, 300,1000,3000,10,000, or 30,000 ppm of APFO for 28 days. The animals were
observed daily and body weights and food consumption were recorded weekly. At necropsy, the
organs were weighed, examined for gross pathology and preserved for histopathology.  All
animals in the 1000 ppm and higher groups died before the end of day 9. All animals in the 300
ppm group died within 26 days except one male.  One animal in each of the 30 and 100 ppm
groups died prematurely. Clinical signs were observed in mice exposed to 100 ppm and higher
doses of APFO. At 100 ppm some animals exhibited  cyanosis on days 10 and 11 of testing, but
appeared normal throughout the rest of the study.  Animals fed 300 ppm exhibited roughed fur
and muscular weakness as well as signs of cyanosis after 9 days of treatment.  Animals fed 1000
ppm exhibited similar effects after 6 days and those receiving 3000 ppm or greater doses
exhibited effects after 4 days. There was a dose-related reduction in mean body weight in all
treated groups. Relative and absolute liver weights were increased in mice fed 30 ppm or more
APFO. Treatment-related changes were observed in the livers among all APFO treated animals
including enlargement and/or discoloration of 1 or more liver lobes.  Histopathologic
examination of all surviving treated mice revealed diffuse cytoplasmic enlargement of
hepatocytes throughout the liver (panlobular hypertrophy) accompanied by focal to multifocal
cytoplasmic lipid vacuoles of variable size which were random in distribution.

Metrick and Marias (1977) administered ChR-CD rats (5/sex/group) dietary concentrations of 0,
30,100, 300,1000, 3000,10,000, or 30,000 ppm APFO for 28 days. The animals were observed
daily and body weights and food consumption were recorded weekly. At necropsy, the organs
were weighed, examined for gross pathology and preserved for histopathology. All animals in
the  10,000 and 30,000 ppm groups died before the end of the first week. Gross pathologic
examination revealed white foci in the cortex and medulla in the kidneys of a 10,000 ppm


                                          52

-------
SAR RPVIPW rtraft? Hn Not Pifp nr
female.  Pelvic dilation was evident in the kidneys of a control male and a 30,000 ppm female.
There were no premature deaths or unusual behavioral reactions in the other groups. Food
consumption was reduced in the 1000 and 3000 ppm groups. Body weight gain was reduced as
dose increased.  The reduction in body weight gain was statistically significant for males at 1000
ppm and males and females at 3000 ppm.  Absolute liver weights were increased in males fed 30
ppm or more and females fed 300 ppm or more. Treatment-related morphologic changes were
observed in the livers of all test animals.  Focal to multifocal cytoplasmic enlargement of
hepatocytes in the centrilobular to midzonal areas was noted in animals fed 30 to 300 ppm, and
multifpcal to diffuse enlargement of hepatocytes throughout the liver lobules (panlobular) was
noted in animals fed 1000 ppm or higher.  The hypertrophy of hepatocytes was accompanied by
acidophilic degeneration and/or necrosis of scattered liver cells with no lobular distribution. The
severity and degree of tissue involvement were more pronounced in males than in females.

Goldenthal (1978a) administered ChR-CD rats (5/sex/group) dietary levels of 0,10, 30,100,
300, and 1000 ppm APFO for 90 days. These dose levels are equivalent to 0.056,1.72, 5.64,
17.9, and 63.5 mg/kg-day in males, and 0.74, 2.3, 7.7, 22.36 and 76.47 mg/kg-day in females.
Animals were observed twice daily and body weight and food consumption were recorded
weekly. Blood and urine samples were collected during the pretest period and at 1 and 3 months
of the study for hematology and clinical chemistry and urinary sis. At necropsy, the organs from
the control, 100, 300, and 1000 ppm groups were weighed and examined for histopathologic
lesions; livers from the 10 and 30 ppm groups were also examined microscopically.

There were no treatment-related changes in behavior or appearance. One female in the 100 and
one female in the 300 ppm group died during collection of blood. These deaths were not
considered to be treatment related.  All other animals survived until scheduled sacrifice. There
was  a decrease in body weight gain for male rats at the 300 and 1000 ppm dose level.  At 13
weeks, mean body weight of males in the 1000 ppm group was significantly less than that of
controls. There were no treatment related effects on the hematologic, biochemical or urine
parameters.

Relative kidney weights were significantly increased in males in the 100,300, and 1000 ppm
groups.  However, absolute kidney weights were comparable among groups, and there were no
histopathological lesions.  Absolute liver weights were significantly increased in males in the 30,
300 and 1000 ppm groups and in females in the 1000 ppm group.  Relative liver weights were
significantly increased in males in the 300 and 1000 ppm groups and in females in the 1000 ppm
group. Discoloration on the surface of the liver was observed in male rats in the 1000 ppm
group. Hepatocellular hypertrophy (focal to multifocal in the centrilobular to midzonal regions)
was  observed in 4/5, 5/5, and 5/5 males in the 100, 300, and 1000 ppm groups, respectively.
Hepatocyte necrosis was observed in 2/5, 2/5,1/5, and 2/5 males in the 30,100, 300, and 1000
ppm groups, respectively.  Under the conditions of this study, the LOAEL for males is 30 ppm
(1.72 mg/kg-day) based on liver effects and the NOAEL is 10 ppm (0.56 mg/kg-day); the
LOAEL for females is 1000 ppm (76.5 mg/kg-day) and the NOAEL is 300 ppm (22.4 mg/kg-
day).

Palazzolo (1993) administered male ChR-CD rats (45-55 per group) dietary concentrations of 1,
10, 30, or 100 ppm APFO for 13  weeks. These doses are equivalent to 0.06, 0.64,1.94, and 6.50
mg/kg-day. Two control groups (a nonpair-fed control group and a control group pair-fed to the
100 ppm dose group) were fed basal diet during that period. Following the 13-week exposure
period, 10 animals per group were fed basal diet for an 8-week recovery period.  The animals
were observed twice daily for clinical signs of toxicity,  and body weights and food consumption

                                          53

-------
SAR RPVIPW Draft; Tin Nnt rifp nr Oiinfp
were recorded weekly. Food consumption was recorded daily for the pair-fed animals. A total
of 15 animals per group were sacrificed following 4, 7, or 13 weeks of treatment; 10 animals per
group were sacrificed after 13 weeks of treatment and following the 8 week recovery period.
Serum samples collected from 10 animals per group at each scheduled sacrifice during treatment
and from 5 animals per group during recovery  were analyzed for estradiol, total testosterone,
luteinizing hormone, and PFOA. The level of palmitoyl  Co A oxidase was analyzed from a
section of liver that was obtained from 5 animals per group at each scheduled sacrifice. Weights
of the brain, liver, lungs, testis, seminal vesicle, prostate, coagulating gland, and urethra were
recorded, and these tissues were examined histologically. In addition, the brain, liver, lungs,
testis, seminal vesicle, and prostate were preserved in glutaraldehyde for electron microscopic
examination.

In the analysis of the data, animals in groups exposed to  1,10, 30, and 100 ppm APFO were
compared to the control animals in the nonpair-fed group, while the data from the pair-fed
control animals were compared to animals exposed to 100 ppm APFO. No treatment-related
clinical signs were noted.  At 100 ppm, significant reductions in body weights were seen
compared to the pair-fed control group during  week 1 and the nonpair-fed control group during
weeks 1-13.  Body weight data in the other dosed-groups were comparable to controls. At 100
ppm, mean body weight gains were significantly higher than the pair-fed control group during
week 1 and significantly lower than the nonpair-fed control group during weeks  1-13. At 10 and
30 ppm, mean body weight gains were significantly lower than the nonpair-fed control group at
week 2.  These differences in body weight and body weight gains were not observed during the
recovery period.  Animals fed 100 ppm consumed significantly less food during weeks 1 and 2,
when compared to the nonpair-fed control group. Overall, there was no significant difference in
food consumption.  There were no significant differences among the groups for any of the
hormones evaluated in the serum although there appeared to be some indication of elevated
estradiol for the 100 ppm group at week 5.

Significant increases in absolute and relative liver weights and hepatocellular hypertrophy were
observed at weeks 4, 7, or 13 in the 10, 30 and 100 ppm groups. There was no evidence of any
degenerative changes or abnormalities associated with the hypertrophy. Hepatic palmitoyl CoA
oxidase  activity was significantly increased at  weeks 4,7, and 13 in the 30 and 100 ppm groups.
At 10 ppm, hepatic palmitoyl CoA oxidase activity was significantly increased at week 4 only.
During recovery, however, none of the liver effects were observed, indicating that these
treatment-related liver effects were reversible.

Serum levels from the male rats in this study are shown below in Table 17.
                                          54

-------
SAR RPVJPW
                   f>n Nnt rite nr Oimfp
                                       Table 17
       Summary of Serum Values (jig/ml) in Male Rats Treated with APFO for 90 Days
Week
5
8
14
22
Dose Group (mg/kg-day)
0.06
6.5 ±1.05"
(8)
7.5 ±1.27
(9)
7.1 ±1.15
(10)
1.2
(1)
0.64
55.4 ±8.08
(9)
45.5 ± 16.20
(10)
41.2 ±12.98
(10)
1.09 ±1.303
(3)
1.94
103.5 ± 14.38
(8)
87.3 ± 27.48
(10)
70.3 ±16.18
(10)
1.64 ±0.918
(3)
6.50
159.3 ±30.16
(10)
149.1 ±34.98
(10)
137.6 ±33.83
(10)
2.45 ± 0.849
(2)
a - Mean ± SD, (N)

Under the conditions of this study, the LOAEL is 10 ppm (0.64 mg/kg-day) based on increases
in absolute and relative liver weights with hepatocellular hypertrophy. The NOAEL is 1.0 ppm
(0.06 mg/kg-day).

3.5.3 Chronic Toxicity and Carcinogenicity Studies in Rats

The chronic toxicity and carcinogenic potential of PFOAhas been investigated in two dietary
studies in rats. In the first study (Sibinski, 1987), groups of 50 male and 50 female Sprague-
Dawley (Crl:CD BR) rats were fed diets containing 0, 30 or 300 ppm APFO for two years.
Groups of 15  additional rats per sex were fed 0 or 300 ppm APFO and evaluated at the one year
interim sacrifice.  In males, the mean test article consumption was 1.3 and 14.2 mg/kg-day for
the 30 and 300 ppm groups, respectively; in females, the mean test article consumption was  1.6
and 16.1 mg/kg-day for the 30 and 300 ppm groups, respectively.  All animals were observed
daily throughout the two year dosing period.  Body weights and feed consumption were recorded
once per week for the first six months, and then once every two weeks for the remainder of the
study.  Clinical pathological examinations including hematology, serum chemistry and urinalysis
were conducted on samples obtained from 15 rats per sex from each group  at 3, 6,12,18 and 24
months. Macroscopic postmortem examinations were performed on all animals that died during
the study and those which were terminated at the one year interim and two  year necropsies.  The
weights of the kidneys, liver, testes, brain, heart, spleen, adrenal glands and uterus were recorded
for 15 randomly selected rats/sex at the interim termination from both the control and high-dose
groups, and from the control and both treated groups at the two year necropsy. Microscopic
evaluation was performed on all tissues from all of the control and high-dose rats.

There was a dose-related decrease in body weight gains in the males rats and to a lesser extent,
in the female rats as compared to the controls; the decreases were statistically significant in the
high-dose groups of both sexes (up to 21% in male rats and 11% in female  rats). The body
weight changes are treatment related since feed consumption was actually increased (rather than
decreased). There were no differences in mortality between the treated and untreated groups; the
                                          55

-------
SAR RPVJPW Draff; Dn Nnf rite nr
survival rates at the end of 104 weeks for the male control, low-, and high-dose groups were
70%, 72% and 88%, respectively; in females, the survival rates were 50%, 48% and 58% for the
control, 30 and 300 ppm groups, respectively.  The only clinical sign observed was a dose-
related increase in ataxia in the female rats; which was most commonly associated with
moribund animals; the incidences in the control, low- and high-dose groups were 4%, 18% and
30%. Significant decreases in red blood cell counts, hemoglobin concentrations and hematocrit
values were observed in the high-dose male and female rats as compared to control values.
Clinical chemistry changes included slight (less than 2-fold), but significant increases (P< 0.05)
in alanine aminotransferase (ALT), aspartate aminotransferase (AST) and alkaline phosphatase
(AP) in both treated male groups from 3-18 months, but only in the high-dose males at 24
months.  Slight (up to about 10%) increases in absolute or relative liver and kidney weights were
noted in both high-dose male and female rats at the 1 year interim sacrifice and at the terminal
necropsy; however, only the relative liver weight (vs. body weight or brain weight) increases in
the high-dose males were statistically significant (PO.05).

Histologic evaluations showed lesions in the liver, testis and ovary. In the liver, the increased
incidence of lesions reached statistical significance only in the high-dose male group.  At the 1
year interim sacrifice, diffuse hepatomegalocytosis (12/15 animals), portal mononuclear cell
infiltration (13/15 animals) and hepatocellular necrosis (6/15 animals) were seen in the high-
dose males while incidences in the control group were, 0/15, 7/15 and 0/15, respectively.
Hepatocellular vacuolation was seen in 11/15 high-dose females as compared to an incidence of
5/15 in the control group. At the 2-year sacrifice, megalocytosis was found at an incidence of
0%,  12% and 80% in the males, and 0%, 2% and 16% in the females from the control, low-, and
high-dose groups, respectively. Hepatic cystoid degeneration, a condition characterized by areas
of multilocular microcysts in the liver parenchyma, was observed in 14% and 56% of the low-
and high-dose males, as compared to a control incidence of 8%. The incidence of hyperplastic
nodules, a localized proliferation of hepatic parenchyma! cells, was slightly increased in the
high-dosed males with an incidence of 6% as compared to 0% in the control males.

At the one-year sacrifice, testicular masses were found in 6/50 high-dose and 1/50 low-dose rats,
but not in any of the controls.  Furthermore, marked aspermatogenesis was found in 2/15 high-
dose males but none in the control males. At the 2-year sacrifice, vascular mineralization of the
testes occurred in 18% of the high-dosed male and 6% of the low-dosed males, but was not seen
in the controls.  These testicular effects reached statistical significance in the high-dose group.

A statistically significant, dose-related increase in the incidence of ovarian tubular hyperplasia
was found in female rats at the 2-year sacrifice. The incidence of this lesion in the control, low-,
and high-dose groups was 0%, 14%, and 32%, respectively. The biological significance of this
effect at the time of the initial evaluation was unknown, as there was no evidence of progression
to tumors. Recently, however, slides of the ovaries from that study were re-evaluated, with
particular emphasis placed on the proliferative lesions of the ovary (Mann and Frame, 2004).
Using more recently published nomenclature, the ovarian lesions were diagnosed and graded as
gonadal stromal hyperplasia and/or adenomas, which corresponded to the diagnoses of tubular
hyperplasia or tubular adenoma by the original study pathologist.  The data are summarized in
Table 18.  No statistically significant increases in hyperplasia (total number), adenomas, or
hyperplasia/adenoma combined were seen in treated groups compared to controls. There was
some evidence of an increase in size of stromal lesions observed at the 300 ppm group; however,
adenomas occurred in greater incidences in the control group than in either of the treated groups.
Results of this follow-up evaluation indicated that rats sacrificed at the one-year interim
                                           56

-------
SAR RPVJPW Draff; Tin Nnt fite nr Oiintp
sacrifice, as well as rats that died prior to the interim sacrifice were not considered at risk for
tumor development.

                                 Table 18
             Incidence of Ovarian Stromal Hyperplasia and Adenoma in Rats
Group
No. examined
Hyperplasia (Total)
Grade 1
Grade 2
Grade 3
Grade 4
Adenoma
Adenoma and/or
Hyperplasia
Control
45
8
6
2
0
0
4
12
30ppm
47
16
7
3
5
1
0
16
300 ppm
46
15
5
1
6
3
2
17
From Mann and Frame (2004)

Based on these toxic effects, the high dose selected in this study appears to have reached the
Maximum Tolerated Dose.  Based on a decrease in body weight gain, increase in liver and
kidney weights and toxicity in the hematological and hepatic systems, the LOAEL for male rats
is 300 ppm and the NOAEL is 30 ppm. The LOAEL for female rats is 300 ppm based on a
decrease in body weight gain and hematologic effects and the NOAEL is 30 ppm.

At the termination of the study, there was a significant increase (P<0.05) in the incidence of
testicular (Leydig) cell adenomas in the high-dose male rats. The incidence of the Leydig cell
tumors (LCT) in the control, low- and high-dose males was 0/50 (0%), 2/50 (4%) and 7/50
(14%), respectively. The increase was also statistically significant when compared to the
historical control incidence of 0.82% observed in 1,340 Sprague-Dawley control male rats used
in 17 carcinogenicity studies (Chandra et al., 1992). The spontaneous incidence of LCT in 2-
year old Sprague-Dawley rats in other studies was reported to be approximately 5% (cited in
Cleggetal., 1997).

There was also a significant increase (P<0.05) in the incidence of mammary fibroadenomas in
both groups of female rats.  The incidence of the mammary fibroadenoma was 10/47 (21%),
19/47 (40%) and 21/49 (43%) in the control, 30, and 300 ppm groups, respectively. The increase
was also statistically significant when compared to the historical control incidence of 19.0%
observed in 1,329 Sprague-Dawley control female rats used in 17 carcinogenicity studies
(Chandra et al., 1992).  The investigators did not consider the mammary fibroadenomas to be
treatment related on the basis of the historical control incidence (24%)  from a study of 181
female rats terminally sacrificed at 18 months, which is considered an inappropriate historical
reference. When the mammary fibroadenoma incidences were compared to the historical control
                                         57

-------
SAB RPVIPW Draft; Dn Nnt fiti> nr
incidence (37%) in 947 female rats in the Haskell Laboratory, however, there does not appear to
be any compound related effect (Sykes,1987).

The induction of Ley dig cell tumors was confirmed in a follow-up 2-year mechanistic study of
PFOA toxicity in male Sprague-Dawley rats at a dietary level of 300 ppm (Cook et aL, 1994;
Biegel et al., 2001). There were 156 animals in the treatment group and 80 animals in the
control group. Cage-side examinations were conducted at least once daily throughout the study.
Rats were weighed once a week during the first 3 months and once every other week for the
remainder of the study.  Rats were euthanized at interim time points of 1, 3, 6, 9,12,15,18, and
21 months.  At each time point, the liver and testis from 6 rats/group were weighed and
evaluated for cell proliferation. Another 6 rats/group were selected for peroxisome proliferation,
and 10 rats/group for serum hormone (estradiol, testosterone, LH, FSH, and prolactin) analysis.
All rats surviving the 24-month test period were necropsied for microscopic examination of
various organs: e.g., kidneys, liver, testes, brain, heart, spleen.

In the treated group, relative liver weights and hepatic p-oxidation activity were statistically
significantly increased at all of the sampling time points when compared to the controls.
Absolute testis weights  were increased only at 24 months.  No hepatic or Ley dig cell
proliferation was observed at any of sampling times. There were no significant differences in
serum testosterone, FSH, LH, or prolactin in the PFOA-treated rats when compared to the
controls.  There were, however, significant increases in serum estradiol concentrations in the
treated rats  at 1, 3, 6, 9, and 12 months.

There was a significant increase in the incidence of LCT in the treated rats (8/76,11 %) as
compared to the controls (0/80, 0%).  In addition, the treated group had a significant increase in
the incidences of liver adenomas and pancreatic acinar cell tumors (PACT).  The incidences of
liver adenomas in the control and treated groups were 2/80 (3%) and 10/76 (13%), respectively,
whereas those for the pancreatic acinar cell adenomas were 0/80 (0%) and 7/76 (9%). There was
one pancreatic acinar cell carcinoma in 76 of the treated rats and none in 80 controls. The
incidence of combined pancreatic acinar cell adenoma/carcinoma in the treated rats (8/76,11%)
was significantly increased as compared with the controls (0/80, 0%).

In the first carcinogenicity study (Sibinski, 1987), there was no reported increase in the
incidence of PACT, and the incidence of pancreatic acinar hyperplasia in the male rats was  0/33,
2/34, and 1/43 in the control, 30 and 300 ppm groups, respectively. To resolve this discrepancy,
the histological slides from both studies were reviewed by independent pathologists.  This
review of the microscopic lesions of the pancreas in the two studies indicate that PFOA
produced increased incidences of proliferative acinar cell lesions of the pancreas in the rats of
both studies at the dietary concentration of 300 ppm. The differences observed were quantitative
rather than  qualitiative; more and larger focal proliferative acinar cell lesions  and greater
tendency for progression of lesions to adenoma of the pancreas were observed in the second
study compared to the first study. The difference between pancreatic acinar hyperplasia
(reported in Sibinski, 1987) and adenoma (reported in Cook et al., 1994; Biegel et al. 2001) in
the rat is a reflection of arbitrary diagnostic criteria and nomenclature by different pathologists.
The basis for the quantitative difference in the lesions observed in not known but was believed to
be due most likely to difference in the diets used in the two laboratories (Frame and McConnell,
2003).

In summary, the two carcinogenicity studies of PFOA have shown that PFOA induced liver
adenomas, Leydig cell adenomas, and pancreatic acinar cell tumors in male Sprague-Dawley


                                           58

-------
SAR RPVJPW Draft; Tin Nnt fifp nr
rats. The evidence for mammary fibroadenomas in the female rats is equivocal since the
incidences were comparable to some historical background incidences. PFOA has also been
shown to promote liver carcinogenesis in rodents (Abdellatif et al.,1991; Nilsson et al., 1991).

3.6 Immunotoxicity Studies in Mice

Four immunotoxicity studies of PFOA have been conducted in mice. In the first study, Yang et
al. (2000) administered 0.02 % PFOA to male C57B1/6 mice in the diet for 2,5,1, or 10 days.
At the end of the feeding period, mice were sacrificed and the liver, spleen, and thymus were
weighed. The effect of PFOA administration on the cellularity, cell surface phenotype, and cell
cycle of thymocytes and splenocytes was determined. In addition, effects of exposure of
thymocytes and splenocytes to  PFOA in vitro were examined. Administration of 0.02% PFOA
for 2, 5, 7, or 10 days resulted in a significant increase, relative to controls, in liver weight, even
at the earliest time point.  Also, a decrease in body weight was observed.  Following five days of
administration, significant decreases in thymus and spleen weight were noted. After
administration of 0.02% PFOA for 7 days, significant decreases (85% and 80%, respectively) in
the total number of thymocytes and splenocytes were observed. In addition, the number of
thymocytes expressing both CD4 and CDS decreased by 95%; the number expressing both CD4
and CDS decreased by 57%; and the number expressing either CD4 or CDS decreased by 64%
and 72%, respectively. For the splenocytes, both T cells (CDS) and B cells (CD 19) decreased by
75% and 86%, respectively. Also, significant decreases in both CD4 helper and CDS cytotoxic
splenic T cells were observed.  Upon administration of 0.02% PFOA to mice for 7 days,
thymocyte proliferation was also inhibited, as detected by cell cycle flow cytometry analyses. In
vitro studies showed that there was spontaneous apoptosis occurring in splenocytes and
thymocytes after 8 or 24 hours  of culturing in the presence of varying concentrations (50,100, or
200 M) of PFOA. However, PFOA did not significantly alter the cell cycle under these
conditions.

In order to examine the dose dependency of the effects, Yang et al. (2001) administered C57B1/6
mice diets consisting of 0.001%-0.05% PFOA (w/w) for 10 days.  For examining the time-
course, a diet containing 0.02% PFOA was given for 2, 5, 7 or 10 days. Effects of withdrawal of
PFOA were also studied. The results showed that, at higher doses, a significant decrease,
relative to controls, in body weight was observed, although no other apparent signs of toxicity
such as sores, lethargy, and poor grooming were noticed. However, a significant decrease in
total water intake was observed. Mice receiving dietary PFOA for 10 days experienced
significant increases in liver weight and peroxisome proliferation,  as measured by induction of
acyl-CoA oxidase with lauroyl-CoA or palmitoyl-CoA as substrate. These increases started at
the lowest dose and reached their maximal values at a dose of 0.003-0.01%. In contrast, the
weight decreases of the spleen  and thymus began at a higher dose (0.01%) with no maximum
reached with the doses given.  The time course studies showed that increased liver weights and
peroxisome proliferation were evident at the earliest time point examined. In contrast,
significant thymus and spleen weight decreases required PFOA administration for a period of at
least 5 days, following which the spleen weight remained constant while the thymus weight
continued to decrease. However, upon prolonged treatment for one month, no further decreases
in thymus and spleen weights were observed. In another set of experiments, animals received
0.02% PFOA for 7 days, and then they received normal chow for a period of 10 days. These
recovery experiments showed that the animals rapidly recovered the body weight the second day
after withdrawal of PFOA.  However, the liver weight did not return to normal even after 10
days of recovery. Thymus recovery  started on day 2 and was completed by day 10. The spleen
weights returned to normal by day 2 post-withdrawal. In addition, the changes in thymus and


                                          59

-------
SAR Rpyjpw Draff; Dn Nnt rite nr Ounfp
spleen weight upon PFOA treatment and withdrawal paralleled the changes in total thymocyte
and splenocyte counts. Furthermore, flow cytometry cell cycle experiments showed that the
decrease in thymocyte number caused by PFOA treatment is due mainly to inhibition of
thymocyte proliferation. In contrast, PFOA treatment caused no changes in the cell cycle of
splenocytes.

A third feeding study (Yang et al. 2002a) was designed to examine the possible involvement of
the peroxisome prohferator-activated receptor alpha (PPARa) in the immunomodulation exerted
by PFOA. This study made use of transgenic PPARa null mice, which are homozygous with
regards to a functional mutation in the PPARa gene.  These mice do not exhibit peroxisome
proliferation or hepatomegaly and hepatocarcinogenesis even after exposure to peroxisome
proliferators.  These mice were fed a diet consisting of 0.02% PFOA (w/w) for 7 days. At the
end of the feeding period, mice were sacrificed and the  liver, spleen, and thymus were removed
and weighed. The effect of PFOA on peroxisome proliferation, cell cycle, and
lymphoproliferation was ascertained.

The results showed that, in contrast to wild-type mice, feeding PPARa null mice PFOA resulted
in no significant decrease in body weight.  As expected, peroxisome proliferation, as measured
by fatty acid oxidation, was totally lacking in PPARa null mice. Also in contrast to wild type
mice,  feeding PPARa null mice PFOA resulted in no significant decrease in the weight of the
spleen or the number of splenocytes. At the same time, the decrease in weight and cellularity of
the thymus was attenuated, but not totally eliminated in the PPARa null mice. In addition, the
decreases in the size of the CD4+CD8+ population of thymus cells and the number of thymus
cells in the S and G2/M phases of the cell cycle, which reflects inhibition of proliferation,
observed in wild type mice administered PFOA were much less extensive in PPARa null mice.
Finally, in contrast to wild type mice, PFOA treatment caused no significant change in
splenocyte proliferation in response to mitogens in PPARa null mice.

A fourth feeding study (Yang et al. 2002b) was designed to examine the effects of PFOA on
specific humoral immune responses in mice.  For this study, 0.02 % PFOA was administered to
male C57B1/6 mice for 10 days. Then the animals were examined, via plaque forming cell
(PFC) and serum antibody assays, for their ability to generate an immune response to horse red
blood cells (HRBCs).  Ex vivo and in vitro splenic lymphocyte proliferation assays were also
performed. The results showed that mice fed normal chow responded to challenge with HRBCs
with a strong humoral response, as measured by the PFC assay. In contrast, mice fed with PFOA
responded to HRBC immunization with no increase in HRBC-specific PFCs, relative to
unimmunized controls. However, in experiments where PFOA-treated mice received normal
chow  following HRBC immunization, there was a significant recovery of the numbers of
specific PFCs stimulated. The suppression of the humoral immune response by PFOA was
confirmed by analysis of the serum anti-HRBC response. In  ex vivo experiments, splenocytes
isolated from control mice responded to both ConA and LPS with lymphocyte proliferation, as
measured by thymidine incorporation. However, treating mice with PFOA (0.02% for 7 days)
attenuated the proliferation. In a set of in vitro experiments, PFOA (1- 200  M) added to the
culture medium of splenocytes cultured from untreated mice did not cause an alteration of
lymphocyte proliferation in response to LPS or ConA.

3.7 Prenatal Developmental Toxicity Studies in Animals

Several prenatal developmental toxicity studies of APFO have been conducted. These include
two oral studies in rats, one oral study in rabbits, and one inhalation study in rats.


                                         60

-------
SAR RPVIPW Draft; fin Not Pifp nr Dnnfp
Gortner (1981) administered time-mated Sprague-Dawley rats (22 per group) doses of 0, 0.05,
1.5, 5, and 150 mg/kg-day APFO in distilled water by gavage on gestation days (GD) 6-15.
Doses were adjusted according to body weight. Dams were monitored on GD 3-20 for clinical
signs of toxicity. Individual body weights were recorded on GD 3, 6, 9,12,15, and 20. Animals
were sacrificed on GD 20 by cervical dislocation and the ovaries, uteri, and contents were
examined for the number of corpora lutea, number of viable and non-viable fetuses, number of
resorption sites, and number of implantation sites.  Fetuses were weighed and sexed and
subjected to external gross necropsy. Approximately one-third of the fetuses were fixed in
Bouin's solution and examined for visceral abnormalities by free-hand sectioning.  The
remaining fetuses were subjected to skeletal examination using alizarin red.

Signs of maternal toxicity consisted of statistically significant reductions in mean maternal body
weights on GD 9, 12, and 15 at the high-dose group of 150 mg/kg-day. Mean maternal body
weight on GD 20 continued to remain lower than controls, although the difference was not
statistically significant. Other signs of maternal toxicity that occurred only at the high-dose
group included ataxia and death in three rat dams.  No other effects were reported.
Administration of APFO during gestation did not appear to affect the ovaries  or reproductive
tract of the dams. Under the conditions of the study, a NOAEL of 5 mg/kg-day and a LOAEL of
150 mg/kg-day for maternal toxicity were indicated.

A significantly higher incidence in fetuses with one missing sternebrae was observed at the high-
dose group of 150 mg/kg-day; however this skeletal variation also occurred in the controls and
the other three dose groups (at similar incidence but lower than the high-dose group) and
therefore was not considered to be treatment-related. No significant differences between treated
and control groups were noted for other developmental parameters that included the mean
number of males and females, total and dead fetuses, the mean number of resorption sites,
implantation sites, corpora lutea and mean fetus weights. Likewise, a fetal lens finding initially
described as a variety of abnormal morphological changes localized to the area of the embryonal
nucleus, was later determined to be an artifact of the free-hand sectioning technique and
therefore not considered to be treatment-related. Under the conditions of the study, a NOAEL
for developmental toxicity of 150 mg/kg-day (highest dose group) was indicated.

A second oral prenatal developmental toxicity study was conducted in rabbits (Gortner, 1982).
Based on the results of a range-finding study, an upper dose level of 50 mg/kg-day was set for
the definitive study in which four groups of 18 pregnant New Zealand White rabbits were
administered 0,1.5, 5, and 50 mg/kg-day APFO in distilled water by gavage on gestation days 6-
18. Pregnancy was established in each sexually mature female by i.y. injection of pituitary
lutenizing hormone in order to induce ovulation? followed by artificial insemination with 0.5 ml
of pooled semen collected from male rabbits; the day of insemination was designated as day 0 of
gestation. A constant dose volume of 1 ml/kg was administered. Individual body weights were
measured on GD 3, 6, 9,12,15,18, and 29. The does were observed daily on GD 3-29 for
abnormal clinical signs.  On GD  29, the does were euthanized and the ovaries, uterus and
contents examined for the number of corpora lutea, live and dead fetuses, resorptions and
implantation sites. Fetuses were examined for gross abnormalities and placed in a 37° C
incubator for a 24-hour survival check. Pups were subsequently euthanized and examined for
visceral and skeletal abnormalities.

Signs of maternal toxicity consisted of statistically significant transient reductions in body
weight gain on GD 6-9 when compared to controls; body weight gains returned to control levels


                                           61

-------
SAR RPVJPW Draff- Dn Nnf Pitp ni- Ountp
on GD 12-29. Administration of APFO during gestation did not appear to affect the ovaries or
reproductive tract contents of the does. No clinical or other treatment-related signs were
reported. Under the conditions of the study, aNOAEL of 50 mg/kg-day, the highest dose tested,
for maternal toxicity was indicated.

No significant differences were noted between controls and treated groups for the number of
male and female fetuses, dead or live fetuses, or fetal weights.. Likewise, there were no
significant differences reported for the number of resorption and implantation sites,  corpora
lutea, the conception incidence, abortion rate, or the 24-hour mortality incidence of the fetuses.
Gross necropsy  and skeletal/visceral examinations were unremarkable. The only sign of
developmental toxicity consisted of a dose-related increase in a skeletal variation, extra ribs or
13th rib, with statistical significance at the high-dose group (38% at 50 mg/kg-day, 30% at 5
mg/kg-day, 20% at 1.5 mg/kg-day, and 16 % at 0 mg/kg-day).  A statistically significant increase
in 13™ ribs-spurred occurred in the mid-dose group of 5 mg/kg-day; however, the biological
significance of this effect is uncertain since in both the high- and low-dose groups, this effect
occurred at the same rate and was not statistically significantly different from controls.
Therefore, under the conditions of the study, a LOAEL for developmental toxicity of 50 mg/kg-
day (highest dose group) was indicated.

Staples et al. (1984) also conducted a developmental toxicity study of APFO in rats. The study
design consisted of an inhalation and an oral portion, each with two trials or experiments. In the
first trial the dams were sacrificed on GD 21; while in the second trial, the dams were allowed to
litter and the pups were sacrificed on day 35-post partum. For the inhalation portion of the
study, the two trials consisted of 12 pregnant Sprague-Dawley rats per group exposed to 0, 0.1,
1,10, and 25 mg/m3 APFO by whole-body dust inhalation for 6 hours/day, on GD 6-15. In the
oral portion of the study, 25 and 12 Sprague-Dawley rats for the first and second trials,
respectively, were administered 0 and 100 mg/kg-day APFO in corn oil by gavage on GD 6-15.
For both routes of administration, females were mated on an as-needed basis and when the
number of mated females was bred, they were ranked within breeding days by body weight and
assigned to groups by rotation in order of rank. Finally, two additional groups (six dams per
group) were added to each trial that was pair-fed to the 10 and 25 mg/m groups.

For trial one, the dams were weighed on GD 1, 6, 9,13,16, and 21 and observed daily for
abnormal clinical signs.  On GD 21, the dams were sacrificed by cervical dislocation and
examined for any gross abnormalities, liver weights were recorded and the reproductive status of
each animal was evaluated. The ovaries, uterus and contents were examined for the number of
corpora lutea, live and dead fetuses, resorptions and implantation sites. Pups (live and dead)
were counted, weighed and sexed and examined for external, visceral, and skeletal alterations.
The heads of all control and high-dosed group fetuses were examined for visceral alterations as
well as macro- and microscopic evaluation of the eyes.

For trial two, in which the dams were allowed to litter, the procedure was the same as that for
trial one up to GD 21.  Two days before the expected day of parturition, each dam was housed in
an individual  cage. The date of parturition was noted and designated Day 1  PP.  Dams were
weighed and examined for clinical signs on Days 1, 7. 14, and 22 PP.  On Day 23 PP all dams
were sacrificed. Pups were counted, weighed, and examined for external alterations.  Each pup
was subsequently weighed and inspected for adverse clinical signs on Days 4, 7,14, and 22 PP.
The eyes of the pups were also examined on Days 15 and 17 PP for the inhalation portion and on
Days 27 and 31 PP for the gavage portion of the study. Pups were sacrificed on Day 35 PP and
examined for visceral and skeletal alterations.
                                          62

-------
SAR RPVIPW Tlraft; Dn Nnt fitp nr
In trial one of the inhalation study, treatment-related clinical signs of maternal toxicity occurred
at 10 and 25 mg/m3 and consisted of wet abdomens, chromodacryorrhea, chromorhinorrhea, a
general unkempt appearance, and lethargy in four dams at the end of the exposure period (high-
concentration group only).  Three out of 12 dams died during treatment at 25 mg/m3 (on GD 12,
13, and 17). Food consumption was significantly reduced at both 10 and 25 mg/m3; however, no
significant differences were noted between treated and pair-fed groups. Significant reductions in
body weight were also observed at these concentrations, with statistical significance at the high-
concentration only. Likewise, statistically significant increases in mean liver weights (p < 0.05)
were seen at the high-concentration group.  Under the conditions of the study, a NOAEL and
LOAEL for maternal toxicity of 1 and 10 mg/m3, respectively, were indicated.

No effects were observed on the maintenance of pregnancy or the incidence of resorptions.
Mean fetal body weights were significantly decreased in the 25 mg/m3 groups (p = 0.002) and in
the control group pair-fed 25 mg/m3 (p = 0.001). However, interpretation of the decreased fetal
body weight is difficult given the high incidence of mortality in rne dams. Under the conditions
of the study, a NOAEL and LOAEL for developmental toxicity of 10 and 25 mg/m3,
respectively, were indicated.

In trial two of the inhalation study, clinical signs of maternal toxicity seen at 10 and 25 mg/m3
were similar in type and incidence to those described for trial one. Maternal body weight gain
during treatment at 25 mg/m3 was less than controls, although the difference was not statistically
significant. In addition, 2 out of 12 dams died during treatment at 25 mg/m3.  No other
treatment-related effects were reported, nor were any adverse effects noted for any of the
measurements of reproductive performance. Under the conditions of the study, a NOAEL and
LOAEL for maternal toxicity of 1 and 10 mg/m3, respectively, were indicated.

Signs of developmental toxicity in this group consisted of statistically significant reductions in
pup body weight on Day 1 PP (6.1 g at 25 mg/m3 vs. 6.8 g in controls). On Days 4 and 22 PP,
pup body weights continued to remain lower than controls, although the difference was not
statistically significant (Day 4 PP: 9.7 g at 25 mg/m3 vs.10.3 in controls; Day 22 PP: 49.0 g at 25
mg/m3 vs. 50.1 in controls). No significant effects  were reported following external examination
of the pups or with ophthalmoscopic examination of the eyes.  Again, interpretation of these
effects is problematic given the high incidence of maternal mortality. Under the conditions of
the study, a NOAEL and LOAEL for developmental toxicity of 10 and 25 mg/m3, respectively,
were indicated.

In trial one of the oral study, three out of 25 dams died during treatment of 100 mg/kg APFO
during gestation (one death on GD 11; two on GD 12). Clinical signs of maternal toxicity in the
dams that died were similar to those seen with inhalation exposure.  Food consumption and body
weights were reduced in treated animals compared to controls. No adverse signs of toxicity
were noted for any of the reproductive parameters such as maintenance of pregnancy or
incidence of resorptions. Likewise, no significant differences between treated and control
groups were noted for fetal weights, or in the incidences of malformations and variations; nor
were there any effects noted following microscopic examination of the eyes.

In trial two of the oral study, similar observations for clinical signs were noted for the dams as in
trial one. Likewise, no adverse effects on reproductive performance or in any of the fetal
observations were noted.
                                          63

-------
SAR RPVJPW Draft' Fin Nnt riff nr
3.8 Reproductive Toxicity Studies in Animals

An oral two-generation reproductive toxicity study of APFO in rats was conducted (York, 2002;
Butenhoff et al., 2004). Five groups of 30 Sprague-Dawley rats per sex per dose group were
administered APFO by gavage at doses of 0,1, 3,10, and 30 mg/kg-day six weeks prior to and
during mating. Treatment of the FO male rats continued until mating was confirmed, and
treatment of the FO female rats continued throughout gestation, parturition, and lactation.

The FO animals were examined twice daily for clinical signs, abortions, premature deliveries,
and deaths. Body weights of FO male rats were recorded weekly during the dosage period and
then on the day of sacrifice. Body weights of FO female rats were recorded weekly during the
pre- and cohabitation periods and then on gestation days (GD) 0, 7,10,14,18, 21, and 25 (if
necessary) and on lactation days  (LD) 1, 5, 8,11,15, and 22 (terminal body weight). Food
consumption values in FO male rats  were recorded weekly during the treatment period, while in
FO female rats, values were recorded weekly during the precohabitation period, on CDs 0, 7,10,
14,18,21, and 25 and  on LDs 1, 5, 8,11, and 15.

Estrous cycling was evaluated daily by examination of vaginal cytology beginning 21 days
before the scheduled cohabitation period and continuing until confirmation of mating by the
presence of sperm in a vaginal smear or confirmation of a copulatory plug. On the day of
scheduled sacrifice, the stage of the  estrous cycle was assessed.

For mating, one male rat and one female rat per group were cohabitated for a maximum of 14
days.  Female rats with evidence of sperm in a vaginal smear or copulatory plug were designated
as GD 0. Parental females were evaluated for length of gestation, fertility  index, gestation index,
number and sex of offspring per litter, number of implantation sites, general condition of the
dam and litter during the postpartum period, litter size and viability, viability index, lactation
index, percent survival, and sex ratio. Maternal behavior of the dams was  recorded on LDs 1, 5,
8,15, and 22.

FO generation animals  were sacrificed by carbon dioxide asphyxiation (day 106 to 110 of the
study  for male rats, i.e., after completion of the cohabitation period; and LD 22 for female rats),
necropsied, and examined for gross lesions. Gross necropsy included examination of external
surfaces and orifices, as well as internal examination of tissues and organs. Individual organs
were weighed and organ-to-body weight and organ-to-brain weight ratios were calculated for the
brain, kidneys, spleen,  ovaries, testes, thymus, liver, adrenal glands, pituitary, uterus with
oviducts and cervix, left epididymis (whole and cauda), right epididymis, prostate and seminal
vesicles, (with coagulating glands and with and without fluid).  Tissues retained in neutral
buffered 10% formalin for possible histological evaluation included the pituitary, adrenal glands,
vagina, uterus, with oviducts, cervix and ovaries, right testis, seminal vesicles, right epididymis,
and prostate. Histological examination was performed on tissues from 10  randomly  selected rats
per sex from the control and high dosage groups. All gross lesions were examined histologically.
All FO generation rats that died or appeared moribund were  also examined.

Histological examination of the reproductive organs in the low- and mid-dose groups was
conducted in rats that exhibited reduced fertility by either failing to mate, conceive, sire, or
deliver healthy offspring; or for which estrous cyclicity or sperm number, motility, or
morphology were altered. Sperm number, motility, and morphology were  evaluated in the left
cauda epididymis of FO generation male rats; testicular spermatid concentrations were evaluated
in the left testis. The number and distribution of implantation sites were recorded in FO


                                          64

-------
SAR Rpyipyy nrflff F>n TSfpt
generation female rats. Rats that did not deliver a litter were sacrificed on GD 25 and examined
for pregnancy status.  Uteri of apparently nonpregnant rats were examined to confirm the
absence of implantation sites. A gross necropsy of the thoracic, abdominal and pelvic viscera
was performed. Female rats without a confirmed mating date that did not deliver a litter were
sacrificed on an estimated day 25 of gestation.

At scheduled sacrifice, after completion of the cohabitation period in FO male rats and on LD 22
in FO female rats, blood samples (10 males and 10 females each for the 10 and 30 mg/kg-day
dose groups; 3 males  and 3 females for the control group) were collected for analysis of PFOA.

The Fl generation pups in each litter were counted once daily.  The litter sizes were not
standardized on day 4 or at anytime during lactation. Physical signs (including variations from
expected lactation behavior and gross external physical anomalies) were recorded for the pups
each day. Pup body weights were recorded on LDs 1, 5, 8,15 and 22.  On LD 12, all Fl
generation male pups were examined for the presence of nipples.  Pups that died before
examination of the litter for pup viability on LD 1 were evaluated for vital status at birth. Pups
found dead on LDs 2  to 22 were examined for gross lesions and for the cause of death. All Fl
generation rats were weaned on LD 22 based on observed growth and viability of these pups.

At weaning (LD 22),  two Fl generation pups per sex per litter per group (60 male and 60 female
pups per group) were selected for continued evaluation, resulting in 600 total rats (300 rats per
sex) assigned to the five dosage groups. At least two male pups and two female pups per litter,
when possible, were selected. Fl generation pups not selected for continued observation for
sexual maturation were sacrificed.  Three pups per sex per litter were examined  for gross lesions.
Necropsy included a single cross-section of the head at the level of the frontal-parietal  suture and
examination of the cross-sectioned brain for apparent hydrocephaly.  The brain, spleen and
thymus from one of the three selected pups per sex per litter were weighed and the brain, spleen,
and thymus from the three selected pups per sex per litter were retained for possible histological
evaluation.  All remaining pups were discarded without further examination.

 The Fl generation rats were given the same dosage level of the test substance and in the same
manner as their respective FO generation sires and dams. Dosages were given once daily,
beginning at weaning and continuing until the day before sacrifice. Fl generation female rats
were examined for age of vaginal patency, beginning on day 28 postpartum (LD 28). Fl
generation male rats were evaluated for age of preputial separation, beginning on day 39
postpartum (LD 39).  Body weights were recorded when rats reached sexual maturation.

Following sexual maturation, a table of random units was used to select one male and one female
per litter per group for continuation through mating to produce the F2 generation.  The
remaining Fl animals were sacrificed.

Estrous cycling was evaluated daily by examination of vaginal cytology beginning 21 days
before the scheduled cohabitation period and continuing until confirmation of mating by the
presence of sperm in a vaginal smear or confirmation of a copulatory plug.  On the day of
scheduled sacrifice, the stage of the estrous cycle was assessed.

A table of random units was used to assign Fl generation rats to cohabitation, one male rat per
female rat.  If random assignment to cohabitation resulted in the pairing of Fl generation
siblings, an alternate assignment was made. The cohabitation period consisted of a maximum of
14 days.


                                          65

-------
SAR Rpvipw Dr^ff; Fin Nnf Tifp nr Qnnfp
Body weights of the Fl generation male rats were recorded weekly during the postweaning
period and on the day of sacrifice.  Body weights of the Fl generation female rats were recorded
weekly during the postweaning period to cohabitation, and on GDs 0, 7,10,14,18, 21 and 25 (if
necessary) and on LDs 1, 5, 8,11,15 and 22.  Food consumption values for the Fl generation
male rats were recorded weekly during the dosage period.  Food consumption values for the Fl
generation female rats were recorded weekly during the postweaning period to cohabitation, on
GDs 0,7,10,14,18,21 and 25 and on LDs 1,5, 8,11 and 15.  Because pups begin to consume
maternal food on or about LD 15, food consumption values were not tabulated after LD 15.

At scheduled sacrifice, the Fl animals were subjected to gross necropsy, and selected organs
were weighed and examined histologically as described above for the FO animals. Sperm
analyses were also conducted as described for the FO animals.

F2 generation litters were examined after delivery to identify the number and sex of pups,
stillbirths, live births and gross alterations. Each litter was evaluated for viability at least twice
each day of the 22-day postpartum period.  Dead pups observed at these times were removed
from the nesting box. Anogenital distance was measured for all live F2 generation pups on LDs
1 and 22, and F2 male pups  were examined for the presence of nipples on LD 12.

3.8.1 FO Generation

3.8.1.1 FO Males

One FO male rat in the 30 mg/kg-day dose group was sacrificed on day 45 of the study due to
adverse clinical signs (emaciation,  cold-to-touch, and decreased motor activity). Necroscopic
examination in that animal revealed a pale and tan liver, and red testes. All other FO generation
male rats survived to scheduled sacrifice.  Statistically significant increases in clinical signs were
also observed hi male rats in the high-dose group that included dehydration, urine-stained
abdominal fur, and ungroomed coat.

Significant reductions in body weight and body weight gain were reported for most of the dosage
period and continuing until termination of the study in the 3,10, and 30 mg/kg-day dose groups.
Absolute food consumption values were also significantly reduced during these periods in the 30
mg/kg-day dose group, while significant increases in relative food consumption values were
observed in the 3,10, and 30 mg/kg-day within those same periods.

No treatment-related effects were reported at any dose level for any of the mating and fertility
parameters assessed, including numbers of days to inseminate, numbers of rats that mated,
fertility index, numbers of rats with confirmed mating dates during the first and second week of
cohabitation, and numbers of pregnant rats per rats in cohabitation. At necropsy, none of the
sperm parameters evaluated (sperm number, motility, or morphology) were affected by treatment
at any dose level.

At necropsy, statistically significant reductions in terminal body weights were seen at 3,10, and
30 mg/kg-day (6%, 11%, and 25% decrease from controls, respectively; p<0.05). Absolute
weights of the left and right epididymides, left cauda epididymis, seminal vesicles (with and
without fluid), prostate, pituitary, left and right adrenals, spleen, and thymus were also
statistically significantly reduced at 30 mg/kg-day.  These absolute organ weight reductions are
probably due to reductions in body weight and not a reflection of target organ toxicity since the
organ-to-body weight ratios were either normal or significantly increased. The biological


                                           66

-------
SAR RPVJP^ Draff; Ttn Nnt Tito nr
significance of the weight changes for the adrenals, however, is unclear since dose-related
histopatholpgical changes were observed. The absolute weight of the seminal vesicles without
fluid was significantly reduced in the 10 mg/kg-day dose group. The absolute weight of the liver
was statistically significantly increased in all dose-groups.  Absolute kidney weights were
statistically significantly increased in the 1, 3, and 10 mg/kg-day dose groups, but significantly
decreased in the 30 mg/kg-day group. Organ weight-to-terminal body weight ratios for the liver
and for the left and right kidney were statistically significantly increased in all treated groups.
The biological significance of the weight changes observed in the liver and kidney is unknown
since no histopathology was conducted on these organs.  Organ weight-to-terminal body weight
ratios for the brain were statistically significantly increased for the 3, 10, and 30 mg/kg-day dose
groups; organ weight-to-brain weight ratios were significantly reduced for some organs at the
high dose group, and significantly increased for other organs among all treated groups.

No treatment-related effects were seen at necropsy or upon microscopic examination of the
reproductive organs, with the exception of increased thickness and prominence of the zona
glomerulosa and vacuolation of the cells of the adrenal cortex in 2/10 males in the 10 mg/kg-day
group and 7/10 males in the 30 mg/kg-day dose group.

Serum analysis for the FO generation males in the control, 10 and 30 mg/kg-day groups sampled
at the end of cohabitation showed that PFOA was present in all samples tested, including
controls.  Control males had an average concentration of 0.0344+ 0.0148 jig/ml PFOA. Levels
of PFOA were similar in the two male dose groups; treated males had 51.1+9.30 and 45.3+12.6
     l, respectively for the 10 and 30 mg/kg-day dose groups.
3.8.1.2 FO Females

No treatment-related deaths or adverse clinical signs were reported in parental females at any
dose level.  No treatment-related effects were reported for body weights, body weight gains, and
absolute and relative food consumption values.

There were no treatment-related effects on estrous cyclicity, mating or fertility parameters.
None of the natural delivery and litter observations were affected by treatment, that is, the
numbers of dams delivering litters, the duration of gestation, the averages for implantation sites
per delivered litter, the gestation index (number of dams with one or more livebom pups/number
of pregnant rats), the numbers of dams with stillborn pups, dams with all pups dying, livebom
and stillborn pups viability index, pup sex ratios, and mean birth weights were comparable to
controls among all treated groups.

Necropsy and histopathological  evaluation were also unremarkable. Terminal body weights,
organ weights, and organ-to-terminal body weight ratios were comparable to control values for
all treated groups, except for kidney  and liver weights. The absolute weights of the left and right
kidney, and the ratios of these organ weights-to-terminal body weight and of the left kidney
weight-to-brain weight were significantly reduced at the highest dose of 30 mg/kg-day.  The
biological significance of these weight changes is not known since histopathology was not
conducted on the kidney.  The ratio of liver weight-to-terminal body weight was significantly
reduced at 3 and 10 mg/kg-day,  but there were no effects observed at 30 mg/kg-day or in
absolute liver weight.

Serum PFOA levels were analyzed from the control, 10 and 30 mg/kg-day groups on LD 22.
The samples were collected 24 hours after dosing. In the controls, serum PFOA was below the


                                           67

-------
SAR Rpyipw Draff; Dn 1\nf riff nr
limits of quantitation (0.00528 ng/ml). Levels of PFOA found in female sera increased between
the two dose groups; treated females had an average concentration of 0.37+0.0805 and
1.02+0.425 jig/ml, respectively for the 10 and 30 mg/kg-day dose groups.

3.8.2 Fl Generation

No effects were reported at any dose level for the viability and lactation indices. No differences
between treated and control groups were noted for the numbers of pups surviving per litter, the
percentage of male pups, litter size and average pup body weight per litter at birth. At 30 mg/kg-
day, one pup from one dam died prior to weaning on lactation day 1 (LD1).  Additionally, on
lactation days 6 and 8, statistically significant increases in the numbers of pups found dead were
observed at 3 and 30 mg/kg-day.  According to the study authors, this was not considered to be
treatment related because they did not occur in a  dose-related manner and did not appear to
affect any other measures of pup viability including numbers  of surviving pups per litter and live
litter size at weighing. An independent statistical analysis was conducted by US EPA (2002b).
No significant differences were observed between dose groups and the response did not have any
trend in dose.

The authors  did not present the mean pup body weights for the male and female pups separately.
Pup body weight on a per litter basis (sexes combined) was reduced throughout lactation in the
30 mg/kg-day group, and statistical significance (p < 0.01) was  achieved on days 1 (9.5%), 5
(9.6%), and  8 (10.5%).  Of the pups necropsied at weaning, no statistically significant,
treatment-related differences were observed for the weights of the brain, spleen and thymus and
the ratios of these organ weights to the terminal body weight and brain weight.

3.8.2.1 Fl Males

Significant increases in treatment-related deaths (5/60 animals) were reported in Fl males in the
high dose group of 30 mg/kg-day between days 2-4 postweaning. One rat was moribund
sacrificed on day 39 postweaning and another was found dead on day 107 postweaning.

Statistically  significant increases in clinical signs of toxicity were also observed in Fl males
during most of the entire postweaning period. These signs included an increased incidence of
annular constriction of the tail at all doses, with statistical significance at the 1,10, and 30
mg/kg-day; a significant increase at 10 and 30 mg/kg-day in the number of male rats that were
emaciated; and a significant increase in the incidence of urine-stained abdominal fur, decreased
motor activity, and abdominal distention at 30 mg/kg-day.

Statistically  significant reductions in body weight gain were observed at 10 and 30 mg/kg-day
during days  8-15, 22-29, 29-36, 43-50, and 50-57 postweaning. Body weight gains were also
significantly reduced in the 30 mg/kg-day group on days 1-8,15-22, 36-43, 57-64, and 64-70
postweaning. Body weights were significantly reduced in the 10 mg/kg-day group beginning on
postweaning day 36, and in the 30 mg/kg-day group beginning on postweaning day 8. In the 3
mg/kg-day group, mean body weight gain was significantly reduced on days 43-50 and 57-64
postweaning, and mean body weights were significantly reduced on days  106 and 113
postweaning. In the 1 mg/kg-day group, mean body weight gain was significantly reduced on
days 15-22 and 43-50 postweaning, and mean body weights were significantly reduced on days
50, 57, 64, 70, 99,106 and 113 postweaning.  For all groups, there was a significant, dose-
related reduction in mean body weight gain for the entire dosing period (days 1-113). Absolute
food consumption values were significantly reduced at 10 and 30 mg/kg-day during the entire


                                           68

-------
SAR RPVIPW Draft- Tin Nnt Cite nr
precohabitation period (days 1-70 postweaning), while relative food consumption values were
significantly increased.

Statistically significant (p< 0.01) delays in sexual maturation (the average day of preputial
separation) were observed in high-dose animals versus concurrent controls (52.2 days of age
versus 48.5 days of age, respectively).

No apparent effects were observed on any of the mating or fertility parameters including fertility
and pregnancy indices (number of pregnancies per number of rats that mated and rats in
cohabitation, respectively), the number of days to inseminate, the number of rats that mated, and
the number of rats with confirmed mating dates during the first week.  No statistically
significant, treatment-related effects were observed on any of the sperm parameters (motility,
concentration, or morphology).

Necroscopic examination revealed statistically significant treatment-related effects at 3,10, and
30 mg/kg-day ranging from tan areas in the lateral and median lobes of the liver to moderate to
slight dilation of the pelvis of one or both kidneys.

Statistically significant, dose-related decreases in terminal body weights were observed in the 1,
3,10, and 30 mg/kg-day dose groups(6%, 6%, 11%, and 22% decrease from controls,
respectively; p<0.01 at 1 and 3 mg/kg-day, p<0.05 at 10  and 30 mg/kg-day). The absolute and
relative weights of the liver were statistically significantly increased in all treated groups (p <
0.01) and was accompanied by histopatholpgical changes. The absolute weights of the left
and/or right kidneys were statistically significantly increased in the 1 and 3 mg/kg-day dose
groups and statistically significantly decreased in the 30 mg/kg-day dose group. Organ weight-
to-terminal body weight and brain weight ratios for the kidney were statistically significantly
increased in all treated groups.  The biological significance of the effects on kidney weight is
unknown since histopathology was  not conducted on that organ. All other organ weight changes
observed (thymus,  spleen, left adrenal, brain, prostate, seminal vesicles, testes, and epididymis)
are probably due to decrements in body weight and not a reflection of target organ toxicity since
the absolute weights of these organs was significantly reduced while the relative weights were
either normal or significantly increased. However, the biological significance of the weight
changes observed in the adrenal is unclear since histopathological changes were also observed.

Histopathologic examination of the reproductive organs was unremarkable; however, treatment-
related microscopic changes were observed in the adrenal glands of high-dose animals
(cytoplasmic hypertrophy and vacuolation of the cells of the adrenal cortex) and in the liver of
animals treated with 3,10, and 30 mg/kg-day (hepatocellular hypertrophy).  No other treatment-
related effects were reported.

3.8.2.2 Fl Females

A statistically significant increase in treatment-related mortality (6/60 animals) was observed in
Fl females on postweaning days 2-8 at the highest dose of 30 mg/kg-day. No adverse clinical
signs of treatment-related toxicity were reported for any dose level during any time of the study
period.

Statistically significant decreases in body weights were observed in high-dose animals on days 8,
15, 22,29, 50, and 57 postweaning, during precohabitation (recorded on the day cohabitation
began, when Fl generation rats were 92-106 days of age), and during gestation and lactation.


                                           69

-------
SAR RPVJPW Draft; Fin Nnf fitf nr Dnnfp
Body weight gain was significantly reduced during days 1-8 and 8-15 postweaning. Statistically
significant decreases in absolute food consumption were observed during days 1-8, 8-15, and 15-
22 postweaning, during precohabitation and during gestation and lactation in animals treated
with 30 mg/kg-day.  Relative food consumption values were comparable across all treated
groups.

Statistically significant (p< 0.01) delays in sexual maturation (the average day of vaginal
patency) were observed in high-dose animals versus concurrent controls (36.6 days of age versus
34.9 days of age, respectively).

Prior to mating, the study authors noted a statistically significant increase in the average numbers
of estrous stages per 21 days in high-dose animals (5.4 versus 4.7 in controls). For this
calculation, the number of independent occurrences of estrus in the 21 days of observation was
determined.  This type of calculation can be used as a screen for effects on the estrous cycle, but
a more detailed analysis should then be conducted to determine whether there is truly an effect.
3M Company (2002) recently completed an analysis that showed there were no effects on the
estrous cycle; mere were no differences in the number of females with > 3 days of estrus or with
> 4 days of diestrus in the control and high  dose groups. Analyses conducted by the US EPA
(2002a) also demonstrated that there were no differences in the estrous cycle among the control
and high dose groups.  The cycles were evaluated as having either regular 4-5 day  cycles, uneven
cycling (defined as brief periods with irregular pattern) or periods of prolonged diestrus (defined
as 4-6 day diestrus periods)  extended estrus (defined as 3 or 4 days of comified smears), possibly
pseudopregnant, (defined as 6-greater days  of leukocytes) or persistent estrus (defined as 5-or
greater days of comified smears).  The two groups were not different in any of the parameters
measured. Thus, the increase in the number of estrous stages per 21 days that was  noted by the
study authors is due to the way in which the calculation was done, and is not biologically
meaningful.

No effects on any of the mating and fertility parameters (numbers of days in cohabitation,
numbers of rats mat mated, fertility index, rats with confirmed mating dates during the first week
of cohabitation and number  of rats pregnant per rats in cohabitation).

All natural delivery observations were unaffected by treatment at any dose level. Numbers of
dams delivering litters, the duration of gestation, averages for implantation sites per delivered
litter, the gestation index (number of dams with one or more livebom pups/number of pregnant
rats), the numbers of dams with stillborn pups, dams with  all pups dying and livebom and
stillborn pups were comparable among treated and control groups.

The terminal body weights and absolute and relative pituitary weights are shown in Table 19.
No treatment-related effects were observed in the terminal body weights of the Fl  females. The
absolute weight of the pituitary, the pituitary weight-to-terminal body weight ratio, and the
pituitary weight-to-brain weight ratio were statistically significantly decreased at 3 mg/kg-day
and higher. Since there is not a clear dose-response relationship and histologic examination did
not reveal any lesions, the biological significance of the pituitary weight data is problematic. No
other differences were reported for the absolute weights or ratios for other organs evaluated. No
treatment-related effects were reported following macroscopic and histopathologic examinations.
                                           70

-------
SAR RPVJPW Tlraft! Dn Nnt fitp nr
                                        Table 19
                   Summary of Body and Pituitary Weights of Fl Females
Group
Control
1 mg/kg-day
3 mg/kg-day
10 mg/kg-
day
30 mg/kg-
day
Terminal
Body Weight (g)a
322.9 + 23.4
321. 7 ±24.2
329.2 ±21.5
325.1+23.5
315.7 + 20.9
Pituitary
Weight (g)a
0.017 ±0.004
0.016 ±0.003
0.015 ±0.003*
0.015 + 0.002*
0.015 + 0.003*
Pituitary/
Body Weight (%)a
5.46+1.6
5.0 ±0.86
4.65 ±0.86*
4.72 + 0.90*
4.75 + 1.03*
Pituitary /Brain
Weight (%)a
0.84 ±0.21
0.76 ±0.14
0.74 ±0.14*
0.72 ±0.1 2*
0.72 + 0.16**
a- Mean ± SD
* Significantly different from control at p < 0.05
** Significantly different from control at p < 0.01

3.8.3 F2 Generation

No treatment-related adverse clinical signs were observed at any dose level. Dead or stillborn
pups were noted in both the control and treated groups.  The deaths occurred on lactation days 1-
18 with the majority occurring on days 1-6. However, there was no dose-response relationship
and therefore were unlikely related to treatment. Statistically significant increases (p< 0.01) in
the number of pups found dead were observed on lactation day 1 in the 3 and 10 mg/kg-day
groups. According to the study authors, this was not considered to be treatment related because
they did not occur in a dose-related manner and did not appear to affect any other measures of
pup viability including numbers of surviving pups per litter and live litter size at weighing.  An
independent statistical analysis was conducted by US EPA (2002b). No significant differences
were observed between dose groups and the response did not have any trend in dose.

No effects were reported at any dose level for the viability and lactation indices. No differences
between treated and control groups were noted for the numbers of pups surviving per litter, the
percentage of male pups, litter size and average pup body  weight per litter when measured on
LDs 1, 5, 8,15, or 22.  Anogenital distances measured for F2 male and female pups on LDs 1
and 22 were also comparable among the five dosage groups and did not differ significantly.

Terminal body weights in F2 pups were not significantly different from controls.  Absolute
weights of the brain, spleen and thymus and the ratios of these organ weights-to-terminal body
weight and to brain weight were also comparable among treated and control groups. There were
no treatment-related effects following necroscopic examination, with the exception of no milk in
the stomach of the pups that were found dead.

3.8.4 Conclusions

Dosing with APFO at 30 mg/kg-day resulted in a delay in the onset of sexual maturation in both
male and female Fl offspring. The authors of the study contend that the delays in sexual
maturation (preputial separation or vaginal patency) observed in high-dose animals are due to the
                                           71

-------
SAR RPVIPW Draft; Ttn Nnf rite nr Oimfp
fact that these animals have a decreased gestational age, a variable which they have defined as
the time in days from evidence of mating in the FO generation until evidence of sexual
maturation in the Fl generation. The authors state that gestational age appeared to be decreased
in high-dose animals at the time of acquisition (the time when sexual maturation was reached),
which they believe meant the animals in that group were younger and more immature than the
control group, in which there was no significant difference in sexual maturation.

In order to test this hypothesis, the authors covaried separately the decreases in body weight and
in gestational age with the delays in sexual maturation in order to determine whether or not body
weights and gestational age were a contributing factor. When the body weight was covaried
with the time to sexual maturation, the time to  sexual maturation showed a dose related delay
that was statistically significant at p< 0.05.  This suggests that the delay in sexual maturation was
partly related to body weight, but not entirely. When gestational age was covaried with the time
to sexual maturation, there was no significant difference in the time of onset of sexual
maturation between controls and high-dose animals. This indicates that the effect of delayed
sexual maturation could possibly be attributed to decreased gestational age.

While it is known and commonly accepted that changes in the body weights of offspring can
affect the time to sexual maturation, whether or not gestational age, as defined by the authors,
also affects the time of sexual maturation is purely speculative, especially since there were no
data provided by the authors to support this relationship.  Additionally, covaring gestational age
with tune to sexual maturation is problematic from a statistical standpoint.  Since there was no
significant change in the length of gestation at 30 mg/kg-day, based on the authors' definition of
'gestational age', the decreases in gestational age would have to be due mostly to changes in
time to sexual maturation.  Therefore, sexual maturation is essentially being covaried with itself.
Still, even if a relationship between gestational age and time to sexual maturation were shown, it
merely offers an explanation for the observed delays in sexual maturation in high-dose animals,
but does not diminish its significance.

A variety of endpoints are evaluated throughout different lifestages in a two generation
reproductive toxicity study. Therefore, some of the endpoints may be indicative of
developmental/reproductive toxicity, while others may be indicative of adult toxicity. The
selection of developmental endpoints (and the appropriate LOAELs and NOAELs) for this study
was based on the Agency's Developmental Toxicity Risk Assessment Guidelines (EPA, 1991)
and the Agency's Reproductive Toxicity Risk Assessment Guidelines (EPA, 1996).  According
to the guidelines, the period of exposure for developmental toxicity is prior to conception in
either parent,  through prenatal development and continuing until sexual maturation.  In contrast,
the  period during which a developmental effect may be manifested includes the entire lifespan of
the  organism. For selection of the developmental  endpoints from the two generation
reproductive toxicity study, attention was focused on effects that were noted during the period of
developmental exposure.  Thus, only effects that occurred up to sexual maturation were
considered relevant for assessing developmental toxicity.  Effects occurring after sexual
maturation were considered relevant for assessing adult toxicity since it was not possible to
determine whether the effects were due to developmental and/or adult exposures.

Therefore, under the conditions of the study, the LOAEL for FO parental males is 1 mg/kg-day,
the  lowest dose tested, based on significant increases in absolute and  relative liver weight. A
NOAEL for the FO parental males could not be determined since the increases in liver weight
were seen at all doses tested. The NOAEL for FO parental females is 30 mg/kg-day, the highest
dose tested.
                                           72

-------
SAR RPVIPW Draft; Tin Nnt fite nr Quote
A variety of developmental/reproductive effects were noted in the Fl generation. During
lactation, there was a significant reduction in Fl mean body weight on a litter basis (sexes
combined) in the 30 mg/kg-day group. Fl males in the 10 and 30 mg/kg-day groups exhibited a
significant reduction in body weight gain during days 8-50 postweaning, and body weights were
significantly reduced in the 10 mg/kg-day group beginning on postweaning day 36, and in the 30
mg/kg-day  beginning on postweaning day 8. Fl females in the 30 mg/kg-day group exhibited a
significant reduction in body weight gain on days 1-15 postweaning, and in body weights
beginning on day 8 postweaning. There was a significant increase in mortality mainly during the
first few days after weaning, and a significant delay in the timing of sexual maturation for Fl
males and females in the 30 mg/kg-day group.  For Fl males, the LOAEL for
developmental/reproductive toxicity was 10 mg/kg-day, and the NOAEL was 3 mg/kg-day.  For
Fl females, the LOAEL for developmental/reproductive toxicity was considered to be 30 mg/kg-
day, and the NOAEL was 10 mg/kg-day.

The LOAEL for adult systemic toxicity in the Fl males is  1 mg/kg-day based on significant,
dose-related decreases in body weights and body weight gains (observed prior to and during
cohabitation and during the entire dosing period), and in terminal body weights; and significant
changes in absolute and relative liver weights.  A NOAEL for the Fl males could not be
determined since these effects were seen at all doses tested.

The NOAEL and LOAEL for adult systemic toxicity in the Fl females are 10 and 30 mg/kg-day,
respectively, based on statistically significant decreases in body weight and body weight gains.

The NOAEL for developmental/reproductive toxicity in the F2 offspring was 30 mg/kg-day. No
treatment-related effects were observed at any doses tested in the study. However, it should be
noted that the F2 pups were sacrificed at weaning, and thus it was not possible to ascertain the
potential post-weaning effects that were noted in the Fl generation.

3.9 Mode of Action and Summary of Weight of Evidence

3.9.1 Epidemiology Studies

All of the epidemiologic data available on PFOA are based on occupational studies, most of
which were routine biomonitoring efforts conducted by 3M Company. With the exception of
one, all of the studies were cross-sectional and mostly analyzed males, the majority of workers at
these plants. Consequently, reproductive and developmental outcomes have not been studied.

In the 3M mortality studies, the only statistically significant association reported was for prostate
cancer mortality and employment duration in a plant that manufactures PFOA; however, this
association was not observed in a follow-up study which included more specific exposure
measures (Gilliland and Mandel, 1993; Alexander, 2001a). In the Dupont cancer study, bladder
and kidney  cancer incidence was elevated among employees (Dupont, 2003). However, very
little other data on exposures, including other chemicals at the plant, were available. A study on
hormone levels in male workers indicated an increase in estradiol levels in workers with the
highest PFOA serum levels; however, the results may have been confounded by body mass
index (Olsen, et al., 1998a).  There were no changes of note in other hormone levels in these
workers.  Another study of CCK levels in employees, in which increased levels have been linked
to pancreas acinar cell adenomas in rats, did not report increases in workers (Olsen, et al.,  1998b;
Olsen et al., 2000). In addition, cholesterol and triglyceride levels in workers were positively
associated with PFOA exposures, inconsistent with the hypolipidemic results reported in rat


                                          73

-------
 SAB RPVIPW Draft: Dn Nnf TifA nr Onnfp
 studies.  There was also a statistically significant positive association reported for PFOA and T3
 but not any other thyroid hormones (Olsen, et al., 2001e).
o
 Overall,  there were no notable health effects reported in fluorochemical workers which can be
 directly attributed to PFOA exposure. However, there were many limitations in these studies
 (detailed in Section 3.1) which need to be considered.

 3.9.2 Metabolism and Pharmacokinetics

 Little information is available concerning the pharmacokinetics of PFOA and its salts in humans.
 An ongoing 5-year, half-life study in 7 male and 2 female retired workers has suggested a mean
 serum PFOA half-life of 4.37 years (range, 1.50 - 13.49 years). -Metabolism and
 pharmacokinetic studies in non-human primates are limited to a study of 3 male and 3 female
 cynomolgus monkey administered a single IV dose of 10 mg/kg potassium PFOA.  In male
 monkeys, the average serum half life was 20.9 days.  In female monkeys, the average serum half
 life was 32.6 days.

 Studies in adult rats have shown that PFOA is absorbed following oral, inhalation, and dermal
 exposures. Serum pharmacokinetic parameters of PFOA have been evaluated in adult Sprague-
 Dawley rats following gavage administration, and in Wistar rats following IV administration.
 The distribution of PFOA has been examined in tissues of adult rats following administration by
 gavage and by i.v.  and i.p. injection. PFOA distributes primarily to the liver, serum, and kidney,
 and to a lesser extent, other tissues of the body.  It does not partition to the lipid fraction or
 adipose tissue. Although investigated in several studies, there is no evidence that PFOA is
 metabolized. In adult rats, there is evidence of enterohepatic circulation of PFOA.  The urine is
 the major route of excretion of PFOA in the female rat, while the urine and feces are both main
 routes of excretion in male rats.  There are gender differences in the elimination of PFOA in rats.
 In female rats, following oral administration, estimates of the serum half-life were dependent on
 dose and ranged from approximately 2.8 to 16 hours, while in male rats estimates of the serum
 half-life  following oral administration were independent of dose and ranged  from approximately
 138 to 202 hours. In female rats, elimination of PFOA appears to be biphasic with a fast phase
 and a slow phase. The rapid excretion of PFOA by female rats is believed to be due to active
 renal tubular secretion (organic acid transport system); this  renal tubular secretion is believed to
 be hormonally controlled.  Hormonal changes during pregnancy do not appear to cause a change
 in the rate of elimination in rats.

 No studies have been conducted to specifically examine the absorption, metabolism or
 elimination of PFOA in the developing rat.  However, recent studies have been conducted to
 examine the concentrations of PFOA in the developing Sprague-Dawley rat, and to determine
 when the gender difference in elimination of PFOA becomes apparent.  In addition, several
 studies have examined the serum and tissue distribution of PFOA in newly weaned Wistar rats.
 These studies have shown that PFOA readily crosses the placenta and is present in the breast
 milk of rats.  During lactation and  for the first several weeks after weaning, the elimination of
 PFOA is similar in males and females. Between 4-5 weeks of age, the elimination in male rats
 assumes the adult pattern and the gender difference becomes readily apparent.  Distribution
 studies in the postweaning rat have shown that PFOA is distributed primarily to the serum, liver,
 and kidney.

 It has been suggested that PFOA is circulated around the body by noncovalently binding to
 plasma proteins. Several studies have investigated the binding of PFOA to plasma proteins of


                                           74

-------
SAR RPVIPW Uraft? Ttn Nnf Pifp nr
rats, humans or monkeys to gain understanding of its absorption, distribution and elimination
and species and gender differences. There is limited additional information on the metabolism
and pharmacokinetics of PFOA in mice, rabbits and dogs.

3.9.3 Mode of Action Analyses and Cancer Descriptor

There has been a great deal of scientific debate and research into rodent liver toxicity and liver
tumors that are induced through the activation of PPARa, and several groups of scientists have
summarized the state of the science over the years (IARC, 1995; Cattley et al. 1998). More
recently, the ILSI Risk Science Institute convened a workgroup to re-examine the mode of action
of PPARa agonist-induced rodent liver tumors as well as to evaluate the mode(s) of action for
Leydig cell and pancreatic acinar cell tumors, which also are observed frequently in rats with
PPARa agonists (Klaunig et al., 2003).  This effort utilized and extended guidance that the US
EPA (1999) and IPCS (Sonich-Mullin et al., 2001) have provided for assessing information for
the delineation of an animal mode of action (MOA) for specific tumor types. This guidance is
also applicable for assessing MOAs for non-tumor responses. In this context, it is important to
note that the term "mode of action" is defined as the sequence of key cellular and biochemical
events (measurable parameters) that result in a toxicological effect.  The analysis does not
require knowledge of "mechanism of action" which implies a more detailed understanding of the
molecular basis of the toxicological effect.

Recently, EPA's Office of Prevention, Pesticides and Toxic Substances (OPPTS) presented a
draft of a proposed OPPTS science policy on PPARa mediated hepatotoxicity and
heptacarcinogenesis to the FIFRA Science Advisory Panel (SAP) in December, 2003 (OPPTS,
2003). The OPPTS guidance document describes the approach the Office proposes to use to
evaluate the scientific information regarding the mode of action of PPARa agonists in rodent
hepatocarcinogenesis and the relevance of this mode of action for human hepatocarcinogenesis.
Other tumor types (e.g. Leydig cell and pancreatic acinar cell tumors) that may be associated
with PPARa-agonists are also briefly described. The document provides an overview of the
evidence for a PP ARa-agonist mode of action for liver tumors in rodents, and an overview of all
known age and species differences in the key events. The document also provides proposed
guidance on the data needed to demonstrate that rodent liver toxicity and tumors have arisen as a
result of a PPARa agonist mode of action, and the relevance  of this mode of action for humans.

As described in section 3.5, several studies of PFOA in adult rats have shown that the liver is a
principle target organ, and long-term exposure results in liver adenomas, Leydig cell adenomas,
and pancreatic acinar cell tumors. The mode of action information for each of these effects are
described below and are described in the context of the work of Klaunig et al. (2003), the
proposed OPPTS science policy and the response from the FIFRA SAP (FIFRA SAP, 2004).

3.9.3.1 Mode of Action Analysis of Liver Toxicity and Liver Adenomas in Rats

There are a number of possible modes of action for hepatocarcinogenesis of chemicals. As
summarized in Section 3.4, the weight of evidence from short-term genotoxicity assays suggest
that PFOA is not a DNA-reactive compound.

As mitochondria play a major role in cell signaling and apoptotic modes of cell death, and several
structurally related perfluorinated compounds have been shown to manifest their toxicity by
interfering with mitochondria biogenesis and bioenergetics, the effects of PFOA on
mitochondria! biogenesis and bioenergetics were investigated.  A number of studies have shown


                                          75

-------
SAR Review T>rflft r>" N"t
that mitochondria biogenesis in liver was increased following treatment of rats with PFOA (e.g.,
Berthiaume and Wallace, 2002).  PFOA has also been demonstrated to uncouple oxidative
phosphorylation in mitochondria of the liver from rats exposed to PFOA in the diet (Keller et al.,
1992). At high concentrations, PFOA caused a small increase in resting respiration rate and
slightly decreases the membrane potential.  The observed effects are believed to be attributed to a
slight increase in nonselective permeability of the mitochondria membranes caused by the
surface-active property of the compound (Starkov and Wallace, 2002).  Treatment with 200 \iM
PFOA was found to cause a dramatic increase in the cellular content of reactive oxygen species
(ROS) in human hepG2 cells. The activation of caspase-9 and apoptosis by PFOA observed was
postulated to be the result of the disruption of mitochondria membrane and accumulation of ROS
(Panaretakis et al., 2001). Further research is needed, however, to elucidate how apoptosis is
involved in tumorigenesis of PFOA.

Gap junctional intercellular communication (GJIC), a process by which cells exchange ions,
second messages, and other small molecules, is important for normal growth, development, and
differentiation, as well as maintenance of homeostasis in muticellular organisms.  Because tumor
formation requires loss of homeostasis and many tumor promoters inhibit GJIC, it has been
hypothesized that GJIC may play a role in carcinogenesis (Trosko et al., 1998).  PFOA has been
demonstrated to inhibit GJIC in liver cells in vitro and in vivo (Upham et al., 1998).  Since
inhibition of GJIC is a widespread phenomenon, and the effect by PFOA was neither species nor
tissue specific and was reversible, the significance of GJIC inhibition in regard to the mode of
carcinogenic action of PFOA is unknown.

Estrogen has been shown to promote hepatocarcinogenesis in rats (Yager and Yager,  1980;
Cameron et al., 1982). However, more research is needed to support the involvement of this
MO A in the hepatocarcinogenesis of PFOA.

Low and sporadic incidences of liver necrosis were noted in both control and treated rats in the
subchronic and chronic toxicity studies of PFOA. Liver necrosis and regenerative cell
proliferation may play a role in promoting pre-initiated cells in carcinogenesis. The involvement
of necrosis in the liver tumor induction by PFOA in rats remains to be investigated.

While available data are not sufficient to support any of the above MOAs for the liver tumor
induction by PFOA, there is strong evidence to conclude that the liver toxicity and liver
adenomas mat are observed in rats following exposure to PFOA result from a PPARa-agonist
MO A. As described in Klaunig et al. (2003) and OPPTS (2003), the MOA of PPARa-agonist
induced liver toxicity and liver tumors involves four causal key events which are shown in Figure
1.  (Page 78, below.) The first key event is activation of  PPARa (which regulates the
transcription of genes involved in peroxisome proliferation, cell cycle control, apoptosis, and
lipid metabolism). Activation of PPARa leads to an increase in cell proliferation and a decrease
in apoptosis, which in turn leads to preneoplastic cells and further clonal expansion and formation
of liver tumors. Of these key events, only PPARa activation is highly specific for this MOA
while cell proliferation/apoptosis and clonal expansion are common to other modes of action.  As
depicted in Figure 1, there are also several "associative" events that are markers of PPARa
agonism but are not directly  involved in the etiology of liver tumors.  These include peroxisome
proliferation (a highly specific indicator that this MOA is operative) and peroxisomal gene
expression. Peroxisomal proliferation may also result in hepatocyte oxidative stress which may
contribute to the mode of action by causing indirect DNA damage and leading to mutations, or by
stimulating cell proliferation. However, increases in oxidative damage to DNA have not been
unambiguously demonstrated for PPARa agonists. Oxidative stress is a general phenomenon,


                                          76

-------
SAB Rpyjew r>**aft  r>n Nnt Pitp nr
and thus does not represent a highly specific marker for PPARa-agonist induced liver
carcinogenesis.

OPPTS (2003) provided some guidance on information that would help establish that a chemical
is inducing liver toxicity and tumors via a PPARoc agonist MO A.  This includes in vitro evidence
of PPARa agonism (i.e., evidence from an in vitro receptor assay), in vivo evidence of an increase
in number and size of peroxisomes, increases in the activity of acyl CoA oxidase, and hepatic cell
proliferation. The in vivo evidence should be collected from studies designed to provide the data
needed to show dose-response and temporal concordance between precursor events and liver
tumor formation. Other information that is desirable and may strengthen the weight of evidence
for demonstrating that a PPARa agonist MOA is operative includes data on hepatic CYP4A1
induction, palmitoyl CoA activity, hepatocyte hypertrophy, increase in liver weights, decrease in
the incidence of apoptosis, increase in microsomal fatty acid oxidation, and enhanced formation
of hydrogen peroxide.

There is sufficient information to demonstrate the key events for a PPARa agonist MOA
following exposure to PFOA in rodents. It has been well documented that PFOA is a potent
peroxisome proliferator, inducing peroxisome proliferation in the liver of rats  and mice (e.g.,
Ikeda et al., 1985; Pastoor et al., 1987; Sohlenius et al., 1992).  Maloney and Waxman (1999)
demonstrated that PFOA activates PPARa using COS1 cells transfected with  a luciferase reporter
gene. Maximal transcriptional activity with PFOA was seen at 10 ^M in the mouse PPARa (and
at 20 uM in human PPARa). Like many other peroxisome proliferators, PFOA has also been
shown to cause hepatomegaly (an early biomarker of peroxisome proliferator
hepatocarcinogenesis) in rats (Takagi et al., 1992; Cook,  1994) and mice (Kennedy, 1986), and
induce oxidative DNA damage in liver of rats (Takagi et al., 1991).
                                          77

-------
SAR RPVJPW Draft; Tin Not fite nr Oiinfp
                                      Figure 1
           Key  Events in the Mode of Action for PPARa-
                Agonist Induced  Rodent Liver Tumors
           PPARa Agonist
                                   Causative Events
                                 Activation of PPAR a
                                    Cell Proliferation
                                  Decreased Apoptosis

                                          1
                                  Preneoplastic Foci
                                          I
                                   Clonal Expansion
                                          I
                                     Liver Tumors
                                                       Associative Events*
                                                     •Expression of Peroxisomal Genes
                                                     •Increase in Peroxisomes (number
                                                     & size)
     Although there are other biological events (e.g., Kupffer cell mediated events, inhibition of gap junctions), the
     measurements of peroxisome proliferation and peroxisomal enzyme activity (in particular acyl-CoA) are widely used as
     reliable markers of PPARa activation.
From OPPTS (2003)
                                         78

-------
SAR RPVJPW Draft; Dn Nnt Pifp nr
Several studies have been conducted to examine the dose-response and temporal relationships
among key endpoints. The temporal and dose-response relationship of measures of peroxisome
proliferation, hepatocellular hypertrophy, liver weight, and liver histopathology have been
examined in male Sprague-Dawley rats following 4, 7 and 13 weeks of administration of dietary
PFOA at doses ranging from 0 - 6.5 mg/kg-day (Palazzolo, 1993).  The results are summarized in
Table 20.  There was no evidence of peroxisome proliferation, hepatocellular hypertrophy or liver
weight increases at 0.06 mg/kg-day. At doses ranging from 0.64 to 6.5 mg/kg-day there is a clear
relationship between peroxisome proliferation (indicated by increased palmitoyl CoA oxidase
activity), hepatocellular hypertrophy and increases in liver weight at all time points. There was
no evidence of hepatocellular necrosis.

                                         Table 20
       Summary of Liver Effects in Male Sprague-Dawley Rats  Fed APFO for 90 Days
Parameter
Palmitoyl CoA
Oxidase (IU/G)
Hepatocellular
Hypertrophy
Hepatocellular
Necrosis,
Coagulative
Absolute Liver
Weight (g)
Liver/Body
Weight (%)
Week
4
7
13
4
7
13
4
7
13
4
7
13
4
7
13
Dose (mg/kg-day)1
0"
8(0.5)
7(1.5)
8 (0.9)
0/15
0/15
0/15
0/15
0/15
0/15
16.34
(2.14)
17.78
(2.12)
19.73
(2.01)
3.97
(0.37)
3.75
(0.29)
3.53
(0.28)
Ob
5 (0.4)
7(1.5)
5(1.1)
0/15
0/15
0/15
1/15
0/15
0/15
15.83
(1.13)
16.91
(2.22)
16.30
(1.62)
4.07
(0.27)
3.76
(0.37)
3.23
(0.23)
0.06
9(1.7)
7 (0.8)
8(1.9)
0/15
0/15
0/15
0/15
0/15
1/15
15.45
(1.71)
17.68
(??)
18.03
(2.81)
3.73
(0.23)
3.64
(0.33)
3.24
(0.30)c
0.64
14 (3.8)c
18(5.5)
10(2.1)
12/15
12/15
13/15
0/15
0/15
0/15
17.89
(2.13)
19.42
(2.10)
20.44
(2.87)
4.48
(0.32)d
4.12
(0.37)
3.69
(0.32)
1.94
24(11.4)c
32(12.2)°
14 (3.4)°
15/15
15/15
14/15
1/15
0/15
1/15
23.23
(2.83)c
27.76
(3.51)'
22.74
(4.21)
5.77
(0.60)d
5.14
(0.53)c
4.21
(0.56)°
6.5
37 (14.8)cd
54 (35.3)cf
17 (4.5)cd
14/15
15/15
15/15
2/15
1/15
0/15
25.44
(1.89)cd
27.76
(3.51)cd
26.78
(5.47)cd
6.73
(0.49)de
6.06
(0.59)cd
5.49
(0.84)cd
1- Mean (SD)
a - non-pair-fed controls
b - pair-fed controls
c - statistically significant at p < 0.05 with the non-pair-fed control
d - statistically significant at p < 0.05 with the pair-fed control
e - calculated using the non-pair-fed control
f - calculated using the pair-fed control
                                            79

-------
                   Tin Nnt fifp nr
Liu et al. (1996) characterized the dose-response relationships of several key endpoints in male
CD rats exposed to doses of 0.2, 2, 20, and 40 mg/kg-day PFOA for 14 days.  These endpoints
included liver weight, hepatic p-oxidation, hepatic aromatase (P450 19A1), and hepatic total
cytochrome P450. The NOEL for these endpoints was found to be 0.2 mg/kg-day with significant
changes observed at > 2 mg/kg-day for all endpoints. Thus, the studies of Palazzolo (1993) and
Liu et al. (1996) demonstrate evidence of peroxisome proliferation at doses close to and below
doses that result in liver adenomas following chronic exposures (1.5 and 15 mg/kg-day).

Biegel et al. (2001) examined the temporal relationship between relative liver weights, hepatic P-
oxidation, and hepatic cell proliferation, and hepatic adenomas were evaluated in CD rats
following PFOA exposure for 1, 3, 6, 9,12,15,18, 21, and 24 months. Relative liver weights and
hepatic p-oxidation were increased at all time points. Hepatic cell proliferation was numerically
increased relative to the pair-fed control at 9,15,18, and 21 months. The liver endpoints (weight,
P-oxidation, and cell proliferation) were all elevated well before the first occurrence of liver
adenomas, which occurred after 12 months of treatment.

In summary, these data clearly demonstrate that PFOA induces liver toxicity and adenomas via a
PPARa agonist MOA in rats.  PFOA activates the PPARa and the requisite dose-response and/or
temporal associations of the key events for the PPARa mode of action with the liver adenomas
have been characterized.

3.9.3.2 Human Relevance of the Rat PPARa-agonist Induced Liver Toxicity and Liver
Adenomas

There has been substantial scientific interest regarding the role of peroxisome proliferation in
rodent hepatocarcinogenesis and its relevance for human carcinogenesis. Several scientific
groups have examined the state of the science on PPARa agonist-induced rodent liver tumors
over the years. In 1995, a workgroup convened under the auspices of the International Agency
for Research on Cancer concluded that the MOA for liver tumors induced in rodents by PPARa
agonists is unlikely to be operative in humans (IARC, 1995). The participants of a workshop held
in 1998 under the auspices of the International Life Sciences Institute Health and Environmental
Sciences Institute concluded that although it appeared unlikely that PPARa agonists could induce
liver tumorigenesis in humans, the possibility could not be ruled out (Cattley et al. 1998).  A
recent analysis by Klaunig et al. (2003) concluded that: (1) the weight of evidence in linking
PPARa to the mode of carcinogenic action of PFOA is high for the liver; (2) the PPARa MOA is
plausible in humans since the PPARa (hPPARa) is, present in humans and human livers possess
PPARa at sufficient levels to mediate the hypolipidaemic response to therapeutic fibrate drugs,
many of which are PPARa agonists; and (3) the weight of evidence, however, suggests that this
MOA is unlikely to occur in humans based on quantitative differences in several of the key
factors. For instance, human livers have been found to have 10-fold less mRNA for PPARa
compared with the rodents (Palmer et al., 1998; Tugwood et al., 1998). A recent study using a
PPARa-humanized mouse in which the human PPARa was expressed in the mouse liver has
demonstrated that following activation by the potent ligand Wy-14643, the PPARa-mediated
pathways controlling lipid metabolism are independent from those controlling the cell
proliferation pathways, and suggested that structural differences between human and mouse
PPARa may be responsible for the differential susceptibility to liver tumor development of
fibrates (Cheung et al., 2004).
                                          80

-------
SAR RPVJPW Draff? Tin Nnf fitp nr Onntp
There are several additional pieces of evidence on PFOA specifically which also suggest that this
MO A is not operative in non-human primates or humans. Maloney and Waxman (1999)
demonstrated that PFOA activated the mouse PPARa to a greater extent than the human PPARa
suggesting that humans would be less responsive to PFOA than rodents. In addition, there was no
indication of hepatic cell proliferation in a 6-month cynomolgus monkey study (Butenhoff et al.,
2002). Although limited, available human data have not shown an association between PFOA
exposure and liver effects.

On the question of the relevance of this mode of action to humans, the majority of the SAP Panel
(FIFRA SAP, 2004) agreed that there are relevant data indicating that humans are less sensitive
than rodents to the hepatic effects of PPARa agonists, and concurred with the premise that when
liver tumors are observed in long term studies in rats and mice, and 1) the data are sufficient to
establish that the liver tumors are a result of a PPARa agonist MOA and 2) other potential MO As
have been evaluated and found not to be operative, the evidence of liver tumor formation in
rodents should not be used to characterize potential human hepatocarcinogenic hazard.

3.9.3.3 Leydig Cell Adenomas in Rats

A series of mechanistic studies have been conducted (Cook et al.,  1994; Biegel et al., 1995; Liu et
al., 1996) to investigate the MOA associated with LCT in male Sprague-Dawley rats exposed to
PFOA. Two MO As have been proposed. One involves inhibition of testosterone biosynthesis,
leading to an imbalance of androgen/estrogen levels.  This leads to an increase in leutinizing
hormone (LH) which promotes the development of LCT. The second MOA involves an increase
in serum estradiol levels via induction of hepatic aromatase activity. Estradiol stimulates the
production of growth factors such as the transforming growth factor a (TGF a) which induces
Leydig cell proliferation.

Administration of PFOA to adult male rats by gavage for 14 days  was shown to decrease
testosterone levels and increase serum estradiol levels (Cook et al., 1994). These endocrine
changes correlate with its potency to induce LCTs in rats and were hypothesized to play a role in
the PFOA-induction of LCTs (Biegel et al., 2001). Other PPARa agonists (e.g., DEHP,
clofibrate) have been shown to increase serum estradiol levels in adult male rats (Eagon et al.,
1994; Rao et al., 1994). Eight out of eleven other peroxisome proliferators have been shown to
increase estradiol production using isolated LCs (Liu et al., 1996). Collectively, these data
suggest that many PPARa agonists can increase estradiol levels.  It was postulated that the
elevated estradiol levels may cause Leydig cell hyperpalsia and tumor formation by acting as a
mitogen and/or enhancing growth factor secretion; the transforming growth factor a (TGF a),
which binds to the epidermal growth factor (EOF) receptor and stimulated cell proliferation, for
instance, has been detected in Leydig cell (Teerds et al., 1990).

Subsequent experiments have shown that PFOA increased the levels of estradiol by inducing
CYP19 (aromatase), which converts testosterone to estradiol. Peroxisome proliferators are
known to induce a-oxidation and cytochrome P-450 monooxygenases by binding to the
peroxisome proliferation activation receptor a (PP AR a; a subfamily of steroid hormone
receptors).  It is possible that PFOA induces cytochrome P450 XIX (aromatase) by binding  to
and activating the PPAR a.

PFOA has also been shown to directly inhibit testosterone production when incubated with
isolated LCs, while ex vivo studies demonstrated that this inhibition was reversible (Biegel et al.,
1995). This inhibition of testosterone biosynthesis appears to be mediated by PPARa (Gazouli et


                                          81

-------
                      N»t
al., 2002) and may contribute to the development of LCT through disruption of the hypothalamus
pituitary thyroid axis. Testosterone, which is synthesized and secreted by the Leydig cells, is
regulated by LH; testosterone and LH form a closed-loop feedback system in the HPT axis. In
order to maintain adequate testosterone plasma levels, reduced testosterone levels is expected to
lead to increased LH levels through the negative feedback mechanism. In a mechanistic bioassay
with PFOA, serum testosterone and LH levels were not significantly altered at the levels of PFOA
that were tested (Biegel et al., 2001).  It has been pointed out that increases in LH may not always
be seen in all studies of chemicals for which the proposed mode of action calls  for elevated LH,
and that compensation may have occurred to restore homeostasis and inappropriate timing of
sampling are some of the explanations for failing to detect changes in LH levels (Clegg et al.,
1997).

Thus, the above data demonstrate that the induction of LCTs by  PFOA may be attributed to a
hormonal mechanism whereby PFOA either inhibits testosterone biosynthesis and/or increases
serum estradiol levels via induction of hepatic aromatase activity.  A critical review of these
mode of action data by Klaunig et al. (2003) led to the conclusion that the evidence is inadequate
at this time  to support a linkage between PPARcc agonism and induction of LCT. A similar
conclusion was made by the FIFRA SAP (FIFRA SAP, 2004). The Panel recognized that only
limited evidence support a link between Leydig cell tumors induction and a PPARa agonist mode
of action, the SAP concurred with the conclusion that chemicals in this subclass that induce
Leydig cell tumors may pose a carcinogenic hazard for humans.

Although the LCT induced in the rat by PFOA are assumed to be relevant to humans, they
probably do not represent a significant cancer hazard for humans.  Testicular tumors comprise
about 1% of all human male neoplasms. LCT are rare as they account for only 1-3% of all
testicular tumors in man (cited in: Prentice and Meikle, 1995). As there is some evidence that
PFOA induced LCT in rats either by inhibiting testosterone biosynthesis and/or by inducing
aromatase and thereby increasing estradiol levels, both the rat and human hypothalamic-pituitary-
testicular axis may respond to inhibition of testosterone with a subsequent increase in LH. Hence,
if PFOA inhibited testosterone biosynthesis in humans an increase in LH is expected to occur.
However, a 6-month cynomolgus monkey study with PFOA did not demonstrate any compound-
related effects on testosterone or estradiol (Butenhoff et al., 2002). These data suggest that PFOA
is unlikely to induce LCT in humans because humans are quantitatively less sensitive than rats to
LH stimulation (Cook et al., 1999). The number of LH receptors per Leydig cell is approximately
1,500 in man and 20,000 in the rat.  Furthermore, the rat appears to be unique in possessing the
Leydig cell luteinising hormone releasing hormone (LHRH) receptors which have the same
effects on Leydig cells and are not present in man, monkey  or mouse (cited in:  Prentice and
Meikle, 1995). The more compelling evidence supporting that testosterone inhibition or estradiol
induction is unlikely to pose a significant cancer hazard/risk to humans may be the human disease
state FMPP where men have a mutated LH receptor that is activated throughout their life and
LCT are not seen (Cook et al., 1999).

3.9.3.4 Pancreatic Acinar Cell Tumors  in Rats

The mechanism by which PFOA induced pancreatic acinar cell tumors is unknown. A number of
other peroxisome proliferates also produce pancreatic acinar cell tumors in rats.  The
development of pancreatic acinar cell hypertrophy, hyperplasia,  and adenomas  in the rat have
been shown to be modified by several factors such as steroids (testosterone and estradiol), growth
factors such as cholecystokinin (CCK), growth factor receptor over-expression (CCKA receptor)
and diet (fat) (Longnecker, 1983; Longnecker, 1987; Longnecker and Sumi, 1990). These


                                          82

-------
SAR RPVIPW Draft; f>n Nnt rite nr Onnfp
potential mechanisms have been investigated in a series of in vitro and in vivo (subacute,
subchronic, and oncogenicity) studies using PFOA (Obourn et al., 1997a; Oboum et al., 1997b).
The available data suggest that PFOA appeared to induce PACTs by an indirect trypsin-inhibition
mechanism where reduced bile flow and/or changes in bile composition produced an increase in
CCK levels secondary to hepatic cholestasis; a sustained increase in CCK levels is responsible for
the development of PACTs in rats (Cook et al., 1994; Oboum et al., 1997). CCK is a growth
factor that has been shown to stimulate normal, adaptive, and neoplastic growth of pancreatic
acinar cells in rats (Longnecker, 1987).  Accordingly, the PACT induced in rats by PFOA (and
other PPARa agonists) may be secondary to the liver effects of PPARa agonism.  However, the
mechanism(s) by which PFOA may induce PACT is based primarily on data with WY14643, a
more potent pancreatic acinar cells carcinogen.

The available MOA data of PFOA and other PPARa agonists which induced PACT in rats was
recently reviewed by Klaunig et al. (2003), and it was concluded that the evidence is inadequate
at this time to support  a linkage between PPARa agonism and induction of PACT. A similar
conclusion was made by the FIFRA SAP (FIFRA  SAP, 2004). The  SAP concurred that chemicals
in this subclass that induce pancreatic acinar cell tumors may pose a carcinogenic hazard for
humans.

While there are some evidence that a sustained increase in CCK levels is responsible for the
development of PACTs in rats by PFOA, expressions of PPARa and CCKA receptors in humans
are much lower than rodents. In addition, humans regulate pancreas exocrine secretion via a
neuronal pathway rather than direct binding of CCK to acinar CCKA receptors as in rodents.
Consistent with this conclusion, a 6-month cynomolgus monkey study with PFOA did not
demonstrate any compound-related effects on CCK levels or clinical pathology evidence of
cholestasis  (Butenhoff et al., 2002). Study  of CCK levels in employees also did not report
increases in workers (Olsen, et al., 1998b; Olsen et al., 2000). Furthermore, while the pancreatic
tumors in the rat are typically derived from acinar cells, the majority of human pancreatic
neoplasms are of the ductal type (Cotran et al.,1989). Therefore, the PACT induced in the rat by
PFOA probably do not represent a significant cancer hazard for humans.

3.9.3.5 Cancer Descriptor

Carcinogenicity studies in Sprague-Dawley (CD) rats show that PFOA induces a "tumor triad"
similar to a number of other PPARa agonists.  This "tumor triad" includes liver tumors, Ley dig
cell tumors, and pancreatic acinar cell tumors. The evidence for mammary fibroadenomas in the
female rats is equivocal since the incidences were comparable to some historical background
incidences. In humans, occupational studies of still relatively young cohorts have not indicated
statistically significant increases in these types of cancer. Not only  were the numbers of cancer
deaths very small in some of the cancer categories, but PFOA exposures were also not adequately
characterized.

As summarized in section 3.9.3.1, there is sufficient evidence to conclude that PFOA is a PPARa-
agonist and that PFOA induces liver toxicity and adenomas via a PPARa agonist MOA in rats.
PFOA activates the PPARa and the requisite dose-response and/or temporal associations of the
key events for the PPARa mode of action with the liver adenomas have been characterized.  A
recent ILSI workgroup as well as the FIFRA SAP have concluded that this MOA is unlikely to
occur in humans based on quantitative differences in several of the key factors (Klaunig et al.,
2003; OPPTS, 2003; FIFRA SAP, 2004).
                                         83

-------
SAR RPVJPW Draft; Hn Nnt Tito nr
The modes of carcinogenic action of PFOA-induced LCTs and PACT have not been fully
elucidated. The induction of LCT by PFOA may involve a hormonal mechanism whereby PFOA
either inhibits testosterone biosynthesis and/or increases serum estradiol levels via induction of
hepatic aromatase activity. The induction of PACT are related to an increase in serum level of the
growth factor, CCK, that appears to be secondary to changes in the liver. As the modes of
carcinogenic action of PFOA-induced LCT and PACT have not been clearly linked to PPARa
agonism, it is assumed that these tumors induced in rats are relevant to humans. However, the
LCT and PACT induced in the rat by PFOA probably do not represent a significant cancer hazard
for humans because of quantitative differences in the expressions of LH and CCKA receptors and
of other toxicodynamic differences between the rat and the human.  For instance, the number of
LH receptors per Ley dig cell is approximately 1,500 in man and 20,000 in the rat. Furthermore,
the rat appears to be unique in possessing the Ley dig cell luteinising hormone releasing hormone
(LHRH) receptors which have the same effects on Ley dig cells and are not present in man,
monkey or mouse (cited in: Prentice and Meikle, 1995).  While there is some evidence that a
sustained increase in CCK levels is responsible for the development of PACTs in rats by PFOA,
expressions of PPARa and CCKA receptors in humans are much lower than rodents. In addition,
humans regulate pancreas exocrine secretion via a neuronal pathway rather than direct binding of
CCK to acinar CCKA receptors as in rodents; the pancreatic tumors in the rat are typically derived
from acinar cells whereas the majority of human pancreatic neoplasms are of the ductal type
(Cotranetal.,1989).

Overall, based on no adequate human studies and uncertain human relevance of the tumor triad
(liver, Leydig cell and pancreatic acinar cell tumors) from the rat studies, PFOA may be best
described as "suggestive evidence ofcarcinogeniciiy, but not sufficient to assess human
carcinogenic potentiarunder the draft 1999 Guidelines for Carcinogen Risk Assessment (U.S.
EPA, 1999).

3.9.4 Toxicity in Adult Repeat-Dose Animal Studies

3.9.4.1 Non-Human Primates

A summary of the toxicity associated with exposure to APFO in adult non-human primates is
presented in Table 21. In both the Rhesus and cynomolgus monkey studies, there was a very
steep dose-response curve for mortality.  The cause of the mortality  is unknown. The Rhesus
monkeys also exhibited several clinical indicative of gastrointestinal distress; the mode of action
for these effects is unknown, but the gastrointestinal effects may be  due to the potent surface
activity properties of APFO. The gastrointestinal effects were not noted in a 6 month study of
cynomolgus monkeys. The lack of gastrointestinal effects may be due to a species difference, or
due to the fact that the monkeys were dosed via a capsule versus gavage in the Rhesus monkey
study. In addition, liver weights were significantly increased in the cynomolgus monkey, but
were unaffected in the Rhesus monkey. The increased liver weight does not appear to be due to
peroxisome proliferation as there was only a minimal increase in palmitoyl CoA oxidase in the
high dose group.  However, there was evidence of mitochondria! proliferation which also
suggests that a mode of action other than PPARa-agonism is operative in this species.  It should
also be noted that a number of hormonal and cell proliferation assays were included in the
cynomolgus monkey study to determine whether there was any indication of a response of
hepatic, Leydig cells and pancreatic acinar cells to PFOA exposure which have been observed in
the rat. There was no indication of cell proliferation in the liver, testes or pancreas, nor were
hormonal measures affected.  There is significant uncertainty associated with the derivation of a
LOAEL and NOAEL for both of these studies due to the small sample size, the variability in


                                          84

-------
SAR RPVJPW Draff; Tin Nnt fito ni-
response of the individual monkeys, and the lack of linear proportionality between administered
dose and serum PFOA levels.

                                        Table 21
                        Summary of Toxicity in Nonhuman Primates

Monkey

Rhesus
2/sex/dose




Cynomolgus
4-6
males/dose









Exposure
Period
13 weeks





6 months











Dose
gavage
0,3,10,30,
lOOmg/kg-day



oral capsule
0, 3, 10, 20/30
mg/kg-day










Response
> 3 - clinical signs
(including G.I)

> 30 - mortality, 1
body weight, clinical
signs
3 - moribund
(sacrificed), clinical
signs

> 3 - 1 absolute and
relative liver weight
20/30 - moribund
(sacrificed), clinical
signs, treatment
stopped, ibody
weight,
LOAEL/
NOAEL
(mg/kg-day)
LOAEL = 3

no NOAEL



LOAEL = 3
NOAEL =
ND









Reference
Goldenthal,
1978b




Thomford,
2001b;
Butenhoff et
al., 2002







3.9.4.2 Adult Male Rats

A summary of the systemic toxicity associated with APFO exposure in adult male rats is
presented in Table 22. The major target organ is the liver. As described in section 3.9.2.1, the
liver toxicity is due to a PPARoc-agonist mode of action which is unlikely to occur in humans.
Aside from the liver toxicity, APFO exposure is not associated with many other effects. Body
weight was reduced at doses of 6.5 mg/kg-day in a 90-day study (Palazzolo, 1993); at > 3 mg/kg-
day in the FO males and > 1 mg/kg-day in the Fl males in the two-generation reproductive
toxicity study (York, 2002; Butenhoff et al., 2004); and at 14.2 mg/kg-day in  the 2-year study
(Sibinski, 1987). Adrenal hypertrophy and vacuolation was noted in the FO adult males at doses
> 10 mg/kg-day and in Fl adult males at 30 mg/kg-day in the two-generation reproductive
toxicity study (York, 2002; Butenhoff et al., 2004).  There is some uncertainty as to the biological
significance of this finding since this effect was  not observed in a 90-day study (Goldenthal,
1978a) or a 2-year study (Sibinski, 1987).  Exposure to APFO was associated with a reduction in
the number of erythrocytes and related parameters following exposure to 14.2 mg/kg-day  APFO
in the 2-year study (Sibinski, 1987).  The testicular masses and testicular vascular mineralization
observed in the 2-year study are probably associated with the etiology of the Ley dig cell
adenomas that were discussed in section 3.9.4.3. The mode of action is unknown for any  of these
effects.
                                           85

-------
SAR RPVJPW Draft; fin Nnf Titp nr Dnnfp
                                       Table 22
                     Summary of Systemic Toxicity in Adult Male Rats

Rat
Strain
CD






Sprague-
Dawley




Sprague-
Dawley














Sprague-
Dawley










Sprague-
Dawley



Exposure
Period
13 weeks






1 3 weeks





FO animals
-15 weeks






Fl animals
-19-20 weeks






2 years











2 years





Dose
diet -0,10,
30, 100, 300,
lOOOppm

0.056, 1.72,
5.64, 17.9,
63.5 mg/kg-day
diet -0,1, 10,
30, 100 ppm

0.06,0.64,1.94,
6.50 mg/kg-day

gavage -
0,1,3,10,
30 mg/kg-day













diet - 0, 30,
300 ppm

1.3,14.2
mg/kg-day







diet-
0, 300 ppm

14.2 mg/kg-day

Effects
(mg/kg-day)
> 1.72- [absolute liver
weight
> 5.64 - 1 hepatocellular
hypertrophy
>17.9- 1 relative liver
weight
>63.5- 1 body weight
> 0.64 - 1 absolute and
relative liver weight,
hepatocellular
hypertrophy,
tpalmitoyl CoA oxidase
6.5 - 1 body weight
> 1 - t absolute and
relative liver
> 3 - 1 body weight
> 10 - adrenal
hypertrophy and
vacuolation
30 - clinical signs,
1/30 moribund
>1 - t absolute and
relative liver weight,
1 body weight
>3 - hepatocellular
hypertrophy
30 - adrenal
hypertrophy and
vacuolation
14.2- 1 relative liver
weight, hepatocellular
hypertrophy, cystoid
degeneration, ALT;
I body weight gain;
lerythrocytes,
hematocrit,
hemoglobin
concentration;
testicular vascular
mineralization,
testicular masses
14.2- t relative
liver weight


LOAEL/
NOAEL
(mg/kg-day)
LOAEL
= 1.72

NOAEL
= 0.56


LOAEL
= 0.64

NOAEL
= 0.06

LOAEL
=1

no NOAEL




LOAEL
=1

no NOAEL




LOAEL
= 14.2

NOAEL
= 1.3







LOAEL
= 14.2

no NOAEL


Reference
Goldenthal,
1978a





Palazzolo,
1993




York, 2002;
Butenhoff
et al, 2004













Sibinski,
1987










Cook et al.,
1994;
Biegel et al.,
2001
                                          86

-------
SAR RPVJPW Draff; Tin Nnt fite nr
3.9.4.3 Adult Female Rats

A summary of the systemic toxicity associated with APFO exposure in the adult female rat is
shown in Table 23. Due to the gender difference in elimination of PFOA in rats, the response of
the female rat to APFO exposure is generally far less than that of the male rat. The effects on the
liver were only observed at relatively high doses (Goldenthal, 1978a).  The effects on the adrenal
that were observed in the male rat were not observed in the female rat in the two-generation
reproductive toxicity study (York, 2002; Butenhoff et al., 2004). Body weight was reduced at
doses of 30 mg/kg-day in the Fl females in the two-generation reproductive toxicity study (York,
2002; Butenhoff et al., 2004), and at 16.1 mg/kg-day in the 2-year study (Sibinski, 1987).
Exposure to APFO was associated with a reduction in the number of erythrocytes and related
parameters following exposure to 16.1 mg/kg-day APFO in the 2-year study (Sibinski, 1987).
The mode of action is unknown for any of these effects.

                                        Table 23
                    Summary of Systemic Toxicity in Adult Female Rats
Rat
Strain
CD
Sprague-
Dawley
Sprague-
Dawley
Exposure
Period
13 weeks
FO animals
-18 weeks
Fl animals
-18 weeks
2 years
Dose
diet
0, 10, 30, 100,
300, 1000 ppm
0.74, 2.3, 7.7,
22.4, 76.5
mg/kg-day
gavage
0, 1, 3, 10, 30
mg/kg-day
diet
0, 30, 300 ppm
1.6, 16.1
mg/kg-day
Effects
(mg/kg-day)
76.5 - 1 absolute and
relative liver weight
No effects
30 - i body weight
16.1 - J body weight
gain, 1 erythrocytes,
hematocrit, hemoglobin
concentration
LOAEL/
NOAEL
(mg/kg-day)
LOAEL
= 76.5
NOAEL
= 22.4
no LOAEL
NOAEL
= 30
LOAEL
= 30
NOAEL
= 10
LOAEL
= 16.1
NOAEL
= 1.6
Reference
Goldenthal,
1978a
.York, 2002;
Butenhoff
et al., 2004
Sibinski,
1987
3.9.4.4 Adult Mice

In mice, a 28 day study demonstrated that the liver is also a target organ. As described above
(section 3.9.3), the liver toxicity is due to a PPARa-agonist mode of action which is unlikely to
occur in humans.  In addition, several studies have shown that PFOA affects the immune system
in mice. Feeding C57B1/6 mice a diet containing 0.02% PFOA resulted in adverse effects to both
the thymus and spleen (Yang et al. 2000, 2001).  In addition, this feeding regimen resulted in
                                           87

-------
SAR Rovipw Draff; Tin Nnf rite nr Oiintp
suppression of the specific humoral immune response to horse red blood cells, and suppression of
splenic lymphocyte proliferation in response to LPS and ConA (Yang et al., 2002b). The
suppressed mice recovered their ability to generate a humoral immune response when they were
fed a diet devoid of PFOA.  Studies using transgenic mice showed that the PPARa was involved
in causing the adverse effects to the immune system (Yang et al., 2002a). Further research is
needed to determine the exact role of PPARa and immunotoxicity in mice.

3.9.5 Developmental and Reproductive Toxicity in Animal Studies

The developmental and reproductive effects associated with oral exposure to APFO in animal
studies are summarized in Table 24.  In New Zealand White rabbits, there was evidence of
prenatal developmental toxicity following exposure to 50 mg/kg-day APFO on gestation days 6-
18 (Gortner, 1982). In contrast, there was no evidence of developmental toxicity following
exposure to doses as high as 150 mg/kg-day APFO during gestation days 6-15 in Sprague-Dawley
rats (Gortner, 1981; Staples, 1984). However, there was evidence of developmental/reproductive
toxicity during the postnatal period in a two generation reproductive toxicity study in Sprague-
Dawley rats (York, 2002; Butenhoff et al., 2004).  In this study, During  lactation, there was a
reduction in Fl mean body weight on a litter basis during lactation (sexes combined) in the 30
mg/kg-day group.  Fl males in the 10 and 30 mg/kg-day groups exhibited a significant reduction
in body weight gain during days 8-50 postweaning, and body weights were significantly reduced
in the 10 mg/kg-day group beginning on postweaning day  36, and in the 30 mg/kg-day group
beginning on postweaning day 8. Fl females in the 30 mg/kg-day group exhibited a significant
reduction in body weight gain on days 1-15 postweaning, and in body weights beginning on day 8
postweaning. There was a significant increase in mortality mainly during the first few days after
weaning, and a significant delay in the timing of sexual maturation for Fl males and females in
the 30 mg/kg-day group. For Fl males, the LOAEL for developmental/reproductive toxicity was
10 mg/kg-day, and the NOAEL was 3 mg/kg-day. For Fl  females, the LOAEL for
developmental/reproductive toxicity was considered to be  30 mg/kg-day, and the NOAEL was  10
mg/kg-day. No effects were observed in the F2 pups. However, it should be noted that the F2
pups were sacrificed at weaning, and thus it was not possible to ascertain the potential post-
weaning effects that were noted in the Fl generation.

-------
SAR RPVJPW Draft; Tin Nnf Pitp nr
                                       Table 24
               Summary of Developmental and Reproductive Toxicity Effects

Species
New Zealand
White rabbit


Sprague-
Dawley
rat
Sprague-
Dawley
rat
Sprague-
Dawley
rat
Sprague-
Dawley
rat
















Dose
(mg/kg-day)
0, 1.5,
5,50


0, 0.05,
1.5,5,150

0,100


0,100


0, 1, 3,
10,30

















Exposure
Duration
GD 6-18



GD 6-15


GD6-15



GD 6-15

Fl - prior to
conception
through
sexual
maturation










F2 - prior to
conception
until
weaning
Time
Assessed
GD29



GD20


GD20



PND35

Variety















Variety



Effects
T 13th rib



none


none


none


>10-
Ipostweaning
body weight
in males

30-
ipreweaning
litter body
weight ,
tpostweaning
mortality,
delayed sexual
maturation,
Ipostweaning
body weight
in females

none

LOAEL/
NOAEL
LOAEL
= 50
NOAEL
= 5
no LOAEL
NOAEL
= 150
no LOAEL
NOAEL
= 100
no LOAEL
NOAEL
= 100
LOAEL
= 10
NOAEL
= 3












no LOAEL
NOAEL
= 30

Reference
Gortner, 1982



Gortner, 1981


Staples, 1984


Staples, 1984


York, 2002;
Butenhoffet
al, 2004
















                                          89

-------
SAB RPVJPVV Draft; fin Nnf Tito nr Oimte
4.0 Biomonitoring Data

PFOA has been measured in human serum of workers occupationally exposed to APFO and in the
general U.S. population. In general, serum PFOA concentrations in the general population are
much lower than in workers exposed to APFO (see Tables 25-27 below). However, it should be
noted that the highest levels reported to date in the general population are similar to some of the
lowest levels in workers exposed to PFOA occupationally. The environmental concentrations of
APFO and the pathways of exposure to the general population are not known.

4.1 Occupational Exposures

3M and DuPont have been the primary producers and users of perfluorinated compounds in the
U.S. Both companies offer voluntary medical surveillance programs to workers at plants that
produce or use perfluorinated compounds.  3M discontinued manufacturing PFOA between 2000-
2002. PFOA serum levels seem to be decreasing slightly at both the 3M and DuPont facilities;
however, the cross-sectional nature of the data cannot provide trends in serum concentrations.
PFOA blood serum data for volunteers in the medical surveillance programs at these plants that
have been submitted to EPA are presented below in ug/mL (ppm).

4.1.1  3M Occupational Data

3M has been offering voluntary medical surveillance to workers at plants that produce or use
perfluorinated compounds since 1976. Serum PFOA levels in 3M workers have been measured
since 1993 (Olsen et al., 2003a; 2003b; 2003c; 2001a; 2001b; 2001c; 2000; 1999; 1998c). Prior
to this time, analytical capabilities precluded the accurate measurement of any specific
fluorochemical analyte and only total organic fluorine was measured. PFOA analysis differed
slightly in each surveillance program year and different laboratories were used to assay PFOA in
each year. The samples were analyzed for PFOA using high performance liquid chromatography
mass spectrometry (HPLC/MS), but the extraction methods differed slightly each year (Olsen, et
al., 2001f; Olsen, et al., 2003a; 2003b).

Serum PFOA concentrations for workers participating in 3M's biomonitoring program have been
reported for 3 plants: Cottage Grove, Minnesota; Decatur, Alabama;and Antwerp, Belgium.
Only the U.S. data are presented here.  Surveillance years included 1993,1995,1997,1998, 2000,
2002, and 2003, although not all of the plants offered surveillance in all of these years.
Participation in all of the program years was voluntary.  The eligible voluntary participation rates
ranged from approximately 70% in 1993 to 50% in 1997 (Olsen, et al., 2001f).  The number of
employees volunteering for participation increased greatly in 2000 and then declined by about
75% in 2002 (Olsen, et al., 2003a; 2003b) (see Table 24 below).  3M suggests that the increase in
2000 was probably due to increased employee awareness of the persistence and prevalence of
perfluorinated chemicals in human tissues. Meanwhile, the decline in 2002 was probably due to
3M's announcement to phase out PFOA production at their plants.  In 1998, a random sample of
Decatur employees was surveyed to determine whether the voluntary nature of the program was
biasing the  results of the surveillance (Olsen et al., 1999). The data presented in Table 24
indicate that there is  little difference in male serum concentrations between 1997 (1.40 |ag/mL,
when participation was voluntary) and 1998 (1.735 ug/mL, when a random sample was taken).

Gender-based data for 3M's biomonitoring program are limited due to the small number of
women participating in the program.  Mean PFOA serum concentrations for female employees at


                                          90

-------
SAB RPVIPW flraff? f>n Nnf Tito nr
all 3 of the plants were lower than those of male employees, even in similar jobs.  Where data are
available by gender, they are provided in Table 25.
                                       Table 25
                  Summary of 3M U.S. Occupational PFOA Biomonitoring
                             (serum concentrations, |4g/mL)*


Plant
Cottage Grove
2002 (n=38)
2000 (n= 148)
Male (n= 131)
Female (n= 17)
1997(n=74)
1995(n = 80)
1993 (n= 111)
Decatur
2002 (n =54)
2000 (n = 263)
Male (n = 215)
Female (n = 48)
1998(n=126)
Male (n =102)
Female (n=24)
1997(n = 84)
(males only)
1995(n = 90)
(males only)
Building 236
(fluorochemical
research) 2000
(n = 45)
Arith-
metic
Mean

4.3
NR
4.51
0.85
6.4
6.8
5.0

1.497
1.78
1.90
1.23
1.54
1.735
0.691
1.40

1.72


0.106



Range


0.07 - 32.6
NR
.007-92.03
.04-4.73
0.1-81.3
0-114.1
0 - 80.0

0.025-4.81
0.04-12.70
0.04-12.70
NR
0.02 - 6.76
0.02 - 6.76
NR
NR

NR


0.008-
0.668

Geometric
Mean


1.7
NR
0.85
0.42
NR
NR
NR

0.713
1.13
NR
NR
0.90
1.142
0.326
NR

NR


0.053


95% CI of
Geometric
Mean

1.02-2.72
NR
0.64- 1.22
0.23 - 0.79
NR
NR
NR

0.483-1.055
0.99-1.30
NR
NR
0.72-1.12
NR
NR
NR

NR


0.037 -0.076



Reference


Olsen, et al., 2003a
Olsen, et al., 2003f


Olsen, et al., 2000
Olsen, etal., 2000
Olsen, et al., 2000

Olsen, et al., 2003b
Olsen, et al., 200 la
Olsen et al., 2003 e
Olsen et al., 2003 e
Olsen, etal., 1999


Olsen, etal., 1998c

Olsen, etal., 1998c


Olsen, etal., 2001 c


NR- not reported

Of the plants listed above, PFOA exposures are highest at the Cottage Grove plant.
Correspondingly, PFOA serum levels are highest at this plant (geometric mean =1.7
range, 0.07 - 32.6 ng/mL). Mean serum PFOA levels increased slightly at both the Cottage
Grove and Decatur plants from 1993 to 2000; however, the latest serum measurements (2002)
show  a slight decline in mean PFOA concentrations. However, it should be noted that the
number of participants in the 2002 sampling period was comparatively small. In addition, all of
the data are cross-sectional and therefore cannot provide any temporal information.
                                          91

-------
SAR RPVJPW Draff! T)n Nnt fifp ni- Oiinfp
4.1.2 DuPont Occupational Data

Dupont has been measuring PFOA in workers' blood serum since 1981 (Dupont, 2001 a, 2001b).
Prior to 1981, total blood fluoride levels were reported. All of Dupont's operations in the U.S.
that use PFOA with significant exposure potential are concentrated at the Washington Works
plant in Washington, West Virginia.  PFOA is used as a fluoropolymer reaction aid at this plant.
The data presented in Table 26 reflect serum concentrations of volunteer workers in the plant who
had potential PFOA exposure.  With the exception of the 2000 sampling period, the same workers
were common to all of the other sampling periods, with additional workers participating in 1995.
The data available to EPA at this time are limited to what is presented here.

                                       Table 26
                  Summary of Dupont Occupational PFOA Biomonitoring
                             (serum concentrations, ug/mL)*
Year
Washington Works
2000 (n= 72)
1995(n=80)
1 989-90 (n =22)
1985(n = 22)
1984 (n= 19)
Arithmetic
Mean
1.53
1.56
1.96
2.34
3.21
Range
0.02 - 9.0
0.12-4.5
0.06-11
0.06-18
0.07 - 24
Reference
Dupont, 2001 a
Dupont, 2001 b
Dupont, 2001b
Dupont, 2001b
Dupont, 2001 b
" ug/mL = ppm
4.2 General Population Exposures

Data on PFOA levels in the general population include both pooled and individual serum
samples.  Mean serum PFOA levels are lower in the general population than in workers exposed
to PFOA (described above).  Individual serum PFOA levels in three separate US cohorts are very
similar.  The geometric mean for all three cohorts, comprised of 3 separate age groups, is
approximately 0.004 ug/mL (4 ppb). All of the data are fairly recent but are cross-sectional;
therefore, temporal trends cannot be established.  The available data are summarized in Table 27.
                                          92

-------
SAR RPVIPW Draff; Dn Nnt fifp nr
                                       Table 27
                   Summary of General Population PFOA Biomonitoring
                              (serum concentration, |ag/mLa)
Sample
Arithmetic
Mean
Range
Geometric
Mean
95% CI
Pooled Samples
Commerical sources of
blood, 1999 (n = 35 lots)
Blood banks, 1998(n =
18 lots, 340-680 donors)
0.003
0.017b
.001-
.013
0.012-
0.022
NA
NA
NA
NA
Individual Samples
American Red Cross
blood banks, 2000 (n =
645)
Elderly (65 - 96 years),
2000 (n = 238)
Children (2-12 years),
1995(n = 598)
0.0056
NR
0.0056
0.0019-
0.0523
0.0014-
0.0167
0.0019-
0.0561
0.0046
0.0042
0.0049
0.0043 -
0.0048
0.0039 -
0.0045
0.0047 -
0.0051
Reference

3M
Company,
1999a
3M
Company,
1999b

Olsen et al.,
2002a,
2003d
Olsen et al.,
2002b,
2004a
Olsen et al.,
2002c,
2004b
* ug/mL = ppm
'PFOA detected in about 1/3 of the pooled samples but quantifiable in only 2.
NR - not reported
Pooled blood samples from U.S. blood banks indicate mean PFOA levels of 0.003 to 0.017
ug/mL (3 -17 ppb) (3M Company, Feb. 5,1999; 3M Company, May 26,1999). The highest
pooled sample reported was 0.022 ug/mL (22 ppb).  Pooled samples were collected in 1998 and
1999. However, it cannot be assumed that these levels can be generalized to the U.S. population
for several reasons: 1) blood donors are a unique group that does not necessarily reflect the U.S.
population as a whole, 2) many of the blood banks originally contacted for possible inclusion in
the study declined to participate, 3) only a small number of samples have actually been analyzed
for PFOA, and 4) no other data such as age, sex, or other demographic information are available
on the donors.

Individual blood samples from 3 different age populations were recently analyzed for PFOA and
other fluorochemicals using high-pressure liquid chromatography/electrospray tandem mass
spectrometry (HPLC/ESMS) (Olsen et al., 2002a, 2002b, 2002c). The studies' participants
included adult blood donors, an elderly population participating in a prospective study in Seattle,
                                          93

-------
SAR RPVJPW rir-aft; Dn Not Pifp ni-
WA, and children from 23 states participating in a clinical trial. Overall, the PFOA geometric
means were similar across all 3 populations (0.0046 ug/mL, 0.0042 ug/mL, and 0.0049 ug/mL,
respectively). The geometric means and 95% tolerance limits (the exposure below which 95% of
the population is expected to be found) and their upper bounds were comparable across all 3
studies.  However, the upper ranges for the children and adults were much higher than for the
elderly population. It is not clear whether this is the result of geographic differences in PFOA
levels or some other factor. It should be noted that PFOS and PFOA were highly correlated in all
three studies (r = .63, r = .70, and r = .75) and that PFOA did not meet the criteria for a log
normal distribution based on the Shapiro-Wilk test in any of the studies.  The authors suggest that
it may be due to the greater proportion of subjects with values less than the lower limit of
quantitation (LLOQ); however, only 12 of the 1481 total samples were below the LLOQ. In
those instances where a sample was measured below the LLOQ, the midpoint between zero and
the LLOQ was used for calculation of the geometric mean. The details of each study are
provided below.

Blood samples from 645 U.S. adult blood donors (332 males, 313 females), ages 20-69, were
obtained from six American Red Cross blood banks located in: Los Angeles, CA;
Minneapolis/St. Paul, MN; Charlotte, NC; Boston, MA; Portland, OR, and Hagerstown, MD
(Olsen et al., 2002a). Each blood bank was requested to provide approximately 10 samples per
10-year  age intervals (20-29, 30-39, etc.) for  each sex. The only demographic factors known for
each donor were age, gender, and location.

The geometric mean serum PFOA level was 0.0046 ug/mL. The range was 
-------
SAR RPVJPW Draft? f>n Nnf fite nr Quote
Male children had significantly (p<.01) higher geometric mean serum PFOA levels than females:
0.0052 ng/mL and 0.0047 u_g/mL, respectively. In simple linear regression analyses, age was
significantly (p < .05) negatively associated with PFOA in both males and females. When
stratified by age, the geometric mean of PFOA was highest at age 4 (0.0057 ng/mL) and lowest at
age 12 (0.0035 ng/mL).  Although the data were not reported, a graphical presentation of log
PFOA levels for each state by gender looked similar across the states. However, it is difficult to
interpret these data without analyzing them and the sample sizes were limited for each
gender/location subgroup.  PFOS and PFOA were highly correlated (r = .70) in this study. PFOA
and PFHS (perfluorohexanesulfonate) were also correlated, although not as strongly (r = .48).

The above 3 studies indicate similar geometric means and ranges of PFOA among sampled
adults, children, and an elderly population. However, an unexpected finding was the level of
PFHS and M570 (N-methyl perfluorooctanesulfonamidoacetate) in children (Olsen, 2002c,
2004b). These serum levels were much higher in the sampled  children than in the sampled adults
or elderly. It is not clear why this occurred, but it is probably due to a different exposure pattern
in children.

In another study, the PFOA concentration was analyzed in human sera and liver samples using
HPLC/ESMS (Olsen et al., 2001d).  Thirty-one donor samples were obtained from 16 males and
15 females.  The average age of the male donors was 50 years  (SD 15.6, range 5-69) and the
average age of the female donors was 45 years (SD 18.5, range 13-74). The causes of death were
intracranial hemorrhage (n = 16 or 52%), motor vehicle accident (n = 7 or 23%), head trauma (n =
4 or 13%), brain tumor (n = 2 or 6%), drug overdose (n = 1 or  3%) and respiratory arrest (n = 1 or
3%). Both serum and liver tissue were obtained from 23 donors; 7 donors contributed liver tissue
only and 1 donor contributed serum only.  Resulting serum values for PFOA ranged from < LOQ
(O.0030) - 0.0070 ug/mL.  Assuming the midpoint value between zero and LOQ serum value for
samples 
-------
SAR RPVIAW Draff; fin Nnt rite nr Quote
Studies in mice have shown an effect on the immune system. These mouse studies were not
considered for the calculation of the MOEs since information on serum levels of PFOA was not
available, nor are there pharmacokinetic data that would allow for the estimation of serum levels.

As summarized in section 3.9.5.1, studies of non-human primates did not establish aNOAEL.
The LOAEL for the 13-week study  in Rhesus monkeys was 3 mg/kg-day based on clinical signs
of toxicity (Goldenthal, 1978b) and the LOAEL for the 6-month study of cynomolgus monkeys
was also 3 mg/kg-day based on increased liver weight and possibly mortality (Thomford, 200 Ib;
Butenhoff et al., 2002). For calculation of MOEs, the cynomolgus monkey study was chosen
since it included a slightly larger sample size (N = 4 to 6), was of longer duration, and included
more robust serum analyses of PFOA.

Several studies have examined the toxicity of APFO in adult male and female rats which are
summarized in sections 3.9.5.2 and  3.9.5.3, respectively.  In adult male rats, the primary target
organ is the liver. However, as described in section 3.9.3, the proposed mode of action for PFOA
(PPARa-agonism) in rodents leading to the observed liver toxicity is unlikely to occur in humans,
and therefore was not considered for calculation of the MOEs. The only other consistent effect
noted across  studies was body weight reduction. Body weight was reduced at doses of 6.5 mg/kg-
day in a 90-day study (Palazzolo, 1993); at > 3 mg/kg-day in the FO males and > 1 mg/kg-day in
the Fl males in the two-generation reproductive toxicity study (York, 2002; Butenhoff et al.,
2004); and at 14.2 mg/kg-day  in the 2-year study (Sibinski, 1987). For adult female rats, there are
very few effects due to the rapid excretion of PFOA. The only consistent effect that was noted
among the studies was reduced body weight; this was observed at doses of 16.1 mg/kg-day in the
2-year study  (Sibinski, 1987) and at 30 mg/kg-day in the adult FO and Fl  females in the two-
generation reproductive toxicity study (York, 2002; Butenhoff et al., 2004). Although, the
significance of reduced body weight for human health is not clear, body weight was used for the
calculation of the MOEs simply because it defines the lowest effect level and therefore serves as a
conservative benchmark.  The lowest dose that was  associated with reductions in body weight in
males was in the Fl males in the two-generation reproductive toxicity study.  For the Fl males,
the LOAEL for body weight is 1 mg/kg-day, and a NOAEL was not established. The lowest dose
for females was in the 2-year study; the LOAEL for body weight was 16.1 mg/kg-day and the
NOAEL was 1.6 mg/kg-day.

Several studies have examined the developmental and reproductive toxicity associated with
exposure to APFO which are summarized in section 3.9.5.5.  Developmental effects were
observed in a prenatal developmental toxicity study in New Zealand White rabbits (Gortner,
1982).  However, this study was not considered for the calculation of the MOEs  since information
on serum levels of PFOA in rabbits was not available, nor are there pharmacokinetic data that
would allow for the estimation of serum levels in rabbits. There were no developmental effects
noted in prenatal developmental toxicity studies of Sprague-Dawley rats following exposure to
APFO during gestation days 6-15 (Gortner, 1981; Staples, 1984). Developmental effects were
noted in the Fl male and female pups in the two-generation reproductive toxicity study.  During
lactation, there was a reduction in Fl mean body weight on a litter basis (sexes combined) in the
30 mg/kg-day group.  Fl males in the 10 and 30 mg/kg-day groups exhibited a significant
reduction in body weight gain during days 8-50 postweaning, and body weights were
significantly  reduced in the 10 mg/kg-day group beginning on postweaning day 36,  and in the 30
mg/kg-day group beginning on postweaning day 8.  Fl females in the 30 mg/kg-day group
exhibited a significant reduction in body weight gain on days 1-15 postweaning, and in body
weights beginning on day 8 postweaning.  There was a significant increase in mortality mainly
during the first few days after weaning, and a significant delay in the timing of sexual maturation


                                          96

-------
SAR RPVIPW Draft; Fin Nnt fitp nr
for Fl males and females in the 30 mg/kg-day group. These endpoints were used for the
calculation of the MOEs.

5.2 Use of Serum Levels as a Measure of Internal Dose for Humans

5.2.1 General Population

A summary of the human serum levels of PFOA that were considered in the calculation of MOEs
is provided in Table 28. These data are derived from two biomonitoring studies performed on
large cohorts from various geographic areas of the US and include both children and adults. They
are described in detail in Section 4.2.  The arithmetic and geometric means, as well as the range,
95% confidence intervals for the geometric means, and the cumulative 90th percentiles are
displayed in order to provide a sense of the distribution of the data.  The frequency distributions
of the serum data appear to be log-normally distributed, although the criteria for the Shapiro-Wilk
test are not met. This may be due to the number of subjects with serum values less than the
LLOQ (n = 2 for <1.9 ng/ml and n = 48 for < 2.1 ng/ml). Gender specific data were available for
the geometric mean and range, but not for the arithmetic mean. These serum values are assumed
to represent steady state levels in the general US population.

                                       Table 28
             Summary of Levels of PFOA (ug/mL) in the Serum of Human Populations
Population
Adults (20 - 69 years, American
Red Cross blood banks, 2000)
Both genders (n = 645)
Males (n = 332)
Females (n = 3 13)
Children (2-12-years, 1995)
Both genders (n = 598)
Males (n = 3 00)
Females (n = 298)
Arith-
metic
Mean
0.0056
NR
NR
0.0056
0.0059
0.0052
90th
percentile
0.0094
0.0101
0.0084
0.0085
0.0090
0.0080
Range
0.019-0.0523
O.0019-0.029
<0.0021-0.0523
0.0019-0.0561
0.0029- 0.0561
O.0019- 0.0186
Geo-
metric
Mean
(GM)
0.0046
0.0049
0.0042
0.0049
0.0052
0.0047
95% CI
GM
0.0043-0.0048
0.0046 - 0.0053
0.0039 - 0.0045
0.0047-0.0051
0.0049 - 0.0052
0.0044 - 0.0049
Unless stated otherwise, the geometric mean and 90th percentiles for the combined male and
female adults and combined male and female children are used in the calculations of the MOEs in
the following sections.

5.2.2 Workers

Although some serum level data were available for workers, the data were not adequate to
calculate MOEs for specific occupational exposures. As described in section 4.1, 3M and DuPont
have provided serum monitoring to workers on a voluntary basis. 3M discontinued
manufacturing PFOA between 2000-2002. The last serum monitoring was offered at the 3M
plants in 2002, but there were about 75-80% fewer volunteers in 2002 than in 2000. Thus, the
sample size for 2002 is very small. In addition, since occupational exposures no longer exist for
3M workers, use of the blood monitoring data for 2000 may overestimate current serum levels.
                                          97

-------
SAR RPVIPW Draff- Tin Not rite nr
The data available for the DuPont workers was very limited and there was no information
available for a variety of critical factors including gender, sampling methods, and occupation.

In general, the mean serum levels following occupational exposures appear to be orders of
magnitude higher than observed in the general population. Thus, MOEs for workers would be
expected to be much less than for the general population.

5.3 Calculation of MOEs Based on Non-Human Primate Studies

As described in section 5.1, two studies have been conducted in non-human primates.  A LOAEL
of 3 mg/kg-day was observed in both studies, and neither study established a NOAEL. For
calculation of MOEs, the 6-month male cynomolgus monkey study was chosen since it was a
more robust study. The effects observed at 3 mg/kg-day included increased liver weight and
possibly mortality. Serum levels of PFOA were measured throughout the study and a steady state
level of 77 + 39  jig/ml was observed in the 3 mg/kg-day group.

The most relevant general human population for use in the calculation of the MOEs is the
population of adults, age 20-69, that was examined in the American Red Cross samples (Table
27). Although the cynomolgus monkey study was conducted in males, it should be noted that
there is no evidence of a gender difference in the half-life of PFOA in humans or non-human
primates. Comparing males and females, there was no evidence of a gender difference in toxicity
in the Rhesus monkey study.  In addition, the serum levels in the adult male and female human
populations are very similar.  Therefore, MOEs were calculated using the serum levels from the
combined  males and females. The MOEs were calculated by dividing the steady state serum
value for the male cynomolgus monkeys in the 3  mg/kg-day  group (77 ^g/ml) by the adult serum
levels from the American Red Cross blood samples presented in Table 28. The MOE calculated
using the geometric mean (.0046 ng/ml) for the human serum level is 16,739.  The MOE
calculated using the 90th percentile value (.0094 jig/ml) for the human serum data is 8191.

5.4 Calculation of MOEs Based on Adult Rat Studies

As described in section 5.1, the effect levels that  are appropriate for calculation of the MOEs are
the LOAEL of 1 mg/kg-day for adult male rats from the two-generation reproductive toxicity
study, and the NOAEL  of 1.6 mg/kg-day for adult female rats from the 2-year study. Information
on the serum levels of PFOA were not available for either study.  Therefore, pharmacokinetic
information was used to estimate the Area Under Concentration curve in plasma (AUC) as
described in Appendix A. The estimated AUC for adult male rats exposed to 1 mg/kg-day islOl 1
[ig-hr/mL and the estimated AUC for adult female rats exposed to 1.6 mg/kg-day is 44 ng-hr/mL.

The most relevant general human population for use in the calculation of the MOEs is the
population of adults, age 20-69, that was examined in the American Red Cross samples (Table
28). Assuming that the serum levels reflect steady state in humans, an AUC was calculated by
multiplying the serum concentrations by 24 hours. Thus, the AUC for the geometric mean is
0.1104 |j.g-hr/mL (0.0046 jig/mL X 24 hours), and the AUC for the 90th percentile is 0.2256 \ig-
hr/mL (0.0094 jig/mL X 24 hours).  MOEs were  calculated by dividing the AUC in the adult
female rat by the AUC for the adult humans which is 398 (195 for the 90th percentile) and by
dividing the AUC for the adult male rat by the AUC for the adult humans which is 9158 (4481 for
the 90th percentile).
                                         98

-------
SAR Rpyipw Draff; Ttn Nnf fito nr
5.5 Calculation of MOEs Based on Rat Developmental Toxicity Studies

As described in section 5.1, developmental effects were observed in a two-generation
reproductive toxicity study in Sprague-Dawley rats (York, 2002). These effects were observed at
various times during the maturation of the Fl pups.  During lactation, there was a reduction in Fl
mean body weight on a litter basis (sexes combined) in the 30 mg/kg-day group. There was a
significant increase in mortality mainly during the first few days after weaning, and a significant
delay in the timing of sexual maturation for Fl males and females in the 30 mg/kg-day group.
Mean body weights were also significantly reduced prior to sexual maturation in the Fl males
exposed to 10 and 30  mg/kg-day APFO and in the Fl females exposed to 30 mg/kg-day APFO.

The critical period of exposure for each of the effects is not known.  For example, it is not known
whether prenatal and/or lactational exposure is important for the reduced body weight that was
observed during lactation. Similarly, it is not known whether the reduced body weight, mortality,
or delayed sexual maturation that occurred during the postweaning period are due to prenatal,
lactational, and/or postweaning exposures.  Therefore, it is necessary to calculate MOEs for each
of these exposure periods.

Only limited information on serum levels of PFOA were obtained in the two-generation
reproductive toxicity study.  This included measurements of serum levels in the FO dams in the 10
and 30 mg/kg-day groups which were taken 24 hours after dosing. Given the rapid excretion of
PFOA by female rats, these values are of limited utility. In addition, serum levels were not
obtained for the Fl animals. Therefore, it was necessary to estimate serum levels based on
pharmacokinetic information.

For the prenatal period, information was available on serum levels in the pregnant rat and fetus,
as well as concentrations of PFOA in the embryo. However, the information on serum levels in
the general U.S. population is restricted to children and adults; there is no information on levels in
the human embryo or fetus.  Therefore, it is only appropriate to derive MOEs based on serum
levels in the pregnant rat and human females. Estimates of Cmax? and AUC in the pregnant rat
were used for calculation of the MOEs since it is not known which dose metric may be more
significant. These dose metrics were estimated for a pregnant rat administered 3 mg/kg-day
APFO since the lowest NOAEL for the  developmental effects in the two-generation reproductive
toxicity  study was 3 mg/kg-day.

As described in section 3.2.4.1, PFOA serum levels were measured in the pregnant rat on
gestation days 10,15  and 21 following exposure to 3,10, or 30 mg/kg-day APFO (Mylchreest,
2003).  The serum levels were determined two hours after dosing. Kemper (2003) determined
that peak serum concentrations of PFOA occur approximately 0.5 to 2 hrs after gavage dosing in
the adult female rat. Thus, the serum levels obtained by Mylchreest  (2003) are approximately
equivalent to Cmax.  An average Cmax for the pregnant rat was then calculated by averaging the
serum levels obtained at gestation days 10,15 and 21; this value is 13 ng/ml for the 3 mg/kg-day
dose group. The AUC was estimated using a one compartment model which is described in
Appendix A.  The AUC for the pregnant rat administered 3 mg/kg-day is approximately 83 \ig-
hr/mL.

The most appropriate human population for comparison is the group of adult females, ages 20 -
69 years (Table 28). MOEs for the prenatal period were calculated by dividing the average Cmax
for the pregnant rat by the measured serum levels for adult human females which is 3,095 (1548
using the 90th percentile for human females), and by dividing the estimated AUC for the pregnant


                                          99

-------
SAR RPVJPW Draft; Tin Nnf Pitp nr
rat by the AUC for adult human females which is 823 (412 using the 90th percentile for human
females).

Estimation of meaningful serum levels in the Fl pups during lactation is problematic.  As
described in section 3.2.4.1,  Mylchreest (2003) measured serum PFOA concentrations on days 3,
7,14, and 21 of lactation. These measurements were taken approximately 2 hours after the dams
were dosed. Pups were separated from the dams 1-2 hours prior to the time that blood samples
were collected, and therefore it is not clear when the pups last nursed relative to the time that their
blood samples were collected. Han (2003) has shown that the clearance in 4-week old male and
female pups is fairly rapid. Assuming that the clearance is also rapid during the lactation period,
steady state would not be achieved. Thus, the measured serum concentrations in the pups would
not be expected to estimate steady  state concentrations, and in fact, it is not clear what dose
metric the  serum levels represent.  The AUC for the pups could also be estimated, but, this
requires knowledge of milk consumption and elimination constants during lactation. While
estimates of milk consumption and elimination constants could be made, such estimates would
introduce considerable  uncertainty in the resulting estimates of the AUC. Therefore, it is
premature  at this point to derive estimates of the AUC for the lactation period.  In addition, there
is no information from the available human biomonitoring studies on serum levels in children less
than two years of age. Consequently, MOEs were not calculated for the lactation period.

MOEs were calculated for the postweaning period.  As described in section 3.2.4.2, Han (2003)
measured serum concentrations of PFOA in rat weanlings following a single dose of 10 mg/kg-
day APFO at 4, 5, 6, 7 and 8  weeks of age. Using a one compartment model, this information
was used to provide estimates of AUC (Appendix A). In the two-generation reproductive toxicity
study, several effects were noted during the postweaning period. One of these effects was
mortality that occurred mainly during the first few days after weaning in the male and female
groups exposed to 30 mg/kg-day APFO. For this effect, the most appropriate dose metric is the
estimated AUC for the 4-week old rat exposed to 30 mg/kg-day APFO; these estimates are 2,022
and 1,383 iig-hr/mL for the males and females, respectively.

The most appropriate human population for comparison is the children of ages 2-12 (Table 28).
The AUC for human children calculated from the geometric mean is 0.1176 ^g-hr/mL (0.0049
lig/mL X 24 hours), and the AUC for the 90th percentile is 0.204 jig-hr/mL (0.0085 jig/mL X 24
hours).  MOEs were calculated using the gender specific estimates for the AUCs for the 4-week
rat weanlings and the AUC for children. The MOEs calculated from the AUCs for male and
female rat  weanlings, respectively, and the AUC for the geometric mean for children are 17,194
and 11,760, respectively.  If the AUC for the 90th percentile for the human children is used, the
resulting MOEs are 9,912 and 6,779, respectively.

Another effect that was observed in the two-generation reproductive toxicity study was a delay in
the sexual  maturation of the Fl females exposed to 30 mg/kg-day APFO. Mean body weights
were also reduced in this group prior to sexual maturation. Since sexual maturation occurs at
approximately 5 weeks of age, the most appropriate estimate of serum concentrations is the
average of the estimated AUCs for the 4 and 5 week female weanling  exposed to 30 mg/kg-day
(Appendix A) which is 1233  jig-hr/mL.  The MOE calculated using the average AUC for the 4
and 5 week female weanling rat and the AUC for the geometric mean  for human children is
10,485 (6,044 using the AUC for the 90th percentile).

Sexual maturation was  also delayed in the Fl males in the two-generation reproductive toxicity
study exposed to 30 mg/kg-day APFO. In addition, mean body weights were reduced prior to


                                          100

-------
SAR RPVIPW Fit-aft; Dn Not fifp nr Ounfp
sexual maturation in Fl males exposed to 10 and 30 mg/kg-day. Sexual maturation occurs at
approximately 7-7.5 weeks, and the most appropriate estimate of serum concentrations is the
average of the estimated AUCs for the 4 - 8 week male weanling (Appendix A). The average
AUC at 10 mg/kg-day is 9237 jig-hr/mL. The MOE calculated from the average AUC for the 4 -
8 week male weanling rat and the AUC for human children is 78,546 (45,279 using the AUC for
the 90th percentile).

5.6 Uncertainties in the Risk Characterization

This assessment relies on a rich hazard data base for an industrial chemical. Extensive
pharmacpkinetic information is available for rats through several life stages, and more limited
information is available for humans and non-human primates.  Similarly, there are animal
toxicology studies for a variety of species including non-human primates, rats, mice and rabbits,
and more limited epidemiological studies.  Extensive mode of action information is available that
has established that the liver toxicity and liver tumors are due to PPARa-agonism. In addition,
serum biomonitoring information is available for humans throughout most of the life stages.
PFOA serum levels were also obtained in some of the animal toxicology studies, and
pharmacokinetic data permitted the estimation of serum levels in some cases where measured
values were not available.  These data permitted an analysis of potential human health effects
associated with exposure to PFOA utilizing a variety of animal models, several life stages, and
fairly precise estimates of actual human exposure.

Nonetheless, there are several uncertainties in this assessment.  These are discussed below in a
qualitative fashion, since it is premature to attempt a quantitative analysis at this time.  One area
of uncertainty pertains  to the potential carcinogenesis associated with APFO exposure.  In rats,
chronic exposure to APFO results in liver adenomas, Ley dig cell adenomas, and pancreatic acinar
cell tumors.  An extensive mode of action analysis has demonstrated that the liver tumors result
from PPARa-agonism. However, there are insufficient information to establish the mode of
action for the Leydig cell and pancreatic acinar cell tumors.  In addition, there is some uncertainty
regarding the development of mammary fibroadenomas following chronic APFO exposure in
rats.

In addition, it is not known which animal species is the most appropriate model for humans. The
prenatal developmental toxicity study in New Zealand White rabbits and the studies of
immunotoxicity in mice were not considered in the risk assessment since serum levels of PFOA
were not measured and there were no pharmacokinetic studies available with which to estimate
serum levels. In addition, the female rat may not be a good model for humans due to the gender
specific differences in elimination in rats.  Information that is available for humans indicates that
there is no gender difference in elimination.  Thus, human females would be expected to be
presented with a fairly  continuous internal exposure to PFOA due to the long half-life, whereas
the female rat experiences an internal exposure that is rapid and discontinuous because of the
short half-life. This may be of particular importance in delineating the potential hazards
associated with exposure to PFOA during development. Indeed, Ihe mouse does not appear to
have a gender difference in the elimination of PFOA, and preliminary results of a developmental
toxicity study in CD-I  mice indicate early  pregnancy loss, compromised postnatal survival and
delays in postnatal growth and maturation following prenatal exposure to PFOA (Lau et al.,
2004).

Non-human primate data were used to assess the potential health effects to adult humans. There
is some uncertainty associated with the two studies that have been conducted due to the small
sample size and the lack of a NOAEL. There is also some uncertainty associated with the exact

                                          101

-------
SAB RPVJPW Draft; Dn Nnf fite nt-
effects in the cynomolgus monkey study as it is not clear whether treatment-related mortality
occurred in the lower dose groups. Finally, there is some uncertainty regarding the serum levels
of PFOA in the cynomolgus monkey study due to the lack of a linear increase in serum PFOA
levels in proportion to dose, the high variability, and the precision of the method itself which was
± 30% (inter-assay, intra-assay, and system).

Data from several toxicology studies in adult rats were also used to estimate the potential for
health effects in adult humans. The only consistent effect that was noted in adult animals was a
reduction in body weight. The implications of this effect for human health are not clear, but the
endpoint was used to provide a conservative benchmark from a second species. As noted above,
the female rat may not be a good model for human females. It is possible, due to the kinetics in
the female rat, that some effects were not observed in the rat studies that may be observed in
species with a longer half-life. In addition, there was a dose-related increase in the incidence  of
ovarian tubular hyperplasia in the 2-year study in rats. This effect was not observed in studies of
shorter duration and did not progress to tumor formation in the 2-year study. Thus, the
significance of this finding is not clear. Further research will need to be conducted to determine
the significance of the finding for humans.

Several developmental effects associated with PFOA exposure were observed in the two-
generation reproductive toxicity study in rats.  The study did not determine the critical period  of
exposure for the developmental effects and serum levels were not measured. Therefore, it was
necessary to use pharmacokinetic data to estimate the serum levels for the prenatal, lactation,  and
post-weaning periods.  For the prenatal period, dose metrics were calculated for the pregnant
dams; dose metrics were not calculated for the embryo/fetus since comparable dose metrics do
not exist for humans. As described in Appendix A, pharmacokinetic data in adult female rats
were used to estimate values for the volume of distribution, and elimination and absorption rate
constants; this information was then used to estimate the AUC in the pregnant rat. Although one
study did not report a difference in the elimination of PFOA during pregnancy in the rat (Gibson
and Johnson, 1983), there is still some uncertainty in whether the kinetics are the same in the
pregnant and non-pregnant rat.  There is also some uncertainty whether the distribution of PFOA
to the embryo/fetus is similar in rats and humans.

Unfortunately, dose metrics during the lactation period were not calculated. Therefore, it is not
known what the potential MOE may be for this exposure period. Methods to estimate the  dose to
the offspring during lactation are under development.

Dose metrics for the post-weaning period were calculated. As described in Appendix A, data for
post-weaning rats permitted estimation of the elimination constants during the post-weaning
period. The adult volume of distribution and absorption constants were used;  thus, there is some
uncertainty whether the volume of distribution and absorption constants apply across all life
stages.

The pharmacokinetic modeling described in Appendix A considered both one and two
compartment models. Given the substantial database created for single dose pharmacokinetic
studies in the rat, there would be expected to be relatively small uncertainties estimating AUCs.
In general the fits to the plasma time course data were not substantially improved with the two
compartment model as indicated by generally similar AIC values (see Tables A-xii and A-xv).
Given the increased number of parameters in the two compartment model AIC values would have
to be substantially lower for that model to give a statistically improved fit. Visual inspection
showed that the divergence in plasma time course from the one compartment fit tended to  occur
at later times in some dose groups. With the females, it was impossible to obtain two-

                                          102

-------
SAR RPVJPW Draff- Tin Nnf Cite nr Quote
compartment fits at lower concentrations due to rapid clearance and analytical detection limits.
Given the relatively short period of dosing for the reproductive/developmental studies, error
introduced through use of the one compartment model would be relatively small.  For the male
animals, doses that appeared well fitted by one compartment were interspersed with doses
suggestive of a second compartment. Use of the one compartment model to estimate AUCs for
chronic duration studies may tend to introduce greater error, though the greater uncertainty for the
chronic studies is whether extended dosing alters physiology and thus pharmacokinetics.

There is also some uncertainty regarding the use of the human biomonitoring data. Although the
available data include a range of populations with various demographic features, there may be
some populations that are not represented. Since it is unknown how the human exposures are
occurring, proximity to a manufacturing plant may be a factor in exposure. However, populations
living near the plants were not sampled.  Therefore, it is possible that PFOA serum levels may be
underestimated for certain portions of the U.S. population. The children's sample was derived
from blood collected in 1994/1995; therefore, it may not reflect the current status of PFOA in
children's blood. It is not clear how PFOA may affect more sensitive subpopulations or if their
exposures would vary. PFOA will be included in future environmental reports of the National
Health and Nutrition Examination Survey (NHANES), in an effort to acquire nationally
representative serum PFOA levels; however,  children's serum levels will not be included.

Finally, there is some uncertainty in the interpretation of the MOEs. MOEs are generally
calculated by dividing an administered dose in an animal study by an estimate of external human
exposure. The value of the MOE that is associated with a concern for toxic effects is generally
expressed as the product of the applicable uncertainty factors.  The uncertainty factors that are
considered are the same as those considered in the derivation of a RfD or RfC.  Generally this
includes consideration of an uncertainty factor for intraspecies variation and another for
interspecies variation. An additional factor may be considered if a LOAEL is used from the
animal toxicology study. In this assessment, MOEs were calculated from serum levels in animals
and humans. Generally, data are not available for this kind of estimate, and therefore there is
little experience in determining whether the same uncertainty factors apply. In addition, this
assessment is unique in that the half-life of PFOA is very different among the various animal
species and humans. Whether this should be a consideration in interpreting the MOE is not clear.

6.0 Overall Conclusions

This risk assessment focused on the potential human health effects associated with exposure to
PFOA and its salts. The results of existing epidemiology studies are not adequate for use in
quantitative risk assessment, and therefore the analysis was restricted to endpoints in the animal
toxicology studies. MOEs were calculated for the general U.S. population.  Although some
serum level data were available for workers, the data were not adequate to calculate MOEs for
specific occupational exposures. In general, the mean serum levels following occupational
exposures appear to be orders of magnitude higher than observed  in the general population. Thus,
MOEs for workers would be expected to be much less than for the general population.

A variety of endpoints from the animal toxicology studies were used to calculate MOEs for this
draft risk assessment. The endpoints encompassed different species, gender and life  stages.  For
this draft assessment, specific recommendations on the most appropriate
endpoint/lifestage/species/gender have not been made; rather, all have been presented to provide
transparency.
                                          103

-------
SAB Rpyifw rifflft f>n Nnf riff nr
For adults, two sets of MOEs were calculated based on the toxicology studies in non-human
primates and rats. First, calculation of MOEs from the cynpmolgus monkey study was based on
increased liver weight and possible mortality. The MOE using the geometric mean for the human
serumlevel is 16,739 (8,191 for the 90th percentile). Second, calculation of MOEs from the adult
rat studies was based on reductions in body weight. MOEs were calculated separately for the
female and male rat due to the gender differences in pharmacokinetics in this species. MOEs
were calculated by dividing the AUC in the adult female rat by the AUC for the adult humans
which is 398 (195 for the 90th percentile) and by dividing the AUC for the adult male rat by the
AUC for the adult humans which is 9158 (4481 for the 90th percentile).

MOEs were calculated for the developmental effects in the two-generation reproductive toxicity
study in rats. These effects were observed at various times during the maturation of the Fl pups;
the critical period of exposure for each of the effects is not known, and could be due to exposure
during the prenatal, lactational, and/or postweaning periods.  Ideally, MOEs should be calculated
for each of these exposure periods; however MOEs were not calculated for the lactation period
due to uncertainties in pharmacokinetics.  For the prenatal period, MOEs were calculated for the
pregnant human female; MOEs were not calculated for the fetus since there is no information on
human serum levels in fetuses.  MOEs were calculated using both Cmax  and AUC; the MOE based
on Cmax is 3,095 (1548 for the 90th percentile) and the MOE based on the AUC is 823 (412 for the
90th percentile).

For the postweaning period, MOEs were calculated for several endpoints including reductions in
body weight, mortality and delayed sexual maturation.  These MOEs were based on the geometric
mean for children and range from 10,484 - 78,546 (the range using the  90th percentile is 6,044 -
45,279).

This assessment has provided a range of MOEs for several life stages.  Several uncertainties have
been discussed in a qualitative fashion in this assessment, which highlight the need to interpret
the MOEs with caution. For example, MOEs were not calculated for the lactation period due to
insufficient data, and this life stage may represent an important exposure period.  Similarly, the
biomonitoring data for the children are from samples collected in 1994  and may not be
representative of current children's serum levels. In addition, ascertaining potential levels of
concern will necessitate a better understanding of the appropriate dose  metric in rats or other
animal models, and the relationship of the dose metric to the human serum levels.

Finally, there is some uncertainty associated with the determination of the adequacy of a specific
MOE in protecting human health in the present context. Generally, MOEs are calculated from
administered dose levels and estimates of human exposure.  In this assessment, the MOEs were
calculated from internal dose metrics in animals and humans.  While use of internal dose metrics
reduces many uncertainties pertaining to exposure, there is little experience or guidance on the
factors that should be considered in making judgements about the level of concern associated
with a given MOE. Approaches that are used for conventional MOEs,  if applied unchanged,
indicate that among the populations of interest some individuals are highly exposed, for reasons
not understood at mis time. However, if conventional approaches for determining levels of
concern are not appropriate for MOEs based  on internal dose metrics, then this conclusion would
have to be re-evaluated as the understanding  of this question evolves.
                                          104

-------
SAR RPVIPW Draft; Ttn Nnt fifp nr
7.0 References

3M (Minnesota Mining and Manufacturing Company).  1999a Letter (dated May 26,1999) re:
TSCA 8(e) Supplemental Notice: Sulfonate-based and carboxylic-based fluorochemicals-Docket
Nos. 8EHQ-1180-373; 8EHQ-1180-374; 8EHQ-0381-0394; 8EHQ-0699-373.

3M (Minnesota Mining and Manufacturing Company).  1999b The Science of Organic
Fluorochemistry" and "Perfluorooctane sulfonate: current summary of human sera, health, and
toxicology data" February 5,1999. 8EHQ-0299-373.

3M (Minnesota Mining and Manufacturing Company).  2001  Material Safety Data Sheet FC-26
FLUORAD Brand Fluorochemical Acid, ID Number/U.P.C.: ZF-0002-0376-8. Online:
http://www.3m. com/US/safety/index.jhtml.

Abdellatif, A.G.; Preat, V.; Taper, H.S. and Roberfroid, M.  1991 The modulation of rat liver
carcinogenesis by perfluorooctanoic acid, a peroxisome proliferator. Toxicol. Appl. Pharmacol.
111:530-537.

Alexander, B.H. 2001a Mortality  study of workers employed at the 3M Cottage Grove facility.
Final Report. April 26, 2001. Division of Environmental and Occupational Health, School of
Public Health, University of Minnesota.  US EPA AR226-1030a018.

Alexander, B.H. 2001b Mortality  study of workers employed at the 3M Decatur facility. Final
Report.  April 26, 2001. Division of Environmental and Occupational Health, School of Public
Health, University of Minnesota. US EPA AR226-1030a019.

Berthiaume,  J. and Wallace, K.B. 2002 Perfluorooctanoate, perfluorooctanesulfonate, and N-
ethyl perfluorooctanesulfonamido ethanol; peroxisome proliferation and mitochondrial
biogenesis. Toxicol. Lett. 129:23-32.

Biegel, L.B.; Liu, R.C.M; Hurtt, M.E. and Cook, J.C. 1995 Effects of ammonium
perfluorooctanate on Leydig  cell function: in vitro, in vivo, and ex vivo  studies. Toxicol. Appl.
Pharmacol. 134:18-25.

Biegel, L.B.; Hurtt, M.E.; Frame, S.R.; O'Conner, J.C. and Cook, J.C. 2001 Mechanisms of
extrahepatic tumor induction by peroxisome proliferators in male CD rats. Toxicol. Sci. 60:44-55.

Bluist, S.; Cherrington, N.J; Choudhuri, J.; Hartley, D.P. and Klaassen, C.D. 2002 Gender-
specific and developmental influences on the expression of rat organic anion transporters. J.
Pharmacol. Exper. Ther. 301:145-151.

Boit, H-G., Editor. 1975 Beilstein Handbook of Organic Chemistry. Springer-Verlag, Berlin.
4th Work, Volume 2,  Part 2, page  994.

Burris, J.M.; Olsen, G.; Simpson, C. and Mandel, J. 2000 Determination of serum half-lives of
several fluorochernicals. 3M Company.  Interim Report #1.  June 8, 2000. US EPA AR226-0611.

Burris, J.M.; Lundberg, J.K.; Olsen, G.W.; Simpson, D. and Mandel, G. 2002 Determination of
serum half-lives of several fluorochernicals. 3M Company. Interim Report #2. January 11,
2002. USEPAAR226-1086.
                                         105

-------
SAR RPVIPW Draft; F>n Nnf Cite nr
Butenhoff, J.; Costa, Q; Elcombe, C.; Farrar, D.; Hansen, K; Iwai, H.; Jung, R; Kennedy, G.;
Lieder, P.; Olsen, G. and Thomford, P. 2002 Toxicity of ammonium perfluorooctanoate in male
cynomolgus monkeys after oral dosing for 6 months. Toxicol. Sci. 69: 244-257.

Butenhoff, J.L.; Kennedy, G.L.; Frame, S.R.; O'Connor, J.C. and York, R.G. 2004. The
reproductive toxicology of ammonium perfluorooctanoate (APFO) in the rat. Toxicology.
196:95-116.

Calfours, J. and Stilbs, P. 1985 Solubilization in sodium perfluorooctanoate micelles: a
multicomponent self-diffusion study.  Colloid Interface Sci. 103:332-336.

Cattley, R.C.; DeLuca, J.; Elcombe, C.; Fenner-Ciisp, P.; Lake, E.G.; Marsman, D.S.; Pastoor,
T.A.; Popp, J.A.; Robinson, D.E.; Schwetz, B.; Tugwood, J. and Wahli, W. 1998 Do peroxisome
proliferating compounds pose a hepatocarcinogenic risk to humans? Regul. Toxicol. Pharmacol.
27:47-60.

Cameron, R.G.; Imaida, K.; Tsuda, H.; and Ito, N. 1982  Promotive effects of steroids and bile
acids on hepatocarcinogenesis initiated by diethylnitrosamine. Cancer Res. 42:2426-2428.

Chandra, M; Riley, M.G.I.; and Johnson, DE. 1992  Spontaneous neoplasms in aged sprague-
dawley rats. Arch. Toxiocol. 66:496-502.

Chapin, R.E.; Stevens, J.T.; Hughes, C.L.; Kelce, W.R.; Hess, R.A;and Daston, G.P.  1996
Symposium overview: endocrine modulation of reproduction. Fund. Appl. Toxicol. 29:1-17.

Cheung, C., Akiyama, TE, Ward, JM, Nicol, CJ, Feigenbaum, L., Vinson,  C., and Gonzalez, FJ.
2004. Diminished hepatocellular proliferation in mice humanized for the nuclear receptor
peroxisome proliferator-activated receptor a. Cancer Res. 64:3849-3854.

Christopher, B. and Marias, A.J.  1977 28-Day oral toxicity study with FC-143 in albino mice,
Final Report, Industrial Bio-Test Laboratories, Inc. Study No. 8532-10655, 3M Reference No. T-
1742CoC, Lot 269.

Clegg, E.D.; Cook, J.C.; Chapin, R.E.; Foster, P.M.D.; and Daston,  G.P. 1997  Leydig cell
hyperplasia and adenoma formation: mechanisms and relevance to humans. Reproduct. Toxicol.
11:107-121.

Cook, J.C.; Hurtt, M.E.; Frame, S.R. and Biegel, L.B. 1994  Mechanisms  of extrahepatic tumor
induction by peroxisome proliferators in Crl:CD BR(CD) rats. Toxicologist, 14:301, (abstract #
1169).

Cook, J.C.; Klinefelter, G.R.; Hardisty, J.F.; Sharpe, R.M. and Foster, P.M. 1999. Rodent Leydig
cell tumorigenesis: A review of physiology, pathology, mechanisms and relevance to humans.
Crit. Rev. Toxicol. 29:169-261.

Cotran, R.S.; Kumar, V. and Robbins, S.L. 1989  The pancreas. In:  Pathologic  Basis of Disease,
pp. 981-1010. W.B. Saunders, Philadelphia

Cutts, J.H. and Noble, R.L. 1964 Estrone-induced mammary tumors in the rat. I. Induction and
behavior of tumors. Cancer Res. 24:1116-1123.
                                          106

-------
SAR RPVJPW Draff; Dn Not fito nr Onnfo
Daikin Chemicals Sales Co., LTD. 2001 Online Catalog. Fluoro aliptiatic compounds.
http://daikin-dcs. co.jp/catalog/

Dean, W.P. and Jessup, D.C. 1978 Acute oral toxicity (LD50) study in rats. International
Research and Development Corporation, Study No. 137-091, May 5,1978. US EPA AR226-
0419.

Dupont (Haskell Laboratory). 2001 a  Letter to Charles Auer from Gerald Kennedy, dated
January 25,2001. FYI-00-001378. US EPAAR226-1000.

Dupont (Haskell Laboratory). 2001b  Letter from Gerald Kennedy re: FYI-010-01378 and FYI-
0301-01378, dated May 29, 2001.  US EPA AR226-1023.

Dupont (Haskell Laboratory). 2003  Epidemiology surveillance report: Cancer incidence for
Washington works site 1959-2001. US EPA AR226-1307-6.

Eagon, P.K;, Chander, N.; Epley, M.J.; Elm, M.S.; Brady, E.P. and Rao, K.N. 1994. Di(2-
ethylhexyl)phthalate-mduced changes in liver estyrogen metabolism and hyperplasia. Int. J.
Cancer 58:736-743

Edwards, P.J.B; Jolley, K.W.; Smith, M.H.; Thomsen, S.J.; and Boden, N. 1997  Solvent isotope
effect on the self-assembly liquid crystalline phase behavior in aqueous solutions of ammonium
pentadecafluorooctanoate..  Langmuir:13(10):2665-2669.

Exfluor Research Company. 1998 Material Safety Data Sheet for Pentadecafluorooctanoic acid,
96%, Acros Organics N.V.  Online: http://216.122.170.58/msds.asp?ExflP=C8AC

FIFRA SAP. 2004 Transmittal of meeting minutes of the FIFRA Scientific Advisory Panel
meeting held December 9, 2003.  Memo, from S. M. Knott to J.J. Jones and C.M. Auer, March 5,
2004. Office of Science Coordination and Policy, U.S. Environmental Protection Agency,
Washington, DC.

Fisher Scientific. 2003 Material Safety Data Sheet for C8AC, Perfluorooctanoic acid, 98%.
Online:  https://fscimage.fishersci.com/msds/97331 .htm

Frame, S.R. and McConnell, E.E. 2003 Review of proliferative lesions of the exocrine pancreas
in two chronic studies in rats with ammonium perfluorooctanoate. DuPont-13788. Oct. 16, 2003.

Gabriel, K. 1976a Primary  eye irritation study in rabbits.  Biosearch, Inc., March 4,1976. US
EPAAR226-0422.

Gabriel, K. 1976b Primary  skin irritation study in rabbits.  Biosearch, Inc., March 4,1976.  US
EPAAR226-0423.

Gabriel, K. 1976c Acute oral toxicity - rats. Biosearch, Inc., September 16,1976. US EPA
AR226-0425.

Gabriel, K. 1976d Primary  eye irritation study in rabbits.  Biosearch, Inc., September 16,1976.
USEPAAR226-0426.
                                         107

-------
SAR RPVJPW Draff; Hn Nnt Cite ni-
Garry, V.F. and R.L. Nelson.  1981 An assay of cell transformation and cytotoxicity in
C3H10T'/2 clonal cell line for the test chemical T-2942 CoC. Stone Research Laboratories,
Minneapolis, MN, March 4,1981. US EPA AR226-0428.

Gazouli, M.; Yao, Z.X.; Boujard, N.; Gorton, J.C.; Culty, M. and Papadopoulos, V. 2002.Effect
of peroxisome prohferators on Ley dig cell periferal-type benzodiazepine receptor gene
expression, hormone-stimulated chloresterol transport, and steroidogenesis: Role ofthe proxisome
proliferator-activated receptor a. Endocrinology 143:2571-2583.

Gibson, S.J. and Johnson, J.D. 1979  Absorption of FC-143-14C in rats after a single oral dose.
Riker Laboratories, Inc., Subsidiary of 3M, St. Paul, Minnesota.

Gibson, S.J. and Johnson, J.D. 1980  Extent and route of excretion and tissue distribution of total
carbon-14 in male and female rats after a single I.V. dose of FC-143-14C.  Riker Laboratories,
Inc., Subsidiary of 3M, St. Paul, MN.

Gibson, S.J. and Johnson, J.D. 1983  Extent and route of excretion of total carbon-14 in pregnant
rats after a single oral dose of ammonium 14 C-perfluorooctanoate. Riker Laboratories, Inc.,
Subsidiary of 3M, St. Paul, MN.

Glaza, S.  1995 Acute dermal toxicity study of T-6342 in rabbits.  Coming Hazelton, Inc.
Madison, WI. Project ID: HWI 50800374.  3M Company. St. Paul, MN. US EPA AR226-0427.

Glaza, S.M. 1997 Acute oral toxicity study of T-6669 in rats. Corning Hazleton Inc.  Study No.
CHW 61001760, January 10,1997. US EPA AR226-0420.

Gilliland, F. 1992  Fluorochemicals and human health: Studies in an occupational cohort.
Doctoral thesis, Division of Environmental and Occupational Health, University of Minnesota.
USEPAAR226-0473.

Gilliland, F.D. and Mandel, J.S.. 1993  Mortality among employees of a perfluorooctanoic acid
production plant. J. Occup. Med. 35(9):950-954.

Gilliland, F.D. and Mandel, J.S.  1996 Serum perfluorooctanoic acid and hepatic enzymes,
lipoproteins, and cholesterol:  A study of occupationally exposed men. Am. J. Ind.  Med.
29:560-568.

Goldenthal, E.I. 1978a Final Report, Ninety Day Subacute Rat Toxicity Study on Fluorad®
Fluorochemical FC-143, International Research and Development Corporation, Study No. 137-
089, 3M Reference No. T-3141, November 6,1978. US EPA AR226-0441.

Goldenthal, E.I. 1978b Final Report, Ninety Day Subacute Rhesus Monkey Toxicity Study,
International Research and Development Corporation, Study No. 137-090, November 10,1978.
USEPAAR226-0447.

Gortaer, E.G. 1981  Oral teratology study of T-2998CoC in rats. Safety Evaluation Laboratory
and Riker Laboratories,  Inc.  Experiment Number: 0681TR0110, December 1981.

Gortner, E.G. 1982 Oral teratology study of T-3141 CoC in rabbits. Safety Evaluation
Laboratory and Riker Laboratories, Inc. Experiment Number:  0681TB0398, February 1982.


                                          108

-------
SAB RPVJPW riraft; Dn Nnf Pifo nr Oiinfp
Griffith, F.D. and Long, I.E. 1980 Animal toxicity studies with ammonium perfluorooctanoate.
Am Ind Hyg Assoc J 41(8):576-583.

Han, H.; Snow, T.A.; Kemper, R.A. and Jepson, G.W.  2003 Binding of perfluorooctanoic acid
to rat and human plasma proteins. Chem, Res. Toxicol. 16:775-781.

Han, X. 2003 Ammonium perfluorooctanoate: Age effect on the plasma concentration in post-
wealning rats following oral gavage. Haskell Laboratory for Health and Environmental Sciences.
Study No. Dupont-13267, December 15,2003 US EPA AR226-1553

Hanhijarvi, H.; Ophaug, R.H. and Singer, L.  1982 The sex-related difference in
perfluorooctanoate excretion in the rat. Proc. Soc. Exp. Biol. Med. 171:50-55.

Hanhijarvi, H.; Ylinen, M.; Kojo, A. and Kosma, V.M. 1987  Elimination and toxicity of
perfluorooctanoic acid during subchronic administration in the Wistar rat.  Pharmacol. Toxicol.
61:66-68.

Hanhijarvi, H.; Ylinen, M.; Haaranen, T. and Nevalainen, T. 1988 A proposed species difference
in the renal excretion of perfluorooctanoic acid in the beagle dog and rat. In: New Developments
in Biosciences: Their Implications for Laboratory Animal Science. Beynen, AC; Solleveld, HA,
Eds. Martinus Nijhoff Publishers. Dordrecht, Netherlands.

Hansch, C.  and Leo, A.; Eds.  1979 The fragment method of calculated partition coefficients. In:
Substituent Constants for Correlation Analysis and Chemistry and Biology, Chapter I.V. John
Wiley and Sons, Inc., New York.

Hinderliter, P.M. 2003 Perfluorooctanoic acid: relationship between repeated inhalation
exposures and plasma PFOA concentration in the rat. Haskell Laboratory for Health and
Environmental Sciences. Study No. DuPont-12944, November 5, 2003.

I ARC (International Agency for Research on Cancer). 1995 WHO International Agency for
Research on Cancer. Peroxisome proliferation and its role in carcinogenesis. IARC technical
report No. 24. IARC Press, Lyon, France.

Ikeda, T.; Aiba, K.; Fukuda, K. and Tanaka, M. 1985 The induction of peroxisome proliferation
in rat liver by perfluorinated fatty acids, metabolically inert derivatives of fatty acids. J. Biochem.
98:475-482.

Johnson, J.D.  1995a Final report, analytical study, single-dose intravenous pharmacokinetic
study of T-6067 in rabbits.  Study Number: AMDT-120694.1. 3M Environmental Technology &
Services, St. Paul,  MN. US EPA AR226-0453.

Johnson, J.D.  1995b Final report, analytical study, single-dose absorption/toxicity study of
T-6067, T-6068, and T-6069 in rabbits.  Study Number: AMDT-011095.1.  3M Environmental
Technology & Services, St. Paul, MN.  US EPA AR226-0454.

Johnson, J.D.; Gibson, S.J. and Ober, RE. 1984  Cholestyramine-enhanced fecal elimination of
carbon-14 in rats after administration of ammonium [14C]perfluorooctanoate or potassium
[MC]perfluorooctanesulfonate. Fundam. Appl. Toxicol. 4:972-976.
                                          109

-------
SAR RPVJPW Draft; Dn Nnf Cite nr
Kams, M.E. and Fayerweather, W.E. 1991  A case-control study of leukemia at the Washington
Works site. Final Report. Dupont Company. December 31,1991.  US EPA AR226-1308-2.

Kawashima, Y., Uy-Yu, N. and Kozuka, H.  1989  Sex-related difference in the inductions by
perfluoro-octaanoic acid of peroxisome a-oxidation, rnicrosomal 1-acylglycerolphosphocholine
acyltransferase and cytosolic long-chain acyl-CoA hydrolase in rat liver. Biochem. I 261:595-
600.

Keller, B.J.; Marsman, D.S.;  Popp, J.A. and Thurman, R.G. 1992  Several nongenotoxic
carcinogens uncouple mitochondria! oxidative phosphorylation. Biochem. Biophys. Acta,
1102:237-244.

Kemper, R.A. 2003 Perfluorooctanoic acid:  Toxicokinetics in the rat. UE EPA AR 116.

Kennedy, G.L.  1985  Dermal toxicity of ammonium perfluorooctanoate. Toxicol. Appl.
Pharmacol. 81(2):348-355.

Kennedy, G.L.; Hall, G.T.; Brittelli, M.R.; Barnes, J.R. and Chen, H.C.  1986  Inhalation toxicity
of ammonium perfluorooctanoate.  FoodChem. Toxicol. 24(12):1325-1329.

Klaunig, J.E.; Babich, M.A.;  Baetcke, KP.;  Cook, J.C.; Corton, J.C.; David, R.M.; DeLuca, J.G.;
Lai, D.Y.; McKee, R.H.; Peters, J.M.; Roberts, R.A. and Fenner-Crisp, P.A. 2003 PPARa
agonist-induced rodent tumors: Mode of action and human relevance. Crit. Rev. Toxicol. 33:655-
780.

Kudo, N.; Katakura, M.; Sato, Y. and Kawashima, Y. 2002 Sex hormone-regulated renal
transport of perfluorooctanoic acid. Chem. Biol. Interact. 139:301-316.

Lau,C.; Thibodeaux, J.R; Hanson, R.G. and Rogers. J.M. 2004 Effects of perfluorooctanoic acid
exposure  during pregnancy in the mouse. Birth Defects Research, IN PRESS

Lawlor, T.E. 1995  Mutagenicity test with T-6342 in the Salmonella-Escherichia
co///mammalian-microsome reverse mutation assay. Laboratory Number: 17073-0-409.  Corning
Hazleton  Inc., Vienna, VA.  3M Company.  St. Paul, MN. US EPA AR226-0436.

Lawlor, T.E. 1996  Mutagenicity test with T-6564 in the Salmonella-Escherichia
co///mammalian-microsome reverse mutation assay with a confirmatory assay. Coming Hazleton
Inc. Final Report.  CHV Study No: 17750-0-409R.  September 13,1996. US EPA AR226-0432.

Lines, D.  and Sutcliffe, H. 1984 Preparation and properties of some salts of perfluorooctanoic
acid. Journal of Fluorine Chemistry 25:505-512.

Liu, R.C.M.;, Hurtt, M.E.; Cook, J.C. and Biegel, L.B. 1996 Effect of the peroxisome
proliferator, ammonium perfluorooctanate (C8), on hepatic aromatase activity in adult male
Crl:CD BR (CD) rats. Fund. Appl. Toxicol.  30:220-228.

Liu, S.C.; Sanfilippo, B.; Perroteau, I.; Derynck, R; Salomon, D.S. and Kidwell, W.R. 1987
Expression of transforming growth factor alpha (TGF alpha) in differentiated rat mammary
tumors: estrogen induction of TGF alpha production. Mol. Endocrinol. 1:683-692
                                         110

-------
SAR RPVIPW nraft! F>n Nnt fite nr
Longnecker, D.S. 1987  Interface between adaptive and neoplastic growth in the pancreas. Gut
28:253-258

Longnecker, D.S. and Sumi, C. 1990 Effects of sex steroid hormones on pancreatic cancer in the
rat. Int. J. Pancreatol. 7:159-165.

Luebker, D.J.; Hansen, K. J.; Bass, N.M.; Butenhoff, J.L. and Seacat, A.M. 2002 Interactions of
fluorochemicals with rat liver fatty acid-binding proteins, Toxicology:  176:175-185.

Maloney, E.K. and Waxman, D.J. 1999 Transactivation of PPARct and PPARy by structurally
diverse environmental chemicals. Toxicol. Appl. Pharmacol. 161:209-218.

Mann, P.C. and Frame, S.R. 2004 FC-143: Two year oral toxicity-oncogenicity study in rats: Peer
review of ovaries.  DuPont Project ID 15261, June 25, 2004. US EPA AR226.

Markoe, D.M. 1983 Primary skin irritation test with T-3371 in albino rabbits. Riker
Laboratories, Study No.0883EB0079, July 13,1983. US EPA AR226-0424.

Merck KgaA. Undated. Chemical/physical Data,  807114, Pentadecafluorooctanoic acid for
synthesis.  Online: http://www.chemdat.de/cat/web2.zoom_in?text=807114&screen=120&cid=-
788291649&pg=0&s=pentadecafluorooctanoic&lang=4

Metrick, M. and Marias, A.J. 1977  28-Day  Oral Toxicity Study with FC-143 in Albino Rats,
Final Report, Industrial Bio-Test Laboratories, Inc. Study No. 8532-10654, 3M Reference No. T-
1742CoC, Lot 269, September 29,1977

Murli, H.  1995 Mutagenicity test on T-6342 in an in vivo mouse micronucleus assay. Coming
Hazleton Inc., Vienna, VA. Study No.  17073-0-455, December 14,1995. US EPA AR226-0435.

Mylchreest E. 2003 PFOA: Lactational and placenta! transport pharmacokinetic study in rats.
Haskell Laboratory for Health and Environmental Sciences, Newark DE, Study No. DuPont-
13309, December 19, 2003. US EPA AR226-1551

Murli, H.  1996a  Mutagenicity test on T-6564 in an in vivo mouse micronucleus assay. Corning
Hazleton Inc., Vienna, VA. Study number 17750-0-455, November 1,1996. US EPA AR226-
0430.

Murli, H.  1996b  Mutagenicity test on T-6564 measuring chromosomal aberrations in Chinese
Hamster Ovary (CHO) cells with a confirmatory assay with multiple harvests. Coming Hazleton
Inc., Vienna, VA. Study No. 17750-0-437CO, September 16,1996. US EPA AR226-0431.

Murli, H. 1996c Mutagenicity test on T-6342 measuring chromosomal aberrations inhuman
whole blood lymphocytes with a confirmatory assay with multiple harvests. Corning-Hazelton,
Inc., Vienna, VA. Study No. 17073-0-449CO, November 1,1996.  US EPA AR226-0433.

Murli, H. 1996d Mutagenicity test on T-6342 measuring chromosomal aberrations in Chinese
Hamster Ovary (CHO) cells with a confirmatory assay with multiple harvests. Corning-Hazelton,
Inc. Vienna, VA. Study No. 17073-0-437CO, September 16,1996.  US EPA AR226-0434.

Nilsson, R.; Beije, B.; Preat, V.; Erxon, K. and Ramel, C. 1991  On the mechanism of the
hepatocarcinogenicity of peroxisome proliferators. Chem. Biol. Interact. 78:235-250.

                                         Ill

-------
SAR RPVJPVV Draft; Tin Nnf riff* nr OuntA
NIOSH (National Institute of Occupational Safety and Health). 1994 International Chemical
Safety Card, Chloroacetic Acid, ICSC: 2035, IPCS, CEC. Online:
http://www.cdc.gov/niosh/ipcsneng/neng0235.html

NOTOX. 2000 Evaluation of the ability of T-7524 to induce chromosome aberrations in
cultured peripheral human lymphocytes.  NOTOX Project Number 292062. Hertogenbosch, The
Netherlands.

Obourn, J.D.; Frame. S.R.; Bell, R.H. Jr.; Longnecker, D.S.; Elliott,  G.S. and Cook, J.C. 1997
Mechanisms for the pancreatic oncogenic effects of the peroxisome proliferator Wyeth-14,643.
Toxicol. Appl. Pharmacol. 145:425-436.

O'Flaherty, E.J. 1981 Toxicants and Drugs: Kinetics and Dynamics. John Wiley and Sons, New
York.

Olsen, G.W.; Gilliland, F.D.; Burlew, M.M.; Burris, J.M.; Mandel, J.S. and Mandel J.H. 1998a
An epidemiologic investigation of reproductive hormones in men with occupational exposure to
perfluorooctanoic acid. JOEM 40(7):614-622.

Olsen, G.W.; Burris, J.M.; Burlew, MM.; Mandel, J.H. 1998b  3M Final report: an epidemiologic
investigation of plasma cholecystokinin, hepatic function and serum perfluorooctanoic acid levels
in production workers. 3M Company. September 4,1998.  US EPA AR226-0476.

Olsen, G.W.; Burris, J.M.; Mandel, J.H.; Zobel, L.R. 1998c An epidemiologic investigation of
clinical chemistries, hematology and hormones in relation to serum levels of perfluorooctane
sulfonate in male fluorochemical production employees. 3M Company.  Final Report. August 22,
1998. USEPAAR226-0030.

Olsen, G.W.; Logan, P.W.; Simpson, C.A.; Hansen, K.J.; Burris, J.M.; Burlew, M.M.; Schumpert,
J.C. and Mandel, J.H.. 1999 Fluorochemical exposure assessment of Decatur chemical and film
plant employees.  3M Company.  Final Report. August 11,1999. US EPA AR226-0950.

Olsen, G.W.; Burris, J.M.: Burlew, M.M. and Mandel, J.H. 2000 Plasma cholecystokinin and
hepatic enzymes, cholesterol and lipoproteins in ammonium perfluorooctanoate production
workers. Drug Chem Tox 23(4):603-620.

Olsen, G.W.; Logan, P.W.; Simpson, C.A.; Burris, J.M.; Burlew, M.M.; Lundberg, J.K.; and
Mandel, J.H. 2001 a Descriptive summary of serum fluorochemical levels among employee
participants of the year 2000 Decatur fluorochemical medical surveillance program.  3M
Company.  Final Report. March 19, 2001. US EPA AR226-1030a020a.

Olsen, G.W.; Schmickler, M.N.; Tierens J.M.; Logan P.W.; Burris, J.M.; Burlew, M.M.;
Lundberg, J.K.; Mandel, J.H. 2001b Descriptive summary of serum fluorochemical levels among
employee participants of the year 2000 Antwerp fluorochemical medical surveillance program.
3M Company. Final Report.  US EPA AR 226-103Oa020b.

Olsen, G.W.; Madsen, D.C.; Burris, J.M.  and Mandel, J.H. 2001c Descriptive summary of serum
fluorochemical levels among 236 building employees. 3M Company. Final Report. March 19,
2001. USEPAAR226-1030a020c.
                                        112

-------
SAR Rpvi^w Tlraft; Tin Nnf Pitp nr
Olsen, G.W.; Hansen, K.J.; Clemen, L.A.; Burris, J.M. and Mandel, J.H. 2001d Identification of
Fluorochemicals in Human Tissue. 3M Company. Final Report. June 25, 2001. US EPA AR226-
1030a022.

Olsen G.W.; Burlew, M.M.; Burris, J.M. and Mandel, J.H. 2001e A cross-sectional analysis of
serum perfluorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA) in relation to clinical
chemistry, thyroid hormone, hematology and urinalysis results from male and female employee
participants of the 2000 Antwerp and Decatur fluorochemical medical surveillance program. 3M
Company. Final Report.  October 11, 2001. US EPA AR226-1087.

Olsen, G.W.; Burlew, M.M.; Burris, J.M..and Mandel, J.H. 2001f 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. 3M Company. Final Report. October
11,2001. USEPAAR226-1088.

Olsen, G.W.; Burlew, M.M.; Hocking, B.B.; Skratt, J.C.; Burris, J.M. and Mandel, J.H. 2001g
An epidemiologic analysis of episodes of care of 3M Decatur chemical and film plant employees,
1993-1998. 3M Company. Final Report.  May 18, 2001. US EPA AJA226-1030a021.

Olsen, G.W.; Burris,  J.M.; Lundberg, J.K.; Hansen, K..; Mandel, J.H. and Zobel, L.R. 2002a
Identification of fluorochemicals in human sera.  I. American Red Cross adult blood donors.  3M
Company. Final Report.  February 25,2002.  US EPA AR226-1083.

Olsen, G.W.; Burris,  J.M.; Lundberg, J.K.; Hansen, K.J.; Mandel, J.H. and Zobel, L.R. 2002b
Identification of fluorochemicals in human sera.  II. Elderly participants of the Adult Changes in
Thought Study, Seattle, Washington. 3M Company. Final Report. February 25,2002.  US EPA
AR226-1084.

Olsen, G.W.; Burris,  J.M.; Lundberg, J.K.; Hansen, K.J.; Mandel, J.H. and Zobel, L.R. 2002c
Identification of fluorochemicals in human sera.  III. Pediatric participants in a Group A
Streptococci clinical trial  investigation.  3M Company. Final Report. February 25, 2002.  US
EPAAR226-1085.

Olsen, G.W. and Mandel, J.H.  2003a Descriptive analysis of serum fluorochemical
concentrations from Cottage Grove employee participants of the 2002 Medical Surveillance
Program. 3M Company.  Final Report. USEPAAR226-1352.

Olsen, G.W. and Mandel, J.H.  2003b Descriptive analysis of serum fluorochemical
concentrations from Decatur employee participants of the 2002 Medical Surveillance Program.
3M Company. Final Report. US EPA AR226-1353.

Olsen, G.W.; Church, T.R.; Miller, J.P.; Burris, J.M.; Hansen, K.J.; Lundberg J.K.; Armitage,
J.B.; Herron, R.M.; Medhdizadehkashi, Z.; Nobiletti, J.B.; O'Neill, E.M.; Mandel, J.H.; and
Zobel, L.R. 2003d Perfluorooctanesulfonate and other fluorochemicals in the serum of American
Red Cross adult blood donors.  Environ Health Perspect 111(16): 1892-1901.

Olsen, G.W.; Burris,  J.M.; Burlew, M.M.; Mandel, J.H.  2003e Epidemiologic assessment of
worker serum perfluorooctanesulfonate (PFOS) and perfluorooctanoate (PFOA) concentrations
and medical surveillance examinations.  JOEM 45(3): 260-270.


                                         113

-------
SAR RPVJPW Draff? Tin Nnf Cite nr Onntp
Olsen, G.W.; Butenhoff, J.L.; Mandel, J.H.  2003f Assessment of lipid, hepatic and thyroid
function in relation to an occupational biologic limit value for perfluorooctanoate. 3M Company.
Final Report.  June 9, 2003. US EPA AR226-1351.

Olsen, G.W.; Church, T.R.; Larson E.B.; van Belle, G; Lundberg J.K.; Hansen, K.J.; Burris J.M.;
Mandel J.H, and Zobel, L.R. 2004a  Serum concentrations of perfluorooctanesulfonate and other
fluorochemicals in an elderly population from Seattle, Washington.  Chemosphere 54:1599-1611.

Olsen, G.W.; Church, T.R.; Hansen, K.J.; Butenhoff, J.L.; Mandel, J.H.; and Zobel, L.R. 2004b
Quantitative evaluation of perfluorooctanesulfonate (PFOS) and other PC's in the serum of
children.  J. Children's Health 2:1-24.

O'Malley, K.D. and Ebbins, K.L. 1981  Repeat application 28 day percutaneous absorption study
with T-2618CoC in albino rabbits. Riker Laboratories, St. Paul, MN. US EPA AR226-0446.

Ophaug, R.H. and Singer, L. 1980 Metabolic handling of perfluorooctanoic acid in rats.  Proc
Soc Exp Biol Med 163:19-23.

OPPTS.  2003  Proposed science policy: PPARa-mediated hepatocarcinogenesis in rodents and
relevance to human health risk assessments. Submitted to: FIFRA Scientific Advisory Panel
(SAP). Office of Prevention, Pesticides & Toxic Substances, November 5, 2003. Available from


Palazzolo, M.J. 1993  Thirteen-week dietary toxicity study with T-5180, ammonium
perfluorooctanoate (CAS No. 3825-26-1) in male rats. Final Report. Laboratory Project
Identification HWI 6329-100. Hazleton Wisconsin, Inc. US EPA AR226-0449.

Palmer, C.N.;  Hsu, M.H;, Griffin, K.J;, Raucy, J.L. and Johnson, E.F. 1998 Peroxisome
proliferator acitivated receptor-alpha expression in human liver. Mol. Pharmacol. 53:14-22.

Panaretakis, T.; Shabalina, I.G.; Grander, D.; Shoshan, M.C. and DePierre, J.W. 2001  Reactive
oxygen species and mitochondria mediate the induction of apoptosis in human hepatoma hepG2
cells by the rodent peroxisome proliferator and hepatocarcinogen, perfluorooctanoic acid.
Toxicol. Appl. Pharmacol. 173:56-64.

Pastoor, T.P.;  Lee, K.P.; Perri, M.A. and Gillies, P.J.  1987 Biochemical and morphological
studies of ammonium perfluorooctanate-induced hepatomegaly and peroxisome proliferation.
Exp. Mol. Pathol. 47:98-109.

Prentice, D.E. and Meikle, A.W. 1995 A review of drug-induced Ley dig cell hyperplasia and
neoplasia in the rat and some comparisons with man.  Human Exper. Toxicol.  14: 562-572.

Rusch,G. 1979 An acute inhalation study of T-2305 CoC in the rat. Bio/dynamics, Inc., Study
No. 78-7184, May 3, 1979. US EPA AR226-0417.

Sadhu, D. 2002 CHO/HGPRT forward mutation assay - ISO (T6.889.7).  Toxicon Corporation,
Bedford, MA. Report No. 01-7019-G1, March 28, 2002. US EPA AR226-1101.

Sibinski, L.J. 1987  Final report of a two year oral (diet) toxicity and carcinogenicity study of
fluorochemical FC-143 (perfluorooctanane ammonium carboxylate) in rats. Vol. 1-4, 3M
Company/RIKER Exp. No.0281CR0012; 8EHQ-1087-0394, October 16, 1987.

                                         114

-------
SAB Hpyipvv Draff; Ttn fJnf Titp nr
Simister, E.A.; Lee, E.M.; Lu, J.R.; Thomas, R.K.; Ottewill, R.H.; Rennie, A.R. and Penfold, J.
1992 Adsorption of ammonium perfluoroocanoate and ammonium decanoate at the air-solution
interface. J Chem. Soc., Faraday Trans. 88(20):3033-41.

Sohlenius, A-K.; Andersson, K. and DePierre, J.W. 1992  The effects of perfiuoro-octanoic acid
on hepatic peroxisome proliferation and related parameters show no sex-related differences in
mice. Biochem. J. 285:779-783.

Sonich-Mullin, C.; Fielder, R.; Wiltse, J.; Baetcke, K.; Dempsey, J.; Fenner-Crisp, P.; Grant, D.;
Hartley, M.; Knaap, A.; Kroese, D.; Mangelsdorf, I.; Meek, E.; Rice, J.M. and Younes, M. 2001
IPCS conceptual framework for evaluating a MOA for chemical carcinogenesis. Regul. Toxicol.
Pharmacol. 34:146-152.

Southern Research Institute. 2003  Protein binding of perfluorohexane sulfonate, perfluorooctane
sulfonate and perfluorooctanoate to plasma (human, rat and monkey), and various human derived
plasma protein fractions.  Study ID: 9921.7. US EPA AR226-1354.

Staples, R.E.; Burgess, B.A. and Kerns, W.D.  1984 The embryo-fetal toxicity and teratogenic
potential of ammonium perfluorooctanoate (APFO) in the rat.  Fundam Appl Toxicol 4:429-440.

Starkov, A.A. and Wallace, K.B.  2002 Structural determinants of fluorochemical-induced
mitochondria! dysfunction.  Toxicol. Sci.  66:244-252.

Sykes, G. 1987 Two-year toxicology/carcinogenicity study of fluorochemical FC-143 in rats.
Memo, from G. Sykes to  C. Reinhardt, Haskell Lab. Toxicol. Ind. Med., October 29,1987.

Takagi,  A; Sai, K.; Ummemura, T.; Hasegawa, R. and Kurokawa, Y. 1991 Short-term exposure
to the peroxisome proliferators, perfluorooctanoic acid and perfluorodecanoic acid, causes
significant increases of 8-hydroxydeoxyguanosine in liver DNA of rats. Cancer Lett. 57:55-60.

Takagi,  A.; Sai, K.; Ummemura, T.; Hasegawa, R.; and Kurokawa, Y. 1992 Hepatomegaly is  an
early biomarker for hepatocarcinogenesis induced by peroxisome proliferators. J. Environ.
Toxicol. Pathol. 11:45-149.

Teerds,  K.J.; Rommerts, F.G.; and Dorrington, J.H. 1990 Immunohistochemical detection of
transforming growth factor-a in Ley dig cell during the development of the rat testis. Molec. Cell
Endocrinol. 69:R1-R6.

Thomford, P.J. 2001 a 4-Week capsule toxicity study with ammonium perfluorooctanoate
(APFO) in cynomolgus monkeys. Study performed by Covance Laboratories Inc., Madison
Wisconsin 53704-2592 for APME Ad-hoc APFO Toxicology Working Group. Study No.
Covance 6329-230, Completion Date December 18, 2001,159 pp. US EPA AR226-1052a..

Thomford, P.J. 2001b 26-Week capsule toxicity study with ammonium perfluorooctanoate
(APFO) in cynomolgus monkeys. Study performed by Covance Laboratories Inc., Madison
Wisconsin 53704-2592 for APME Ad-hoc APFO Toxicology Working Group. Study No.
Covance 6329-231, Completion Date December 18, 2001, 463 pp. US EPA AR226-1052a..

Trosko,  J.E.; Chang, C.C.; Upharn, B.L. and Wilson, M. 1998 Epigenetic toxicology as toxicant-
induced changes in intracellular signaling leading to altered gap junctional intercellular
communication. Toxicol. Lett. 102-103:71-78.

                                         115

-------
SAR Rpvifw Draft; Tin Nnf Tifp ni-
Tugwood, J.D.; Holden, P.R.; James, N.H.; Prince, R.A. and Roberts, R.A. 1998 A PPAR alpha
cDNA cloned from guinea pig liver encodes a protein with similar properties to the mouse PPAR
alpha: Implications for species differences in response to peroxisome proliferates. Arch. Toxicol.
72:169-177.

Ubel, F.A.; Sorenson, S.D. and Roach, D.E.  1980  Health status of plant workers exposed to
fluorochemicals~a preliminary report. Am. Ind. Hyg. Assoc. J. 41:584-589.

Upham, B.L.; Deocampo, N.D.; Wurl, B. and Trosko, I.E.  1998 Inhibition of gap junctional
intercellular communication by perfluorinated fatty acids is dependent on the chain length of the
fluorinated tail. Intl. J. Cancer 78:491-495.

U.S. EPA 1991  Guidelines for developmental toxicity risk assessment.  Federal Register
56(234):63798-63826.

U.S. EPA 1999  Guidelines for carcinogen risk assessment [review draft] MCEA-F-)644. Risk
Assessment Forum, Washington, DC. Available from 

U.S. EPA 2002a Memorandum from Dr. Ralph Cooper, NHEERL, to Dr. Jennifer Seed, dated
October 2, 2002.

U.S. EPA 2002b Memorandum from Dr. Elizabeth Margosches to Dr. Katherine  Anitole, dated
October 21, 2002.

U.S. EPA 2003 Letter from Robert A. Bilott to Richard H. Hefter, dated July 3, 2003. U.S. EPA
AR226-1372; 8EHQ-0703-00373.

Vanden Heuvel, J.P.; Kuslikis, B.I.. and Peterson, R.E.  1991a Covalent binding of perfluorinated
fatty acids to proteins in the plasma, liver and testes of rats. Chem-Biol.  Interact. 82:317-328.

Vanden Heuvel, J.P.; Kuslikis, B.I.; Van Rafelghem, M.L. and Peterson, RE. 1991b Tissue
distribution, metabolism, and elimination of perfluorooctanoic acid in male and female rats. J.
Biochem. Toxicol. 6(2):83-92.

Vanden Heuvel, J.P.; Davis, J.W.; Sommers, R. and Peterson, R.E. 1992 Renal excretion of
perfluorooctanoic acid in male rats: Inhibitory effect of testosterone.  J. Biochem. Toxicol.
7(l):31-36.

Walrath J. and Burke, C. 1989 An investigation into the occurrence of leukemia at Washington
Works. E. I. Dupont De Nemours and Company, April 1989. US EPA AR226-1308-1.

Welter, A.N.  1979 Technical report summary - final comprehensive report: FC-143. 3M
Company. U.S. EPA AR-226 528.

Wolter, J. 1993  Impinger studies of volatility of FC-95 and FC-143.  St. Paul, MN. 3M
Environmental Laboratory. Lab Request Number L3306. U.S. EPA AR 226-1030a087.

Yager, J.D. Jr. and Yager, R. 1980 Oral contraceptive steorids as promoters of
hepatocarcinogenesis in female Sprague-Dawley rats. Cancer Res. 40:3680-3685.
                                         116

-------
SAR RPVJPW Draft; Tin Nnf Tito nr
Yang, Q., Xie, Y. and Depierre, W. 2000 Effects of peroxisome proliferators in the thymus and
spleen of mice.  Clin. Exp. Immunol. 122:219-226.

Yang, Q.; Xie, Y.; Ericksson, A.M.; Nelson, B.D. and DePierre, J.W. 2001 Further evidence for
the involvement of inhibition of cell proliferation and development in thymic and splenic atrophy
induced by the peroxisome prohferator perfluoroctanoic acid in mice. Biochem. Pharmacol.
62:1133-1140.

Yang, Q.; Xie, Y.; Alexson, S.E.H.; Nelson, B.D. and DePierre, J.W. 2002a Involvement of the
peroxisome proliferator-activated receptor alpha in the immunomodilation caused by peroxisome
proliferators in mice.  Biochem. Pharmacol. 63:1893-1900.

Yang, Q.; Abedi-Valugerdi, M.; Xie, Y.; Zhao, X.; Moller, G; Nelson, B.D. and DePierre, J.W.
2002b  Potent suppression of the  adaptive immune response in mice  upon dietary exposure to the
potent peroxisome prohferator, perfluorooctanoic acid. Internal Immunopharmacol. 2:389-397.

Ylinen, M.; Hanhijarvi, H.; Jaakonaho, I. and Peura, P. 1989  Stimulation by estradiol of the
urinary excretion of perfluorooctanoic acid in the male rat. Pharmacol. Toxicol. 65:274-277.

Ylinen, M.; Kojo, A.; Hanhijdrvi, H. and Peura, P.  1990 Disposition of perfluorooctanoic acid in
the rat after single and subchronic administration.  Bull. Environ. Contam. Toxicol. 44:46-53.

York, RG. 200;  Oral (gavage) two-generation (one litter per generation) reproduction study of
ammonium perfluorooctanoic (APFO) in rats. Argus Research Laboratories, Inc. Protocol
Number: 418-020, Sponsor Study Number: T-6889.6, March 26, 2002. U.S. EPA AR226-1092.
                                          117

-------
SAR RPVJPW Draft; Tin Nnt dtp nr Oiintp
             Appendix A: Predicting serum concentrations in rats

A-1.0. Introduction

Analysis of pharmacokinetics for PFOA could be done using a number of approaches including
physiologically based pharmacokinetic (PBPK) modeling, classical compartmental modeling, and
non-parametric analysis.  Each of these has strengths and limitation given the available data.
Some of the literature and submitted pharmacokinetic studies have been analyzed using non-
parameteric analyses. This provides a description of the data that have been collected, but has
fairly limited ability to make predictions across species or variations in exposures. Classical
compartmental modeling can be used to analyze the existing data on blood concentrations in rats,
monkeys, and humans. Much of the published analyses use this approach and the data in humans
and perhaps monkeys is only adequate to support this  approach. Comparisons of serum protein
binding across species indicated a high degree of binding in all species eliminating the apparent
need to address this factor in the compartment modeling.  PBPK modeling is perhaps the ideal
approach for addressing PFOA for purposes of cross-species extrapolation.  Extensive
pharmacokinetic studies have been undertaken in rodents demonstrating complex phenomenon
such as high tissue concentrations in liver, kidney and serum and enterohepatic recirculation of
the parent compound. These could be addressed using PBPK modeling for the rodents, but the
more limited information in monkeys and humans would either require substantial assumptions or
preclude use of this approach. In light of the documented differences in clearance of PFOA
across sexes in rats  and across species, compartmental modeling of serum concentrations
provides a sound approach for estimating internal dosimetry without exceeding the limits of the
available data, so this approach was selected for this risk assessment.

The calculations (see Sections 5.4 and 5.5) of margins of exposures (MOEs) using human and
rodent plasma concentrations or the Area Under Concentration curve in plasma (AUC) require
either direct measurements or prediction of these concentrations and AUCs. AUC is a measure of
dose that reflects both concentration and time; i.e., it increases with higher concentrations and
longer durations. The AUCs estimated here are the average daily AUC under conditions of
repeated dosing. Concentrations in plasma at the time of measurement reflect the time course for
circulating chemical including the maximum levels achieved following a bolus dose; with slow
clearance and repeated dosing they can reflect steady  state conditions.  Pharmacokinetic studies in
Sprague-Dawley rats (Kemper (2003); Han (2003); see Section 3.2) provided information for
parameterizing and calibrating a predictive pharmacokinetic model using standard software
(WinNonlin® Version 4.0.1). Measurements of plasma concentrations in several other studies
with Sprague-Dawley rat (York, 2002; Mylchreest, 2003; Palazzolo, 1993; Goldenthal, 1978a)
provide data for evaluating the model (see Section A-4.0) and in some cases for direct
comparison to human measured plasma concentrations in Section 5.5.

Compartmental models were fitted to the data for individual adult animals in Kemper (2003) (see
Section A-5.0).  WinNonlin® Version 4.0.1 was applied to oral plasma time course data to
estimate and compare one-compartment and two-compartment models with both absorption and
elimination constants. Similar fitting was done for the intravenous plasma time course data.
Akaike Information Criteria and correlations of observed and predicted values at each dose
holistically afforded a basis for the comparisons of the fits and the one- and two-compartment
models. For either  sex, in the range of dosing of most interest, there was little difference among
the models, although there is a large difference in elimination rate between adult male and female
rats.
                                          A-l

-------
SAR RPVJPW Draft; Tin Not fito nr Onnfp
A one-compartment model was selected for predicting plasma concentrations and AUCs due to its
ease of use and consistency with the data. The plasma concentration in such a model with first-
order oral absorption and excretion can be expressed (see O'Flaherty, 1981, equation 3.42) as:

                         C = -^(e-'.'-e-'.<).
where C is the plasma concentration at a given time (mg/L), D is the dose (mg), Vis the volume
of distribution (L), ka is the absorption rate constant (/hr), k, is the elimination rate constant (/hr),
and t is time (hr).  Then AUC = D/(Vk)\ this is expressed in mg-hr/L or equivalentiy, ug-hr/mL.
Sex-specific averages of V, ke and ka  were calculated across the one-compartment model fits to
all the orally-administered doses.  Further, AUC can be expressed in terms of the frequently
reported body-weight-adjusted values for dose and volume of distribution. D can be expressed as
D = body weight x d, where <5is the administered dose rate in mg/kg per day, and V can be
written V= body weight x /?, where ft is a per-unit-weight value in L/kg. Thus, AUC =
also in mg-hr/L or ug-hr/L.
Table A-i
Pharmacokinetic Model Values for Adult Rats

Male Rat
Female Rat
/J(L/kg)
0.172
0.168
ka(/hr)
3.87
3.73
ke(/hr)
0.0058
0.216
ti/z (hr)
119
3.2
Estimates of AUC were determined for the doses, exposure durations, and life stages relevant to
the toxicity studies for which calculations were planned for Section 5. For gavage dose studies,
the oral dose received by the animals was known. For dietary studies, the daily dose was
estimated based upon food consumption and chemical concentration in the diet.  AUCs using the
parameter values as described above were predicted for toxicity studies involving adult male and
nonpregnant female rats. The two-generation reproductive toxicity study includes males,
pregnant females, lactating females, developing fetuses, lactationally exposed offspring, and
juvenile offspring, so estimates of AUC are desirable for each of these lifestages. Estimates of
AUC for gavage-dosed males, pregnant females, and lactating females used the parameter values
as described above. As indicated in Section A-4.0, plasma data from pregnant and lactating
females were reasonably predicted by the modeling without adjustments for these lifestages.
Owing to the evidence in the Han (2004) study with juvenile animals that the clearance in males
drops dramatically with sexual maturation after week 4, these data were used to re-estimate
weekly values of kt for gavage dosed juvenile males and females. No estimates for AUC for
lactationally-exposed offspring were  developed owing to uncertainties in the dose they received,
as discussed in Section 5.

A-2.0. Adult Male and Female Rats

A-2.1. Adult Male Rats

Based on Kemper (2003), adult male rats have relatively slow clearance, with average half life
just under a week (about 5 days). It takes 21 days to be within 95% of steady state levels.
Consequently, adult male rats in toxicity studies longer than 21 days would be expected to be at
steady state, though neither the single dose nor the extended course pharmacokinetic studies
would have achieved steady state.  Data from the subchronic toxicity study by Palazzolo (1993),
                                          A-2

-------
SAR RPVJPW Hr-flfti F>n Nnf fite ni-
which extended 14 weeks, can be used to evaluate the adequacy of predictions based upon single
doses for repeated dosing of male rats to steady state (see Section A-4.0).

At steady state,

      state or Qs = (dose rate mg/kg)/((male ke /hour) *(male pi/kg) *(24hour/day))
     = 24*(conc at steady state) — 24*CSS
TableA-ii.a.
Values for Adult Male Rats
administered dose
(mg/kg)
AUC (ng-hr/mL)
Q5(mg/L)
1
1011
42
3
3032
126
10
10108
421
30
30324
1263
A-2.2. Adult Female Rats

Females have rapid clearance, with an estimated half-life of 0.134 days or 3.2 hours. The female
will clear virtually the entire gayage or dietary dose each day, so plasma concentrations will vary
throughout the day, but successive days will appear similar. The process does not reach steady
state, but an AUC can be estimated.
Table A-ii.b.
Values for Adult Female Rats
administered dose
(mg/kg)
AUC (ng-hr/mL)
1.6
44.16
3
82.80
10
276.01
30
828.03
A-3.0. Post-weaning, Juvenile Rats

To estimate AUCs for the gavage dosed juvenile rats in the two-generation toxicity study, the
data from Han (2003) were used to estimate ke in a week- and sex-specific manner. Supposing
the adult per-unit-weight volume of distribution /? estimated from Kemper (2003) holds for the
offspring, the elimination constant can be obtained from the Han (2003) data using
where C(T) is the concentration at 24 hours after gavage dosing, d= 10 mg/kg as used in Han
(2003), and /?is the adult sex-specific weight-adjusted volume of distribution from the
WinNonlin® solutions to Kemper (2003). Absorption is long completed by 24 hours, so an
instantaneous absorption assumption is not very influential. That is, the concentration equation
above is essentially the same as the one in Section A-1.0, when ka is large relative to ke.  In
Section A-5.0 it will be seen kjkt is about three orders of magnitude for male adult rats and one
and a half for female adults. The impact of their relative size has been further verified for the rats
by successfully re-estimating the 24-hour plasma concentrations of Han (2003) using the
expanded equation, the weekly elimination rates,  and the adult absorption rates (results not
shown).
                                          A-3

-------
SAB KPVIPW Draft? f>n Nnt Tito nr Quote
Then,

            = (-ln(pL/kg*(C(24 hr) mg/L)))/d mg/kg)/(24 hr)

This sex-specific k, is then used with the equation AUC = D/(Vk)= (d/(fk)) to obtain the AUCs.

A-3.1 Post-weaning, Juvenile Male Rats

For the males, half lives range from a third of a day to 2.5 days throughout the post-weaning
period of four to eight weeks of age in Han (2003). The longer half-life observed from weeks 5
to 8 would require 10 days to be within 95% of steady state. Treating the juvenile males as if
they are at steady state is, thus, an approximation till steady state would be achieved around week
6 or 7.

             AUC =24hr* 5mg/kg/day /(pL/kg *ke/hr* 24 hr/day)

Table A-iii. Post-weaning Male Rats, AUC, using 0.172 = adult male
dose rate
4wk AUC (jig-hr/mL)
5wk AUC (ng-hr/mL)
6wk AUC (ng-hr/mL)
7wk AUC (jig-hr/mL)
8wk AUC (M,g-hr/mL)
mean wks4-8
1
67.40
354.25
468.13
310.66
339.11
307.91
3
202.19
1062.76
1404.39
931.98
1017.32
923.73
10
673.96
3542.53
4681.29
3106.62
3391.07
3079.09
30
2021.87
10627.59
14043.85
9319.84
10173.20
9237.27
A-3.2.  Post-weaning, Juvenile Female Rats

For the females, ke estimated in a week-specific manner from Han (2003) can also be applied. A
sex-specific ke is again used with the equation AUC = D/(Vk)= (di(fik)) to obtain the AUCs.
Half-lives of the female pups range about 3.5-5.5 hours, so steady state would not be achieved
following gavage dosing.

                 AUC = 5mg/kg/(pL/kg *ke/hr)

Table A-iv Post-weaning Female Rats, AUC, using 0.168 = adult female
dose (daily)
4wkAUC(ng-hr/mL)
5wk AUC (ng-hr/mL)
6wk AUC dig-hr/mL)
7wk AUC (ng-hr/rnL)
8wk AUC (ng-hr/mL)
mean wks4-5
1
46.11
36.07
36.47
30.76
33.28
41.09
3
138.33
108.21
109.40
92.28
99.82
123.27
10
461.11
360.69
364.67
307.60
332.75
410.90
30
1383.34
1082.06
1094.01
922.81
998.25
1232.70
A-4.0. Validation of One-compartment Model for Predicting Dose Metrics

      The parameters for the one-compartment pharmacokinetic model applied to PFOA were
derived from fits using WinNonlin®4.0.1 with single oral dose pharmacokinetic studies in male
and female Sprague-Dawley rats (Kemper 2003).  These studies measuring plasma time courses
used doses of 0.1,1 (two different studies with different durations), 5, and 25 mg/kg.  The
average parameter values across these five data sets were used when making predictions. A
                                         A-4

-------
SAR RPVIPW Draft; Dn Nnf Cite nr
single intravenous dose (1 mg/kg) in male and female Sprague-Dawley rats provides independent
estimates of the volume of distribution and elimination rate that are reasonably consistent with the
average values from the oral studies (Kemper, 2003). Another Kemper (2003) study in male and
female Sprague-Dawley rats showed no major changes in the extent of total urinary and fecal
elimination for a 1 mg/kg oral dose given following 14 days of dosing at 1 mg/kg/day (i.e., there
were no apparent pharmacokinetic changes due to repeated dosing; see Section 3.2.3.5.2.1).
Additional urinary and fecal elimination studies following single doses are available, but have not
yet been subjected to quantitative analysis using the one-compartment model.

       The parameters for males and females can be used to predict blood plasma concentrations
that were measured in several of the toxicity studies. Plasma levels following gavage dosing also
have been reported for adult female rats in the tissue distribution studies of pregnant rats and
lactationally exposed offspring (Mylchreest, 2003).  Parental male and female Sprague-Dawley
rats in the 2-generation study had blood sampled at the end of mating and lactation, respectively
(York,  2002). Two dietary studies exposed rats for 90 days at the end of which blood samples
were obtained.  One study used male Sprague-Dawley rats (Palazzolo, 1993). The other used
male and female Sprague-Dawley rats (Goldenthal,  1978a).

       Different analyses are required for predicting the plasma concentrations in these studies
using the one-comparment model. Owing to rapid clearance in females, it is important to predict
the concentration at the appropriate time following dosing, even when it is measured in a repeated
dosing study. Modeling dietary exposure for females is likely to be somewhat problematic using
a simple bolus approach. For males, their relatively slow clearance would result in achieving
steady  state plasma levels following sufficient duration of dosing; it should make little difference
whether the exposure was by gavage or diet.

       The plasma concentrations in pregnant and lactating rats 2 hrs after dosing are reasonably
well predicted using the average values from the oral pharmacokinetic studies in females for the
volume of distribution, elimination rate, and oral absorption rate (Mylchreest, 2003).  The
predicted concentrations at the higher doses exceed the measured levels by less than a factor of
two, while the lower dose is well predicted. Small differences in the kinetic parameters would be
expected for pregnant animals; pregnancy might lower their predicted plasma concentrations
(owing to a larger volume of distribution, reduced protein binding  and increased clearance).
Lactating females eliminate some fraction of their body burdens in their milk, which may also
reduce their plasma concentrations relative to these predictions.

Table A-v.  Predicting Mylchreest (2003) 2-hour dam plasma concentrations (gavage)*
Time
(far)
2


Daily Dose
(mg/kg)
3
10
30
predicted C(2hr)
(ug/mL)
12.29
40.96
122.87
measured
(ug/mL)
11.19
26.84
66.64
measured SD
2.76
4.21
9.8
    *C(2hr) calculated from

             c(0 =
kaD
         ,-u
                     V(ka-ke}
       The plasma concentrations 24 hrs following gavage dosing of lactating female rats also
are reasonably well predicted at both the 10 and 30 mg/kg/day dose levels (York, 2002).
                                          A-5

-------
SAR RPVJPW Draft; F>n Nnt fite nr
Table A-vi.  Predicting York (2002) 24-hr dam plasma concentrations (gavage)*
Time (hr)
24

Daily Dose
(mg/kg)
10
30
predicted C(24 hr)
(ug/mL)
0.35
1.06
measured
(ug/mL)
0.37
1.02
measured SD
0.0805
0.425
  *C(24hr) calculated from
       Steady state plasma concentrations in male rats exposed via diet for 90 days were
predicted within a factor of two at a wide range of doses (Palazzolo, 1993).

Table A-vii. Predicting Palazzolo (1993) male plasma concentrations (diet)*
Time
(hr)
24



Diet
(ppm)
1
10
30
100
Dose Rate
(mg/kg/day)
0.06
0.64
1.94
6.5
Dose Rate SD
(mg/kg/day)
0.013
0.138
0.408
1.214
predicted Css
(ug/mL)
3
27
82
274
measured avg 5, 8,. 14 wks
(ug/mL)
7.0
47.4
87.0
148.7
   •"calculated as Css (mg/L) = (dose rate mg/Kg/day)/((pL/Kg) *(ke /hour) *(24hour/day))

       These data and predicted levels provide confidence that the model using parameters
independently determined in the pharmacokinetic study is able to predict dose metrics in the
toxicity studies.

       There are two groups of data that are not successfully predicted, the male plasma
concentrations from the York (2002) and Goldenthal (1978a) studies. It is important to note that
the measured concentrations in these studies are not consistent with the results of the Palazzolo
(1993) study, so it would be impossible for a single set of parameter values to predict all these
data sets.  The source of this difference in measured results is unclear, particularly for the York
(2002) results since the female data from that study appear reasonably predicted. The analysis of
plasma from the Goldenthal (1978a) study involved considerably different analytical methods,
which may be an important factor.

       The York (2002) male data reports plasma concentrations at 10 and 3*0 mg/kg/day that are
similar to those observed at 0.6 and 1.9 mg/kg/day (estimated for dietary exposure) in the
Palazzolo (1993) study.  Treating the data as concentrations 24 hrs following a single gavage
gives a reasonable prediction for the lower dose, but not the higher one. Average concentrations
at 24 hr following a single oral dose of 25 mg/kg in Kemper (2003) are 127 ug/mL, clearly
inconsistent with the 45 ug/mL plasma level measured following repeated gavage in York (2002)
males. The males in York (2002) were being mated, but the potential for that to contribute to
changes in pharmacokinetics is unclear.
                                           A-6

-------
SAR RPVJPW Draft! Tin Nnt fifp nr
Table A-viii. Predicting York (2002) 24-hr male plasma concentrations (gavage)*
Time (hr)


24

Dose Rate
(mg/kg/day)

10
30
predicted Css
(ug/mL)

421
1263
measured
(ug/mL)

51.1
45.3
measured SD


9.3
12.6
predicted C(24hr)
(single dose)
(ug/mL)
50.77
152.31
                  ow
from a single dose as
                         kaD
                                                                   _ e
                                                                      -*„<•
       The plasma data in male rats exposed by diet for 90 days should similarly represent
pseudo-steady state concentrations (Goldenthal, 1978a). Plasma from these animals was analyzed
and reported separately by Belisle (1978). While the plasma levels at the lowest dietary exposure
are well predicted, the measured plasma values at higher dietary levels show little increase with
dose and are not well predicted.
                                                   \
Table A-ix.  Predicting Goldenthal (1978a) male plasma concentrations (diet)*
Time (hr)
24




Diet (ppm)
10
30
100
300
1000
Daily Dose
(mg/kg)
0.6
1.7
5.6
17.9
63.5
predicted Css (ug/mL)
25
72
236
754
2674
measured
(ug/mL)
21
34
36
38
49
•"calculated as Css (mg/L) = (dose rate mg/kg/day)/((/}L/kg)*(ke/hour)*(24hour/day))
       The plasma data from female rats exposed by diet for 90 days are more difficult to predict
owing to the rapid clearance and the uncertainty in when the rats ate, so these data have not been
considered.
                                           A-7

-------
SAR RPVIPW llraft; Dn Nnt f"ifp nr Oiinfp
                                  TABLES





                   A-5.0.  Results of the WinNonlin® Analyses
                                     A-8

-------
   SAB Review Draft; Do Not Cite or Quote


                             ®
A-5.0.  Results of the WinNonlin Analyses



The WinNonlin solutions corresponding to V, ka and ke are denoted in its output by V_F,

KOI, and K10, respectively, for the 1-compartment model (additional different coefficients are

estimated for a 2-compartment model as shown in tables in A-5.0 for these analyses).
                                            A-9

-------
SAB Review Draft; Do Not Cite or Quote



  Table A-x. Female Rats, One-Compartment Model
Variable |N |Nmiss
0.1 mg/kg female
V F
K01
K10
BW
beta (L/kg)
3
3
3
4

0
0
0


Nobs

3
3
3


Mean

28.31038
6.522584
0.190299
176.4
0.16049
0.1 mg/kg female (extended time course)
V F
K01
K10
BW
beta (Ukg)
4
4
4
4

0
0
0


4
4
4


1.0 mg/kg female intravenous
V(mL)
K10(1/hr)
BW
beta (Ukg)
4
4
4

0
0


1.0 mg/kg female oral
V F
K01
K10
BW
beta (Ukg)
4
4
4
4

0
0
0


5 mg/kg female oral
V F
K01
K10
BW
beta (Ukg)
4
4
4
4

0
0
0


25 mg/kg fema e oral
V F
K01
K10
BW
beta (Ukg)
4
4
4


0
0
0


4
4



4
4
4



4
4
4



4
4
4


26.85968
1.731362
0.221401
201
0.13363

44.7764
0.1703
200.6
0.223212

35.77252
4.989439
0.161888
197
0.181586

31.74406
2.152185
0.334105
187.3
0.169482

37.75743
3.941804
0.171864
194.8
0.193827
SD

0.953683
1.525262
0.035546
8.6


5.712687
0.732875
0.042599
11.8


4.3264
0.029
2.7


7.683779
2.068961
0.021 4
9.7


13.54995
1.982879
0.121041
3.6


9.235502
1.961216
0.009952
14.4

SE

0.550609
0.88061
0.020522



2.856344
0.366437
0.021299



2.1632
0.0145



3.841889
1.034481
0.0107



6.774977
0.99144
0.06052



4.617751
0.980608
0.004976


Variance

0.909511214
2.326424021
0.001263511



32.63479772
0.537105296
0.001814668



38.7175
0.0008



59.04045812
4.280600063
0.000457953



183.6012501
3.931809422
0.014650842



85.29449457
3.846368692
9.90461 E-05


Win

27.26332
4.821422
0.151857



21.29838
1 .034098
0.172791



41.0724
0.1391



24.93616
2.599878
0.148536



18.41039
0.412166
0.230922



24.25588
2.886762
0.159052


Median

28.5385
6.978249
0.19706



25.76741
1.717244
0.220222



43.8474
0.1713



37.81364
5.147627
0.152607



30.6327
2.072876
0.31729



41.51663
2.99903
0.174024


Max

29.12932
7.768082
0.221978



34.60553
2.456863
0.272367



50.3385
0.1995



42.52666
7.062624
0.193803



47.30047
4.050822
0.470918



43.74058
6.882396
0.180355


Range

1.865992
2.94666
0.070121



13.30716
1.422765
0.099576



9.2661
0.0603



17.59049
4.462746
0.045267



28.89008
3.638655
0.239996



19.48471
3.995634
0.021303


CV%

3.368669
23.38432
18.67902



21.26863
42.32936
19.24064



9.6622
17.0288



21.47955
41.46681
13.21891



42.685
92.13331
36.22836



24.46009
49.75427
5.79074


Geometric

28.29959
6.393595
0.187981



26.42605
1.61078
0.218303



44.623
0.1684



35.07346
4.633671
0.160907



29.53732
1.302084
0.317682



36.74818
3.65612
0.171644


Harmonic h

28.28871
6.257274
0.18559



26.01866
1.498008
0.215218



44.4733
0.1665



34.30106
4.272174
0.160005



27.48214
0.787231
0.302361



35.58839
3.452334
0.171421


Pseudo SC

0.970518
1.778985
0.038933



5.13963
0.656779
0.042948



4.1263
0.0291



9.588217
2.298085
0.018132



12.25893
0.794923
0.107012



12.54934
1.045302
0.010261


Mean Log

3.342847
1.855297
-1.671412



3.27435
0.476718
-1.52187



3.7982
-1.7814



3.557445
1.533349
-1.826932



3.385655
0.263966
-1.146704



3.604089
1.296402
-1.762331


SDLog

0.033893
0.250227
0.194163



0.206269
0.444473
0.194633



0.0952
0.1726



0.237019
0.45924
0.124934



0.443016
1.256424
0.367695



0.281402
0.422148
0.058639


CV% Geor

3.390233
25.41958
19.60074



20.84821
46.73563
19.64913



9.5431
17.3878



24.03874
48.4551 7
12.54231



46.56689
196.1674
38.04798



28.7066
44.16727
5.868926


Skewness

-0.414295
-0.499847
-0.336812



0.567924
0.017135
0.078793



0.4393
•0.0523



-0.77394
-0.147969
1.115285



0.14896
0.017262
0.16179



-0.993507
1.15179
-0.443669


Kurtosis

-1.5
-1.5
-1.5



-1.079306
-1 .93209
-1.319645



-1.4204
-1.7964



-0.951693
-1.656947
-0.697342



-1.689577
-1.975965
-1.786693



-0.815105
-0.668879
-1.416326


                                                                    A-10

-------
SAB Review Draft; Do Not Cite or Quote
 Table A-xi. Female Rats, Two-Compartinent Model
Variable
N
1 mg/kg oral
V1 F
V2 F
K01
CL F
CLD2 F
4
4
4
4
4
5 mg/kg oral
V1 F
V2 F
K01
CL F
CLD2 F
3
3
3
3
3
25 mg/kg oral
V1 F
V2 F
K01
CL F
CLD2 F
LT
4
4
4
4
4
4
Nmiss

0
0
0
0
0

0
0
0
0
0

0
0
0
0
0
0
Nobs

4
4
4
4
4

3
3
3
3
3

4
4
4
4
4
4
Mean

18.6234
9.0363
2.7225
6.0166
22.5062

29.4424
15.4949
1.6408
9.3461
0.2948

37.9011
81.1708
13.3420
6.3948
0.0520
0.0599
SD

21.3178
6.1941
2.7532
1.7649
15.8217

15.5140
20.3798
2.0681
1.1661
0.0961

8.5209
126.4525
18.3460
1.4077
0.0517
0.1197
SE

10.6589
3.0970
1.3766
0.8825
7.9109

8.9570
11.7663
1.1940
0.6733
0.0555

4.2605
63.2263
9.1730
0.7039
0.0259
0.0598
Variance

454.4486
38.3667
7.5800
3.1150
250.3273

240.6835
415.3361
4.2771
1.3598
0.0092

72.6061
15990.2410
336.5755
1.9816
0.0027
0.0143
Min

0.1737
2.0499
0.3420
3.9358
9.7486

18.7050
3.0245
0.4126
8.4231
0.1958

25.4681
3.0616
2.7921
4.3052
0.0022
0.0000
Median

16.9581
9.2846
2.2161
5.9444
17.3694

22.3928
4.4472
0.4814
8.9586
0.3009

41.1476
26.0788
4.9360
6.9487
0.0421
0.0001
Max

40.4038
15.5262
6.1158
8.2420
45.5376

47.2294
39.0132
4.0285
10.6567
0.3877

43.8412
269.4641
40.7041
7.3764
0.1218
0.2395
Range

40.2301
13.4762
5.7739
4.3062
35.7890

28.5244
35.9887
3.6160
2.2336
0.1919

18.3732
266.4025
37.9120
3.0712
0.1196
0.2395
CV%

114.4676
68.5465
101.1273
29.3344
70.2993

52.6926
131.5255
126.0400
12.4771
32.5940

22.4820
155.7857
137.5052
22.0134
99.3637
199.6773
Geometric
Mean

3.2585
6.9706
1.4950
5.8175
19.0904

27.0454
8.0658
0.9283
9.2992
0.2837

37.0614
25.4384
6.9181
6.2576
0.0254
0.0004
Harmonic
Mean

0.5067
4.9841
0.8091
5.6171
16.6218

25.1483
5.1624
0.6317
9.2540
0.2725

36.1080
9.2358
4.5843
6.1006
0.0078
0.0000
Pseudo
SD

1.4802
7.8707
1.2047
1.8777
10.1250

10.1302
4.3887
0.4303
1.0768
0.1066

11.2420
30.9321
3.4425
1.8870
0.0620
0.0002
Mean Log

1.1813
1.9417
0.4021
1.7609
2.9492

3.2975
2.0876
-0.0743
2.2299
-1.2598

3.6126
3.2363
1.9341
1.8338
-3.6724
-7.7893
SDLog

2.8296
0.9196
1 .3888
0.3036
0.6438

0.4911
1 .3786
1.2735
0.1220
0.3453

0.2546
1.8883
1 .2556
0.2509
1 .7350
4.4339
CV%
Geometric
Mean

5476.9987
115.2987
242.4962
31.0689
71.6701

52.2281
238.5370
201.5338
12.2418
35.5869

25.8795
586.2240
195.9054
25.4927
439.1947
1857561.7895
Skewness

0.0528
-0.0805
0.3257
0.1398
0.9516

0.6624
0.7032
0.7062
0.5431
-0.1164

-0.9811
1.1056
1.1165
-1.0813
0.5552
1.1547
Kurtosis

-1.9296
-1.6551
-1.5664
-1.0179
-0.8009

-1.5000
-1.5000
-1.5000
-1.5000
-1.5000

-0.8231
-0.7073
-0.6990
-0.7155
-1.1206
-0.6667
                                                                    A-ll

-------
SAB Review Draft; Do Not Cite or Quote
     Table A-xii. Female Rats, Comparisons of One- and Two-Compartment Models
       Individual Animal Results within Dose Groups

0.1 mg/kg female





one compartment
Correlation
(obs vs pred)
0.94
0.955
0.9463


0. 1 mg/kg female (extended time course)





1.0 mg/kg female intravenous




1.0 mg/kg female oral





5 mg/kg female oral





25 mg/kg female oral





0.9748
0.9946
0.9965
0.9914


0.9896
0.9847
0.9748
0.9538

0.9654
0.9826
0.9952
0.9949


0.9868
0.9846
0.9816
0.9946


0.9348
0.9641
0.9468
0.9873

AIC
-15.90436
-18.71725
-13.88521



-22.51555
-32.64225
-30.0737
-27.13124


-2.59243
6.47658
5.75399
14.09414

13.33264
4.95944
3.2396
-7.62189


40.17303
37.67916
36.31119
26.33912


123.12264
93.547
99.72258
76.97422





































two compartment
Correlation
(obs vs pred)












;




0.9952
0.9828
0.9981
0.9949


0.9846
0.9815
0.9947



0.9639
0.9641
0.9467
0.9873

AIC

















-1.90618
8.82158
-3.38119
-3.64701


41.67803
40.29375
30.01372



121.62356
99.5497
105.71465
82.97019

                                              A-12

-------
SAB Review Draft; Do Not Cite or Quote
     Table A-xiii. Male Rats, One-Compartment Model
Variable
N
Nmiss
0. 1 mg/kg male
V F
K01
K10
BW
beta (Ukg)
4
4
4
4

0
0
0


Nobs

4
4
4


Mean

52.9274
5.5522
0.0041
216.3
0.245
0.1 mg/kg male (extended time course)
V F
K01
K10
BW
beta (Ukg)
4
4
4
4

0
0
0


4
4
4


1 .0 mg/kg male intravenous
V
K10
BW
beta (Ukg)
4
4
4

0
0


1.0 mg/kg male oral
V F
K01
K10
BW
beta (Ukg)
4
4
4
4

0
0
0


5 mg/kg male oral
V f
K01
K10
BW
beta (Ukg)
4
4
4
4

0
0
0


25 mg/kg male oral
V F
K01
K10
BW
beta (Ukg)
4
4
4


0
0
0


4
4



4
4
4



4
4
4



4
4
4


37.4251
9.1214
0.0042
272.5
0.137

32.5522
0.0106
248.3
0.131

36.9875
1.7823
0.0066
248.7
0.149

32.3112
0.6954
0.0068
218.0
0.1482

40.5204
1.5115
0.0071
225.0
0.1801
SD

10.5147
6.6896
0.0008
11.2


4.3777
4.4205
0.0009
5.8


1.9014
0.0015
9.3


5.0695
1.7370
0.0025
1.6


4.8117
0.1078
0.0017
3.4


2.9308
0.6513
0.0020
3.7

SE

5.2573
3.3448
0.0004



2.1888
2.2103
0.0004



0.9507
0.0008



2.5348
0.8685
0.0013



2.4058
0.0539
0.0008



1.4654
0.3256
0.0010


Variance

110.5579
44.7506
0.0000



19.1639
19.5409
0.0000



3.6153
0.0000



25.6999
3.0173
0.0000



23.1522
0.0116
0.0000



8.5897
0.4242
0.0000


Mm

44.1028
0.4407
0.0032



34.0091
3.0705
0.0035



29.9308
0.0085



33.5521
0.6855
0.0050



26.2113
0.5956
0.0049



36.8396
0.6062
0.0043


Median

50.6338
3.5082
0.0041



36.1936
10.2760
0.0040



32.8975
0.0109



34.9621
1.0358
0.0055



32.5260
0.6730
0.0067



40.8789
1.6718
0.0075


Max

66.3393
14.7519
0.0050



43.3040
12.8631
0.0053



34.4829
0.0121



44.4738
4.3723
0.0103



37.9817
0.8399
0.0090



43.4842
2.0962
0.0090


Range

22.2364
14.3112
0.0019



9.2948
9.7927
0.0018



4.5521
0.0036



10.9217
3.6868
0.0054



11.7704
0.2444
0.0040



6.6447
1.4900
0.0047


CV%

19.8662
120.4849
18.7383



11.6971
48.4629
20.7981



5.8411
14.2809



13.7060
97.4575
38.1633



14.8916
15.5025
24.1979



7.2329
43.0894
27.9451


Geometric
Mean

52.1710
2.3561
0.0040



37.2396
7.9849
0.0041



32.5096
0.0105



36.7458
1.3350
0.0063



32.0354
0.6893
0.0067



40.4397
1.3691
0.0068


Harmonic
Mean

51.4522
1.0479
0.0040



37.0614
6.6324
0.0041



32.4661
0.0104



36.5234
1.0983
0.0061



31.7538
0.6836
0.0065



40.3579
1.1990
0.0066


Pseudo
SD

9.5789
1.5954
0.0008



4.0574
7.4220
0.0008



1.9867
0.0017



4.3194
0.6166
0.0016



5.0832
0.0988
0.0017



3.0151
1.0492
0.0026


Mean Log

3.9545
0.8570
-5.5153



3.6174
2.0776
-5.4902



3.4815
-4.5533



3.6040
0.2890
-5.0667



3.4668
-0.3720
-5.0071



3.6998
0.3141
-4.9859


SDLog

0.1945
1.6751
0.1911



0.1142
0.6600
0.2032



0.0594
0.1502



0.1298
0.8188
0.3365



0.1524
0.1510
0.2441



0.0732
0.5605
0.3187


CV%
Geometric
Mean

19.6356
394.2644
19.2810



11.4535
73.8877
20.5287



5.9472
15.1057



13.0362
97.7228
34.6264



15.3329
15.1834
24.7813



7.3309
60.7491
32.6975


Skewness

0.4188
0.6881
0.0136



0.5927
-0.6576
0.4414



-0.6005
-0.6236



1.0518
1.1136
1.0913



-0.1542
0.5545
0.2402



-0.3107
-0.7038
-0.6739


Kurtosis

-1.4551
-1.1415
-1.0632



-1.2525
-1.1410
-1.4335



-0.9207
-0.9151



-0.7523
-0.6980
-0.7174



-0.9947
-1.1896
-1.0168



-1.3840
-1.0216
-0.8978


                                                                       A-13

-------
SAB Review Draft; Do Not Cite or Quote
 Table A-xiv. Male Rats, Two-Compartment Model
Variable
N
0.1 mg/kg oral
V1 F
V2 F
K01
CL F
CLD2 F
4
4
4
4
4
Nmiss

0
0
0
0
0
Nobs

4
4
4
4
4
0.1 mg/kg oral extended
V1 F
V2_F
K01
CL F
CLD2 F
4
4
4
4
4
0
0
0
0
0
1 mg/kg intravenous
V1
V2
CL
CLD2
4
4
4
4
1 mg/kg oral
V1 F
V2 F
K01
CL F
CLD2 F
4
4
4
4
4
5 mg/kg oral
V1 F
V2 F
K01
CL F
CLD2 F
4
4
4
4
4
25 mg/kg oral
V1 F
V2 F
K01
CL F
CLD2 F
4
4
4
4
4
0
0
0
0

0
0
0
0
0

0
0
0
0
0

0
0
0
0
0
4
4
4
4
4

4
4
• 4
4

4
4
4
4
4

4
4
4
4
4

4
4
4
4
4
Mean

48.4637
11.1415
2.2464
0.2009
0.3049

30.3408
19.1722
2.8750
0.1397
0.6509

32.7068
64.5419
0.1751
0.3502

25.5409
17.3278
0.9458
0.2136
1.1540

25.8013
38.1791
0.5151
0.1594
0.8826

37.8559
33.0604
1.3515
0.2287
0.2269
SD

12.2283
7.6879
2.0255
0.0606
0.1821

9.0159
14.7579
1.1408
0.0497
0.9006

2.4987
53.4475
0.0417
0.1161

9.9245
10.0791
1.1459
0.0561
1.2031

6.9297
14.8007
0.1483
0.0324
1.5119

1.6269
23.1410
0.5876
0.0616
0.1410
SE

6.1141
3.8439
1.0127
0.0303
0.0911

4.5080
7.3790
0.5704
0.0249
0.4503

1.2493
26.7238
0.0208
0.0581

4.9622
5.0395
0.5729
0.0280
0.6016

3.4649
7.4004
0.0742
0.0162
0.7560

0.8135
11.5705
0.2938
0.0308
0.0705
Variance

149.5305
59.1033
4.1025
0.0037
0.0332

81.2872
217.7959
1.3015
0.0025
0.8111

6.2433
2856.6356
0.0017
0.0135

98.4956
101.5874
1.3130
0.0031
1.4475

48.0208
219.0613
0.0220
0.0010
2.2860

2.6468
535.5047
0.3452
0.0038
0.0199
Win

39.3870
2.7865
0.4244
0.1265
0.0413

19.2578
10.0028
1.2007
0.1064
0.1371

30.3884
34.5031
0.1170
0.2216

13.2942
4.4310
0.3389
0.1612
0.0953

16.2165
27.9870
0.3092
0.1184
0.0996

35.5592
11.4257
0.4909
0.1615
0.0985
Median

44.5696
10.1815
2.2807
0.2096
0.3730

30.3915
12.7323
3.2665
0.1202
0.2337

32.3988
39.5514
0.1854
0.3454

26.9683
19.0114
0.3901
0.2004
0.8290

27.7764
32.6645
0.5526
0.1622
0.1402

38.3425
29.3070
1 .5868
0.2362
0.1933
Max

65.3285
21.4164
3.9999
0.2581
0.4324

41.3225
41.2217
3.7664
0.2120
1.9993

35.6412
144.5620
0.2128
0.4883

34.9329
26.8573
2.6642
0.2922
2.8628

31.4357
59.4005
0.6461
0.1946
3.1503

39.1797
62.2020
1.7415
0.2811
0.4225
Range

25.9416
18.6299
3.5755
0.1317
0.3910

22.0646
31.2189
2.5657
0.1056
1.8622

5.2527
110.0588
0.0959
0.2668

21.6387
22.4263
2.3253
0.1310
2.7675

15.2192
31.4135
0.3369
0.0761
3.0507

3.6204
50.7763
1.2505
0.1197
0.3240
CV%

25.2318
69.0023
90.1638
30.1333
59.7312

29.7156
76.9754
39.6809
35.5854
138.3527

7.6396
82.8105
23.7936
33.1568

38.8573
58.1671
121.1510
26.2483
104.2553

26.8580
38.7665
28.7941
20.3126
171.3095

4.2976
69.9961
43.4743
26.9140
62.1400
Geometric
Mean

47.3934
8.8677
1.3974
0.1935
0.2224

29.2785
16.0769
2.6356
0.1340
0.3474

32.6357
52.7869
0.1709
0.3354

23.8838
14.1875
0.6088
0.2085
0.6469

24.9914
36.3128
0.4962
0.1568
0.2802

37.8292
26.8232
1.2093
0.2223
0.1968
Harmonic
Mean

46.4259
6.6436
0.8626
0.1856
0.1252

28.1789
14.2145
2.3383
0.1293
0.2411

32.5654
46.1275
0.1662
0.3208

22.1208
10.6781
0.4732
0.2038
0.2979

24.0855
34.7743
0.4749
0.1541
0.1625

37.8019
21.7307
1.0310
0.2158
0.1721
Pseudo
SD

10.1673
9.8784
0.9233
0.0692
0.4264

10.1562
6.0913
1.9543
0.0349
0.1712

2.4510
1 9.3541
0.0527
0.1180

12.0959
16.3905
0.2306
0.0476
1.1733

8.9814
10.4622
0.1976
0.0356
0.1015

1.6942
18.0979
1.0644
0.0626
0.1117
Mean Log

3.8585
2.1824
0.3346
-1.6427
-1.5032

3.3769
2.7774
0.9691
-2.0099
-1.0573

3.4854
3.9663
-1.7667
-1.0925

3.1732
2.6524
-0.4963
-1.5680
-0.4355

3.2185
3.5922
-0.7007
-1.8531
-1.2721

3.6331
3.2893
0.1901
-1.5038
-1.6256
SDLog

0.2399
0.8478
1.2197
0.3267
1.1290

0.3149
0.6382
0.5284
0.3224
1.1997

0.0760
0.6770
0.2648
0.3436

0.4412
0.8207
0.9864
0.2499
1.4271

0.3037
0.3544
0.3303
0.2125
1.6211

0.0437
0.7691
0.6063
0.2794
0.6157
CV%
Geometric
Mean

24.3361
102.5682
185.1179
33.5623
160.5430

32.2851
70.9071
56.7537
33.0970
179.3915

7.6129
76.2544
26.9549
35.4041

46.3571
98.0334
128.2850
25.3828
258.1437

31.0890
36.5867
33.951 1
21.4907
358.3916

4.3679
89.8223
66.6481
28.4933
67.8954
Skewness

0.7009
0.4231
-0.0023
-0.2982
-0.9355

-0.0194
1.1270
-1.0019
0.9639
1.1417

0.2107
1.1418
-0.7195
0.1087

-0.3050
-0.3978
1.1531
0.7246
0.8059

-0.7071
0.8863
-0.7046
-0.2609
1.1541

-0.7976
0.3596
-1.0041
-0.1164
0.6903
Kurtosis

-1.1286
-0.9616
-1.9969
-1.5302
-0.8716

-1.0069
-0.6864
-0.7601
-0.8445
-0.6768

-1.6772
-0.6770
-0.9901
-1.4102

-1.5179
-1.3841
-0.6678
-0.9303
-O.9071

-1.1008
-0.9251
-1.0405
-1.2287
-0.6671

-0.9871
-1.4551
-0.8037
-1.8463
-1.0213
                                                                    A-14

-------
SAB Review Draft; Do Not Cite or Quote
 Table A-xv.  Male Rats, Comparisons of One- and Two-Compartment Models
   Individual Animal Results within Dose Groups

0.1 mg/kgmale





one compartment
Correlation
(obs vs pied)
0.9157
0.9605
0.9272
0.9497

0.1 mg/kg male (extended time course)





1.0 mg/kg male intravenous




1.0 mg/kg male oral





5 mg/kg male oral





25 mg/kg male oral





0.8177
0.9829
0.9804
0.9757


0.79
0.956
0.9114
0.9338

0.9267
0.9266
0.9162
0.9413


0.9576
0.9399
0.8966
0.9668


0.973
0.9815
0.9876
0.9732

AIC
-64.4
-83.4
-53.6
-64.2


17.3
-95.7
-96.8
-77.8


126.7
81.6
91.1
89.7

47.7
74.5
67.2
76.5


136
128.4
156.6
139.4


186
187.6
176
184.2





































two compartment
Correlation
(obs vs pred)
0.9471
0.9605
0.9356
0.9595


0.905
0.9903
0.9824
0.985


0.9367
0.9696
0.914
0.9528

0.9711
0.9649
0.9252
0.9366


0.9587
0.9659
0.9043
0.9708


0.9822
0.9857
0.9877
0.9822

AIC
-58.6
-79.4
-51.2
-65.4


-8.1
-125.6
-97.9
-91.55


77.9
72.4
94
82.7

36.2
61.5
69
82.4


138.2
120
158.6
139.7


180
183
178.8
178.2

                                       A-15

-------