A EPA
EPA/635/R-20/424a
External Review Draft
www.epa.gov/iris
Toxicological Review of Perfluorobutanoic Acid [CASRN 375-22-4]
and Related Salts
October 2021
Integrated Risk Information System
Center for Public Health and Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
-------
Toxicological Review ofPFBA and Related Salts
DISCLAIMER
This document is an external review draft for review purposes only. This information is
distributed solely for the purpose of public comment It has not been formally disseminated by
EPA. It does not represent and should not be construed to represent any Agency determination or
policy. Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
This document is a draft for review purposes only and does not constitute Agency policy.
ii DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
CONTENTS
AUTHORS| CONTRIBUTORS| REVIEWERS ix
EXECUTIVE SUMMARY xi
1. OVERVIEW OF BACKGROUND INFORMATION AND ASSESSMENT METHODS 1-1
1.1. BACKGROUND INFORMATION ON PERFLUOROBUTANOIC ACID (PFBA) 1-1
1.1.1. Physical and Chemical Properties 1-1
1.1.2. Sources, Production, and Use 1-3
1.1.3. Environmental Fate and Transport 1-3
1.1.4. Potential for Human Exposure and Populations with Potentially Greater
Exposure 1-5
1.2. SUMMARY OF ASSESSMENT METHODS 1-6
1.2.1. Literature Search and Screening 1-6
1.2.2. Evaluation of Individual Studies 1-9
1.2.3. Data Extraction 1-11
1.2.4. Evidence Synthesis and Integration 1-11
1.2.5. Dose-Response Analysis 1-13
2. LITERATURE SEARCH AND STUDY EVALUATION RESULTS 2-1
2.1. LITERATURE SEARCH AND SCREENING RESULTS 2-1
2.2. STUDY EVALUATION RESULTS 2-2
3. TOXICOKINETICS, EVIDENCE SYNTHESIS, AND EVIDENCE INTEGRATION 3-1
3.1.TOXICOKINETIC S 3-1
3.1.1. Absorption 3-2
3.1.2. Distribution 3-2
3.1.3. Metabolism 3-5
3.1.4. Excretion 3-5
3.1.5. Summary 3-8
3.2. NONCANCER EVIDENCE SYNTHESIS AND INTEGRATION 3-10
3.2.1. Thyroid Effects 3-11
3.2.2. Hepatic Effects 3-23
3.2.3. Developmental Effects 3-42
3.2.4. Reproductive Effects 3-48
This document is a draft for review purposes only and does not constitute Agency policy.
iii DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
3.2.5. Other Noncancer Health Effects 3-51
3.3. CARCINOGENICITY 3-53
4. SUMMARY OF HAZARD IDENTIFICATION CONCLUSIONS 4-1
4.1. SUMMARY OF CONCLUSIONS FOR NONCANCER HEALTH EFFECTS 4-1
4.2. SUMMARY OF CONCLUSIONS FOR CARCINOGENICITY 4-4
4.3. CONCLUSIONS REGARDING SUSCEPTIBLE POPULATIONS AND LIFESTAGES 4-4
5. DERIVATION OF TOXICITY VALUES 5-1
5.1. NONCANCER AND CANCER HEALTH EFFECT CATEGORIES CONSIDERED 5-1
5.2. NONCANCER TOXICITY VALUES 5-1
5.2.1. Oral Reference Dose (RfD) Derivation 5-1
5.2.2. Subchronic Toxicity Values for Oral Exposure (Subchronic Oral Reference Dose
[RfD]) Derivation 5-23
5.2.3. Inhalation Reference Concentration (RfC) 5-25
5.3. CANCER 5-25
5.3.1. Cancer Weight-of-Evidence Descriptor and Derivation of Cancer Risk Values 5-25
REFERENCES R-l
This document is a draft for review purposes only and does not constitute Agency policy.
iv DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
TABLES
Table ES-1. Health effects with evidence available to synthesize and draw summary judgments
and derived toxicity values xiii
Table 1-1. Predicted or experimental physicochemical properties of perfluorobutanoic acid
(PFBA; CASRN 375-22-4) and ammonium perfluorobutanoate (NH4+PFB;
CASRN 10495-86-0) 1-2
Table 1-2. Perfluorobutanoic acid (PFBA) levels in water, soil, and air at National Priority List
(NPL) sites 1-5
Table 1-3. Populations, Exposures, Comparators, and Outcomes (PECO) criteria 1-8
Table 3-1. Serum and liver concentrations of perfluorobutanoic acid (PFBA) following
subchronic or gestational exposure 3-3
Table 3-2. Summary of toxicokinetics of serum perfluorobutanoic acid (PFBA) (mean ± standard
error) 3-10
Table 3-3. Percent change in thyroid hormones due to perfluorobutanoic acid (PFBA) exposure
in short-term and subchronic oral toxicity studies 3-13
Table 3-4. Incidence and severity of thyroid follicular hypertrophy/hyperplasia due to
perfluorobutanoic acid (PFBA) exposure in short-term and subchronic oral
toxicity studies 3-15
Table 3-5. Evidence profile table for thyroid effects 3-21
Table 3-6. Percent increase in relative liver weight due to perfluorobutanoic acid (PFBA)
exposure in short-term and subchronic oral toxicity studies 3-25
Table 3-7. Incidence and severity of liver histopathological lesions due to perfluorobutanoic acid
(PFBA) exposure in short-term and subchronic oral toxicity studies 3-29
Table 3-8. Evidence profile table for hepatic effects 3-39
Table 3-9. Developmental effects observed following perfluorobutanoic acid (PFBA) exposure in
a developmental toxicity study 3-44
Table 3-10. Evidence profile table for developmental effects 3-47
Table 3-11. Evidence profile table for reproductive effects 3-50
Table 4-1. Evidence integration summary for health effects for which evidence indicates a
hazard exists 4-3
Table 5-1. Endpoints considered for dose-response modeling and derivation of points of
departure 5-5
Table 5-2. Benchmark response levels selected for benchmark dose (BMD) modeling of
perfluorobutanoic acid (PFBA) health outcomes 5-6
Table 5-3. Rat, mouse, and human clearance values and data-informed dosimetric adjustment
factors 5-10
Table 5-4. Points of departure (PODs) considered for use in deriving candidate reference values
for perfluorobutanoic acid (PFBA) 5-12
Table 5-5. Uncertainty factors for the development of the candidate values for
perfluorobutanoic acid (PFBA) 5-13
Table 5-6. Comparison of liver-weight effects across species and durations of exposure 5-15
Table 5-7. Candidate values for perfluorobutanoic acid (PFBA) 5-19
Table 5-8. Confidence in the organ/system-specific oral reference doses (osRfDs) for
perfluorobutanoic acid (PFBA) 5-20
Table 5-9. Organ/system-specific oral reference dose (osRfD) values for perfluorobutanoic acid
(PFBA) 5-22
This document is a draft for review purposes only and does not constitute Agency policy.
v DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Table 5-10. Candidate subchronic oral reference dose (RfD) values for perfluorobutanoic acid
(PFBA) 5-24
FIGURES
Figure 1-1. Chemical structures of perfluorobutanoic acid (PFBA) and ammonium
perfluorobutanoate (NH4+PFB) 1-2
Figure 2-1. Literature search and screening flow diagram for perfluorobutanoic acid (PFBA) 2-2
Figure 2-2. Evaluation results for animal studies assessing effects of perfluorobutanoic acid
(PFBA) exposure (see interactive data graphic for rating rationales) 2-4
Figure 2-3. Evaluation results for epidemiological studies assessing effects of perfluorobutanoic
acid (PFBA; interactive data graphic for rating rationales) 2-5
Figure 3-1. Evaluation results for animal studies assessing effects of perfluorobutanoic acid
(PFBA) exposure on the thyroid (see interactive data graphic for rating
rationales) 3-12
Figure 3-2. Thyroid hormone response to ammonium perfluorobutanoate (NH4+PFB) exposure
(see interactive data graphic and rationale for study evaluations for thyroid
hormone effects in Health Assessment Workspace Collaborative [HAWC]) 3-14
Figure 3-3. Thyroid histopathology and organ-weight responses to ammonium
perfluorobutanoate (NH4+PFB) exposure (see interactive data graphic and
rationale for study evaluations for other thyroid effects in Health Assessment
Workspace Collaborative [HAWC]) 3-15
Figure 3-4. Evaluation results for animal studies assessing effects of perfluorobutanoic acid
(PFBA) exposure on the liver (see interactive data graphic for rating rationales) 3-24
Figure 3-5. Liver-weight response to ammonium perfluorobutanoate (NH4+PFB) or
perfluorobutanoic acid (PFBA) exposure (see interactive data graphic and
rationale for study evaluations for liver-weight effects in Health Assessment
Workspace Collaborative [HAWC]) 3-27
Figure 3-6. Liver histopathology response to ammonium perfluorobutanoate (NH4+PFB) or
perfluorobutanoic acid (PFBA) exposure (see interactive data graphic and
rationale for study evaluation for liver histopathology effects in Health
Assessment Workspace Collaborative [HAWC]) 3-30
Figure 3-7. Evaluation results for animal studies assessing developmental effects of
perfluorobutanoic acid (PFBA) exposure (see interactive data graphic for rating
rationales) 3-43
Figure 3-8. Pre- and postnatal developmental responses to gestational ammonium
perfluorobutanoate (NH4+PFB) exposure (see interactive data graphic and
rationale for study evaluations for developmental effects in Health Assessment
Workspace Collaborative [HAWC]) 3-44
Figure 3-9. Reproductive responses to ammonium perfluorobutanoate (NH4+PFB) exposure (see
interactive data graphic and rationale for study evaluations for reproductive
effects in Health Assessment Workspace Collaborative [HAWC]) 3-49
This document is a draft for review purposes only and does not constitute Agency policy.
vi DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
ABBREVIATIONS AND ACRONYMS
ACO
acyl-CoA oxidase
HAWC
Health Assessment Workspace
ADME
absorption, distribution, metabolism,
Collaborative
and excretion
HED
human equivalent dose
AFFF
aqueous film-forming foam
HERO
Health and Environmental Research
AIC
Akaike's information criterion
Online
ALP
alkaline phosphatase
HISA
highly influential scientific information
ALT
alanine aminotransferase
HPT
hypothalamic-pituitary-thyroid
AST
aspartate aminotransferase
IRIS
Integrated Risk Information System
atm
atmosphere
i.v.
intravenous
ATSDR
Agency for Toxic Substances and
IQ
intelligence quotient
Disease Registry
IQR
interquartile range
AUC
area-under-the-concentration curve
ISI
influential scientific information
BMD
benchmark dose
IUR
inhalation unit risk
BMDL
benchmark dose lower confidence limit
LLOQ
lower limit of quantitation
BMDS
Benchmark Dose Software
LN
log-normal
BMR
benchmark response
LOAEL
lowest-observed-adverse-effect level
BW
body weight
MBq
megabecquerel
Cavg
average concentration
MOA
mode of action
Cmax
maximum concentration
NCEA
National Center for Environmental
CA
Cochran-Armitage
Assessment
CAR
constitutive androstane receptor
NCV
nonconstant variance
CASRN
Chemical Abstracts Service registry
NIOSH
National Institute for Occupational
number
Safety and Health
CDR
Chemical Data Reporting
NIS
sodium-iodide symporter
CI
confidence interval
NOAEL
no-observed-adverse-effect level
CL
clearance
NPL
National Priority List
CLa
clearance in animals
NTP
National Toxicology Program
CLh
clearance in humans
OAT
organic anion transporter
CPAD
Chemical and Pollutant Assessment
OECD
Organisation for Economic Co-
Division
operation and Development
CPHEA
Center for Public Health and
OMB
Office of Management and Budget
Environmental Assessment
ORD
Office of Research and Development
CV
constant variance
OSF
oral slope factor
CYP
cytochrome P450 superfamily
PC
partition coefficient
DAF
dosimetric adjustment factor
PBPK
physiologically based pharmacokinetic
DNA
deoxyribonucleic acid
PBTK
physiologically based toxicokinetic
DNT
developmental neurotoxicity
PECO
Populations, Exposures, Comparators,
DOD
Department of Defense
Outcomes
EPA
Environmental Protection Agency
PFAA
perfluoroalkyl acid
EOP
Executive Office of the President
PFAS
per- and polyfluoroalkyl substances
ER
extra risk
PFBA
perfluorobutanoic acid
FLR
full-litter resorption
PFBS
perfluorobutane sulfonate
FTOH
fluorotelomer alcohol
PFCA
perfluoroalkyl carboxylic acid
GD
gestation day
PFDA
perfluorodecanoic acid
GFR
glomerular filtration rate
PFHxA
perfluorohexanoic acid
GGT
y-glutamyl transferase
PFHxS
perfluorohexane sulfonate
GRADE
Grading of Recommendations
PFNA
perfluorononanoic acid
Assessment, Development, and
PFOA
perfluorooctanoic acid
Evaluation
PFOS
perfluorooctane sulfonate
GSH
glutathione
PK
pharmacokinetic
PND
postnatal day
This document is a draft for review purposes only and does not constitute Agency policy.
vii DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
POD
point of departure
TRI
Toxic Release Inventory
PODhed
human equivalent dose POD
TSCA
Toxic Substances Control Act
PPAR
peroxisome proliferator-activated
TSCATS
Toxic Substances Control Act Test
receptor
Submissions
PQAPP
Programmatic Quality Assurance
TSH
thyroid-stimulating hormone
Project Plan
TSHR
thyroid-stimulating hormone receptor
PT
prothrombin time
UCMR
Unregulated Contaminant Monitoring
PXR
pregnane X receptor
Rule
QA
quality assurance
UDP-GT
uridine 5'-diphospho-
QAPP
Quality Assurance Project Plan
glucuronosyltransferase
QMP
Quality Management Plan
UF
uncertainty factor
RBC
red blood cell
UFa
animal-to-human uncertainty factor
RD
relative deviation
UFc
composite uncertainty factor
RfC
inhalation reference concentration
UFd
database deficiencies uncertainty factor
RfD
oral reference dose
UFh
human variation uncertainty factor
RS
Rao-Scott
UFl
LOAEL-to-NOAEL uncertainty factor
SD
standard deviation
UFs
subchronic-to-chronic uncertainty
S-D
Sprague-Dawley
factor
SE
standard error
Vd
volume of distribution
TD
toxicodynamic
VOC
volatile organic compound
TH
thyroid hormone
WOS
Web of Science
TK
toxicokinetic
TPO
thyroid peroxidase
This document is a draft for review purposes only and does not constitute Agency policy.
viii DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
AUTHORS | CONTRIBUTORS | REVIEWERS
Assessment Managers (Lead Authors)
J. Allen Davis, M.S.P.H. U.S. EPA/Office of Research and Development/Center for
Michele M. Taylor, Ph.D. Public Health and Environmental Assessment
Authors
Andrew Kraft, Ph.D. (IRIS PFAS Lead) U.S. EPA/Office of Research and Development/Center for
Jason C. Lambert, Ph.D., DABT Public Health and Environmental Assessment
Elizabeth Radke, Ph.D.
Paul Schlosser, Ph..D
Contributors
Michelle Angrish, Ph.D. U.S. EPA/Office of Research and Development/Center for
Xabier Arzuaga, Ph.D. Public Health and Environmental Assessment
Johanna Congleton, Ph.D.
Ingrid Druwe, Ph.D.
Kelly Garcia, B.S.
Andrew Greenhalgh, B.S.
Carolyn Gigot, B.A.
Mary Gilbert, Ph.D.
Christopher Lau, Ph.D.
April Luke, M.S.
Pam Noyes, Ph.D.
Katherine O'Shaughnessy, Ph.D.
Elizabeth Oesterling Owens, Ph.D.
Tammy Stoker, Ph.D.
Andre Weaver, Ph.D.
Amina Wilkins, M.P.H.
Michael Wright, Sc.D.
Jay Zhao, Ph.D.
Chris Corton, Ph.D. U.S. EPA/Office of Research and Development/Center for
Computational Toxicology and Exposure
Production Team
Maureen Johnson U.S. EPA
Ryan Jones Office of Research and Development
Dahnish Shams Center for Public Health and Environmental Assessment
Vicki Soto
Samuel Thacker
This document is a draft for review purposes only and does not constitute Agency policy.
ix DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Executive Direction
Wayne Casio
Kay Holt
CPHEA Center Director
CPHEA Deputy Center Director
CPHEA Associate Director
CPHEA Senior Science Advisor
CPAD Division Director
CPAD Associate Division Director
CPAD Senior Science Advisor, Integrated Risk Information
System, IRIS PFAS Team Lead
CPAD Senior Science Advisor
CPAD Senior Science Advisor
CPHEA/CPAD/Toxic Effects Assessment (DC) Branch Chief
CPHEA/CPAD/Toxic Effects Assessment (RTP) Branch Chief
CPHEA/CPAD/Science Assessment Methods Branch Chief
CPHEA/CPAD/Quantitative Assessment Branch Chief
CPHEA/CPAD/Emeritus
NJDEP (retired), Contractor
*Executive Review Committee
Samantha Jones
Emma Lavoie
Kristina Thayer*
James Avery
Andrew Kraft
Belinda Hawkins
Paul White*
Ravi Subramaniam*
Janice Lee
Barbara Glenn
Viktor Morozov
Karen Hogan*
Alan Stern*
Reviewers
This assessment was provided for review to scientists in EPA's program and regional offices.
Comments were submitted by:
Office of the Administrator/Office of Children's Health Protection
Office of Air and Radiation/Office of Air Quality Planning and Standards
Office of Chemical Safety and Pollution Prevention/Office of Pollution Prevention and Toxics
Office of Land and Emergency Management
Office of Water
Region 2, New York
Region 3, Philadelphia
Region 8, Denver
This assessment was provided for review to other federal agencies and the Executive Office of the
President (EOP). A summary and EPA's disposition of major comments from the other federal
agencies and EOP is available on the IRIS website. Comments were submitted by:
The White House
Office of Management and Budget
Department of Defense
Department of Health and Human Services
Agency for Toxic Substances and Disease Registry
National Institute for Occupational Safety and Health
This document is a draft for review purposes only and does not constitute Agency policy.
x
DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Toxicological Review ofPFBA and Related Salts
EXECUTIVE SUMMARY
Summary of Occurrence and Health Effects
Perfluorobutanoic acid (PFBA, CASRN 375-22-4)1 and its related salts are members of the
group of per- and polyfluoroalkyl substances (PFAS). This assessment applies to PFBA as well as
salts (including alkali metal salts) of PFBA that would be expected to fully dissociate in aqueous
solutions of pH ranging from 4-9 (e.g., in the human body). Thus, while this assessment would not
necessarily apply to non-alkali metal salts ofPFBA (e.g., silver heptafluorobutyrate; CASRN 3794-
64-7) due to the possibility of PFBA-independent contributions of toxicity, it does apply to PFBA
salts including ammonium perfluorobutanoate (CASRN 10495-86-0), sodium perfluorobutanoate
(CASRN 2218-54-4), potassium heptafluorobutanoate [2966-54-3], and other non-metal or alkali
metal salts ofPFBA. The synthesis of evidence and toxicity value derivation presented in this
assessment focuses on the free acid ofPFBA and ammonium perfluorobutanoate given the
currently available toxicity data
Concerns about PFBA and other PFAS stem from the resistance of these compounds to
hydrolysis, photolysis, and biodegradation, which leads to their persistence in the environment.
PFAS are not naturally occurring in the environment; they are manmade compounds that have been
used widely over the past several decades in consumer products and industrial applications
because of their resistance to heat, oil, stains, grease, and water. PFBA is a breakdown product of
other PFAS that are used in stain-resistant fabrics, paper food packaging and carpets; it was also
used for manufacturing photographic film, and it is used as a substitute for longer chain
perfluoroalkyl carboxylic acids (PFCAs) in consumer products. PFBA has been found to accumulate
in agricultural crops and has been detected in household dust, soils, food products, and surface,
ground, and drinking water. As such, exposure is possible via inhalation of indoor or outdoor air,
ingestion of drinking water and food, and dermal contact with PFBA-containing products.
Human epidemiological studies have examined possible associations between PFBA
exposure and health outcomes, such as thyroid hormones or disease, hepatic enzymes, birth
outcomes (e.g., birth weight, gestational duration), semen parameters, blood lipids, and blood
pressure. The ability to draw conclusions regarding these associations is limited due to the
methodological conduct of the studies (studies were generally considered low confidence for these
outcomes; two studies on congenital hypothyroidism and birth weight and gestational duration
1 The CASRN given is for linear PFBA; the source PFBA used in toxicity studies was reported to be 98% pure
fDas et al.. 2008) and reagent grade fButenhoff et al„ 2012). Neither study explicitly states that only the linear
form was used. Therefore, there is the possibility that a minor proportion of the PFBA used in the studies
were branched isomers and thus observed health effects may apply to the total linear and branched isomers
in a given exposure source.
This document is a draft for review purposes only and does not constitute Agency policy.
xi DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Toxicological Review ofPFBA and Related Salts
were considered uninformative); the small number of studies per health outcome; and the generally
null findings coincident with notable sources of study insensitivity due to lack of detecting
quantifiable levels ofPFBA in blood samples or a narrow concentration range across exposure
groups. No studies were identified that evaluated the association between PFBA exposure and
carcinogenicity.
Animal studies ofPFBA exposure in rats and mice have exclusively examined the oral route
(i.e., no inhalation or dermal studies were identified during the literature search) and have
examined noncancer endpoints only.
Altogether, the available evidence indicates that developmental, thyroid, and liver effects in
humans are likely caused by PFBA exposure in utero or during adulthood. There was inadequate
evidence to determine whether reproductive effects might represent a potential human health
hazard following PFBA exposure.
The few epidemiological studies did not inform the potential for effects in the thyroid, liver,
reproductive system, or developing offspring. Liver effects manifested as increased relative liver
weight in adult animals and increased incidence of hepatocellular hypertrophy. Thyroid effects in
adult exposed rats were expressed through decreases in free and total thyroxine (T4) and increased
incidence of thyroid follicular hypertrophy and hyperplasia. Developmental effects in exposed
animals were expressed as the loss of viable offspring (total litter resorption), and delays in
developmental milestones: eye opening, vaginal opening, and preputial separation.
Table ES-1 summarizes health effects that had enough evidence available to synthesize and
draw hazard conclusions and the toxicity values derived for those health effects.
This document is a draft for review purposes only and does not constitute Agency policy.
xii DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Table ES-1. Health effects with evidence available to synthesize and draw
summary judgments and derived toxicity values
Health system
Evidence
integration
judgment
Toxicity
value
Value PFBA
(mg/kg-d)
Value NH4+
PFB
(mg/kg-d)a
Confidence
UFC
Basis
Hepatic
Evidence
indicates
(likely)
osRfD
1 X 10"3
1 X 10"3
Medium
1,000
Increased
hepatocellular
hypertrophy in adult
rats
Subchronic
osRfD
1 x 10"2
1 x 10"2
Medium
100
Increased
hepatocellular
hypertrophy in adult
rats
Thyroid
Evidence
indicates
(likely)
osRfD
1 x 10"3
1 x 10"3
Medium-low
1,000
Decreased total T4 in
adult rats
Subchronic
osRfD
1 x 10"2
1 x 10"2
Medium-low
100
Decreased total T4 in
adult rats
Developmental
Evidence
indicates
(likely)
osRfD
6 x 10"3
7 x 10"3
Medium-low
100
Developmental delays
in miceb
Subchronic
osRfD
6 x 10"3
7 x 10"3
Medium-low
100
Developmental delays
in miceb
Reproductive
Evidence
inadequate
osRfD
Not derived
Not derived
NA
NA
NA
Subchronic
osRfD
Not derived
Not derived
NA
NA
NA
RfD
1 x 10"3
1 x 10"3
Medium
1,000
Hepatic and thyroid
effects
Subchronic RfD
6 x 10"3
7 x 10"3
Medium-low
100
Developmental effects
RfD = reference dose (in mg/kg-day) for lifetime exposure; subchronic RfD = reference dose (in mg/kg-day) for less-
than-lifetime exposure; osRfD = organ-specific oral reference dose (in mg/kg-day); UFC = composite uncertainty
factor; NA = not applicable.
a See Tables 5-7 and 5-10 for details on how to calculate candidate values for salts of PFBA. The osRfDs presented
in this table have been rounded to 1 significant digit from the candidate values presented in Tables 5-7 and 5-10.
bThe point of departure represents three types of developmental delays observed in the same study.
Chronic Oral Reference Dose (RfD) for Noncancer Effects
1 From the identified human health hazards of potential concern for adults and developing
2 offspring (liver, thyroid, developmental toxicity), increased liver hypertrophy and decreased T4 in
3 adult male rats, as reported in Butenhoff et al. f20121. were selected as the basis for the oral
4 reference dose (RfD). A benchmark dose lower confidence limit (BMDL) of 5.4 mg/kg-day NH4+PFB
5 was identified for increased liver hypertrophy, and a no-observed-adverse-effect level (NOAEL) of
6 6 mg/kg-day NH4+PFB was identified for decreased T4. These values were used as the points of
7 departure (PODs). After converting the PODs from units of mg/kg-day NH4+PFB to units of mg/kg-
This document is a draft for review purposes only and does not constitute Agency policy.
xiii DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Toxicological Review ofPFBA and Related Salts
day PFBA (by multiplying by the ratio of the molecular weights of the free acid and the ammonium
salt), the ratio of serum clearance values between rats and humans was used to account for
toxicokinetic differences between species, resulting in the human equivalent doses (PODhed) of
1.15 mg/kg-day and 1.27 mg/kg-day for increased liver hypertrophy and decreased T4,
respectively. The RfD for PFBA was calculated by dividing the PODhed values by a composite
uncertainty factor (UFc) of 1,000 to account for residual toxicokinetic and toxicodynamic
uncertainty in the extrapolation from rats to humans (UFA =3), interindividual differences in
human susceptibility (UFH = 10), extrapolation from a subchronic-to-chronic duration (UFS = 10),
and deficiencies in the toxicity database (UFd = 3). The selected overall RfD for PFBA derived based
on liver and thyroid effects is 1 x 10~3 mg/kg-day.2,3
Confidence in the Oral Reference Dose (RfD)
The overall confidence in the RfD is medium. The subchronic toxicity exposure study
conducted by Butenhoff et al. (2012) reported on administration of NH4+PFB by gavage to
Sprague-Dawley (S-D) rats for 90 days. This study is rated as high confidence with adequate
reporting and appropriate study design, methods, and conduct (see study evaluation analysis in
Health Assessment Workspace Collaborative [HAWC]).4 Confidence in the oral toxicity database for
derivation of the RfD is medium because consistent and coherent effects occurred within both
individual organ systems used to support the RfD, although important uncertainties remain.
Confidence in the quantification of the PODs supporting the RfD is medium, given the use of BMD
modeling within the observed range of the data for liver effects, use of a NOAEL roughly equivalent
with a decrease of one standard deviation for thyroid effects (suggesting that this POD might not be
substantially more uncertain than a BMD-based POD, although one source of uncertainty
influencing confidence is the observation of responses only in the high dose group), and dosimetric
adjustments using PFBA-specific toxicokinetic information (see Table 5-8).
2 See Table 5-7 for details on how to calculate candidate values for salts ofPFBA; briefly, the candidate values
for different salts of PFBA would be calculated by multiplying the candidate value for the free acid of PFBA by
the ratio of molecular weights. For example, for the ammonium salt the ratio would be: MW ammonmm salt —
MW free acid
231
— = 1.079. This same method of conversion can be applied to other salts of PFBA, such as the potassium or
sodium salts, using the corresponding molecular weights.
3 Note that the RfD for the free acid presented in this document and an RfD for the anion of PFBA
(perfluorobutanoate, C3F7CO2", CASRN 45048-62-2) would be practically identical given the molecular
weights between the two compounds differ by less than 0.5%, (i.e., by the weight of a single hydrogen atom).
4HAWC is a modular content management system designed to store, display, and synthesize multiple data
sources for the purpose of producing human health assessments of chemicals. This online application
documents the overall workflow of developing an assessment from literature search and systematic review,
to data extraction (human epidemiology, animal bioassay, and in vitro assay), dose-response analysis, and
finally evidence synthesis and visualization. In order to view HAWC study evaluation results, visualizations,
etc., users must create first create a free account; see https: //hawcprd.epa.gov/about for more details.
This document is a draft for review purposes only and does not constitute Agency policy.
xiv DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Toxicological Review ofPFBA and Related Salts
Noncancer Effects Observed Following Inhalation Exposure
No studies are available that examine toxicity in humans or experimental animals following
inhalation exposure, and no physiologically based pharmacokinetic (PBPK) models exist to allow a
route-to-route extrapolation; therefore, no inhalation reference concentration (RfC) was derived.
Evidence for Carcinogenicity
Under EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA. 20051. EPA concluded
there is inadequate information to assess carcinogenic potential for PFBA by either oral or inhalation
routes of exposure. Therefore, the lack of data on the carcinogenicity ofPFBA precludes the
derivation of quantitative estimates for either oral (oral slope factor [OSF]) or inhalation
(inhalation unit risk [IUR]) exposure.
Subchronic Oral Reference Dose (RfD) for Noncancer Effects
In addition to providing organ/system-specific RfDs for lifetime exposures in multiple
systems (see Table 5-9), less-than-lifetime (subchronic) RfDs also were derived (see Table 5-10).
In the case ofPFBA, all studies used to calculate the subchronic values were subchronic or
gestational in duration. Therefore, the method to calculate the organ/system-specific subchronic
RfDs is identical to that used for calculating the organ/system-specific RfDs, except in the
application of the UFs (e.g., UFs = 1 rather than 10). Thus, the individual organs and systems for
which specific subchronic RfD values were derived were the liver, thyroid, and developing fetus.
The value for the developing fetus was selected for the subchronic RfD. A BMDL of 3.8 mg/kg-day
NH4+PFB for increased time to vaginal opening in neonatal female mice was used as the basis for the
POD. After converting the PODs from units of mg/kg-day NH4+PFB to units of mg/kg-day PFBA (by
multiplying by the ratio of the molecular weights of the free acid and the ammonium salt), the HED
was calculated by multiplying the POD for the free acid by the ratio of serum clearance values
between mice and humans. The subchronic RfD for PFBA was calculated by dividing the PODhed of
0.62 mg/kg-day PFBA by a composite uncertainty factor of 100 to account for extrapolation from
rats to humans (UFa = 3), for interindividual differences in human susceptibility (UFH = 10), and
deficiencies in the toxicity database (UFd = 3). The subchronic RfD derived from the effects on
delayed time to vaginal opening, as representative of general developmental delays, was
6 x 10"3 mg/kg-day5.
5 See Table 5-10 for details on how to calculate subchronic candidate values for salts of PFBA; briefly, the
candidate values for different salts of PFBA would be calculated by multiplying the candidate value for the
free acid of PFBA by the ratio of molecular weights. For example, for the ammonium salt the ratio would be:
mw ammonium salt _ 231 _ ^ Qyg same method of conversion can be applied to other salts of PFBA, such as
MW free acid 214
the potassium or sodium salts, using the corresponding molecular weights.
This document is a draft for review purposes only and does not constitute Agency policy.
xv DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Toxicological Review ofPFBA and Related Salts
1. OVERVIEW OF BACKGROUND INFORMATION
AND ASSESSMENT METHODS
A series of five PFAS assessments (PFBA, perfluorohexanoic acid [PFHxA], perfluorohexane
sulfonate [PFHxS], perfluorononanoic acid [PFNA], perfluorodecanoic acid [PFDA], and their
associated salts; see December 2018 IRIS Outlook) is being developed by the Integrated Risk
Information System (IRIS) Program at the request of the U.S. Environmental Protection Agency
(EPA) national programs and regions. Appendix A is the systematic review protocol for these five
PFAS assessments. The protocol outlines the scoping and problem formulation efforts relating to
these assessments, including a summary of other federal and state reference values for PFBA. The
protocol also lays out the systematic review and dose-response methods used to conduct this
review (see also Section 1.2). This systematic review protocol was released for public comment in
November 2019 and was subsequently updated on the basis of those public comments. Appendix A
includes the updated version of the protocol, including a summary of the updates in the protocol
history section (see Appendix A, Section 12).
1.1. BACKGROUND INFORMATION ON PERFLUOROBUTANOIC ACID
(PFBA)
Section 1.1 provides a brief overview of aspects of the physicochemical properties, human
exposure, and environmental fate characteristics of perfluorobutanoic acid (PFBA,
CASRN 375-22-4) and its related salt ammonium perfluorobutanoate (PFB, CASRN 10495-86-0)
that might provide useful context for this assessment. This overview is not intended to provide a
comprehensive description of the available information on these topics. The reader is encouraged
to refer to source materials cited below, more recent publications on these topics, and the
assessment systematic review protocol (see Appendix A).
1.1.1. Physical and Chemical Properties
PFBA and its related salts are members of the group of per- and polyfluoroalkyl substances
(PFAS). Concerns about PFBA and other PFAS stem from the resistance of these compounds to
hydrolysis, photolysis, and biodegradation, which leads to their persistence in the environment
fSundstrom etal.. 2012al. The specific chemical formula ofPFBA is C4HF7O2 and the chemical
formula of NH4+PFB is C4H4F7NO2. More specifically, these PFAS are classified as perfluoroalkyl
carboxylic acids [PFCAs; OECD f20181]. Because PFBA and NH4+PFB are PFCAs containing less than
seven perfluorinated carbon groups, they are considered short-chain PFAS (ATSDR. 2018a). The
This document is a draft for review purposes only and does not constitute Agency policy.
1-1 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
chemical structures ofPFBA and NFU+PFB are presented in Figure 1-1, and select physicochemical
properties are provided in Table 1-1.
Figure 1-1. Chemical structures of perfluorobutanoic acid (PFBA) and
ammonium perfluorobutanoate (NH4+PFB).
Table 1-1. Predicted or experimental physicochemical properties of
perfluorobutanoic acid (PFBA; CASRN 375-22-4) and ammonium
perfluorobutanoate (NH4+PFB; CASRN 10495-86-0)
Property (unit)
Value
PFBA (free acid)
NhVPFB
Molecular weight (g/mol)
214a
230. la
Melting point (°C)
-17.5a
ND
Boiling point (°C)
12 la
ND
Density (g/cm3)
1.65a
ND
Vapor pressure (mm Hg)
6.37a
ND
Henry's law constant (atm-m3/mole)
4.99 x 10"5a b
ND
Water solubility (mol/L)
2.09 x 10"3a
ND
PKa
0.08bc
ND
Octanol-water partition coefficient (Log Kow)
1.43a
ND
Soil adsorption coefficient (L/kg)
47.9a'b
ND
Bioconcentration factor (BCF)
7.61a
ND
ND = no data.
aU.S. EPA (2018a) Chemicals Dashboard (PFBA DTXSID: 4059916):
https://comptox.epa.gov/dashboard/dsstoxdb/results?utf8=%E2%9C%93&search=375-22-4. Median or average
experimental values used where available; otherwise median or average predicted values used depending on
which was available.
Predicted.
c ATS PR (2018a).
This document is a draft for review purposes only and does not constitute Agency policy.
1-2 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Toxicological Review ofPFBA and Related Salts
1.1.2. Sources, Production, and Use
PFAS are not naturally occurring in the environment fATSDR. 2018al. They are manmade
compounds that are or have been used widely over the past several decades in consumer products
and industrial applications because of their resistance to heat, oil, stains, grease, and water. PFBA is
a breakdown product of other PFAS used in stain-resistant fabrics, paper food packaging, and
carpets; it was also used for manufacturing photographic film (MDH. 2017b). Shorter-chain PFAS
like PFBA are also being used as substitutes for longer chain PFAS in consumer products (Liu etal..
20141. Kotthoff et al. (20151 analyzed a variety of consumer products for PFAS. PFBA was detected
in nano- and impregnation-sprays, outdoor textiles, carpets, gloves, paper-based food contact
materials, ski wax, and leather.
The U.S. Environmental Protection Agency (EPA) has been working with companies in the
fluorochemical industry since the early 2000s to phase out the production and use of PFAS [ATSDR
(2018a): https://www.epa.gov/assessing-and-managing-chemicals-under-tsca/risk-management-
and-polyfluoroalkyl-substances-PFAS], The production and use of these chemicals, however, have
resulted in their release to the environment through various waste streams (NLM. 2016. 2013).
Also, because products containing PFAS are still in use, they could continue to be a source of
environmental contamination due to disposal or breakdown in the environment fKim and Kannan.
20071.
No Chemical Data Reporting (CDR) on production volume for PFBA or its salts are available
in EPA's ChemView (U.S. EPA. 2019a). Also, because facilities manufacturing, processing, or
otherwise using PFAS are not required to report on releases to the environment, no quantitative
information on PFBA is available in EPA's Toxic Release Inventory [TRI: (U.S. EPA. 2019a)].6
Wang etal. f 20141 estimated global emission estimates ofPFBA from direct and indirect
(i.e., degradation of precursors) sources between 1951 and 2030 to be between 15 and 915 metric
tons. The lower estimate assumes that producers cease production and use of long-chain PFCAs
and their precursors in line with global transition trends. The higher estimate assumes the
emission scenario in 2015 remains constant until 2030.
1.1.3. Environmental Fate and Transport
PFAS are stable and persistent in the environment fATSDR. 2018al. and many are found
worldwide in the air, soil, groundwater, and surface water, and in the tissues of plants and animals
(https://www.epa.gov/assessing-and-managing-chemicals-under-tsca/risk-management-and-
polyfluoroalkyl-substances-PFAS).
PFAS released to air exist in the vapor phase in the atmosphere and resist photolysis, but
particle-bound concentrations also have been measured (NLM. 2017. 2016. 2013: Kim and Kannan.
6As part of the National Defense Authorization Act for Fiscal Year 2020 (Section 7321), 172 per- and
polyfluoroalkyl substances will be added to the TRI list; however, neither PFBA nor its ammonium salt is on
the January 1,2020 list of 172 PFAS subject to TRI reporting requirements in Reporting Year 2020.
This document is a draft for review purposes only and does not constitute Agency policy.
1-3 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
20071. Wet and dry deposition are potential removal processes for particle-bound PFAS in air
fATSDR. 2018b: Barton etal.. 2007: Prevedouros etal.. 2006: Hurley etal.. 20041.
PFBA would be expected to be mobile in soil based on its soil adsorption coefficient (see
Table 1-1). Zhao etal. f 20161 observed that shorter chain PFAS like PFBA were transported more
readily from the roots to the shoots of wheat plants than longer chain PFAS. Venkatesan and
Halden (2014) analyzed archived samples from outdoor mesocosms to investigate the fate over
3 years of PFAS in agricultural soil amended with biosolids. The mean half-life for PFBA in these
environmental samples was estimated to be 385 days.
The potential for PFAS to bioconcentrate in aquatic organisms depends on their
bioconcentration factors (see Table 1-1), with longer chain PFAS accumulating to a greater degree.
Thus, the potential for PFBA to bioaccumulate is low compared with other PFAS (bioconcentration
factor of 7.61 vs. 789 and 752 for perfluorodecanoic acid [PFDA] and perfluorononanoic acid
[PFNA], respectively). PFBA has been found to bioaccumulate in foods grown on PFAS-containing
soil. Blaine etal. (2013) conducted a series of greenhouse and field experiments to investigate the
potential for PFAS to be taken up by lettuce, tomatoes, and corn when grown in industrially
impacted biosolids-amended soil and municipal biosolids-amended soil. PFBA was found to
bioaccumulate more readily than other PFAS (e.g., PFOA, PFOS, PFHxA, PFHxS, PFDA, and PFNA)
with bioaccumulation factors of 28.4-56.8 for lettuce and 68.4 for corn. PFBA had a
bioaccumulation factor of 12.2-18.2 for tomatoes, which was higher than all other PFAS studied
except perfluoropentanoic acid (bioaccumulation factor of 14.9-17.1).
PFBA has not been evaluated under the National Air Toxics Assessment program
(https: //www.epa.gov/national-air-toxics-assessment). Likewise, although EPA conducted
monitoring for several PFAS in drinking water as part of the third Unregulated Contaminant
Monitoring Rule [UCMR; U.S. EPA f2019bl], PFBA was not among the 30 contaminants monitored.
PFBA can be detected in most dust samples obtained from U.S. homes and vehicles,
however, and has been measured at higher levels in the soil and sediment surrounding
perfluorochemical industrial facilities, at U.S. military facilities, and at training grounds where
aqueous film-forming foam (AFFF) has been used for fire suppression (see Appendix A, Section 2.1).
PFBA also has been measured in the surface water and groundwater at military installations, AFFF
training grounds, and industrial sites, although data are sparse. PFBA levels in water at these sites
seem to exceed those identified in drinking water (see Appendix A, Section 2.1).
PFBA also can be detected in food. PFBA has been found in fish at 16% of sites sampled in
the U.S. Great Lakes (Stahl etal.. 2014) and, although most of the available data are from samples
from outside the United States, PFBA has been detected in grocery items including dairy products,
meats and seafood, fruits and vegetables, food packaging, and spices (see Appendix A, Section 2.1).
Specifically regarding drinking water, PFBA concentrations ranged from 0.0855 to
2.04 |ig/L in seven municipal wells in Oakdale, Minnesota fU.S. EPA. 2019al. In New Jersey public
water systems, only 3% of raw water samples contained PFBA, and did so at concentrations much
This document is a draft for review purposes only and does not constitute Agency policy.
1-4 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Toxicological Review ofPFBA and Related Salts
less than those reported in Minnesota [range from nondetectable to 0.006 |ig/L; (Post etal.. 20131].
Heo etal. f20141 detected PFBA in tap water and bottled water in Korea at mean concentrations of
2.02 and 0.039 ng/L, respectively. The concentrations ofPFBA measured at National Priorities List
(NPL) sites are provided in Table 1-2 fATSDR. 20171.
Table 1-2. Perfluorobutanoic acid (PFBA) levels in water, soil, and air at
National Priority List (NPL) sites
Media
Value
Number of NPL sites with
detections
Water (ppb)
Median
Geometric mean
2.15
1.03
3
Soil (ppb)
Median
Geometric mean
1,600
1,600
2
Air (ppbv)
Median
Geometric mean
ND
ND
ND = No data.
Source: (ATSDR, 2017).
1.1.4. Potential for Human Exposure and Populations with Potentially Greater Exposure
The general population could be exposed to PFAS via inhalation of indoor or outdoor air
(with PFAS possibly being released to the atmosphere via manufacturing processes or via disposal,
i.e., incineration), ingestion of drinking water and food, and dermal contact with PFAS-containing
products (ATSDR. 2018a). Exposure might also occur via hand-to-mouth transfer of materials
containing these compounds fATSDR. 2018al. The oral route of exposure has been considered the
most important one among the general population, however fKlaunig etal.. 20151. Contaminated
drinking water is likely to be a significant source of exposure. Due to the high water solubility and
mobility of PFAS in groundwater (and lack of remediation technology at water treatment facilities),
populations consuming drinking water from any contaminated watershed could be exposed to
PFAS fSun etal.. 20161. Gebbinketal. f20151 modeled exposure to PFBA among the adult general
population using a number of exposure scenarios based on the 5th, median, and 95th percentiles of
all input exposure parameters. "Intermediate" exposure (i.e., based on median inputs for all
exposure parameters) from direct and indirect (i.e., precursor) sources was estimated to be
19 pg/kg-day. Of the pathways evaluated (i.e., ingestion of dust, food, water; inhalation of air),
direct intake of PFBA in water accounted for the largest portion (approximately 90-100%) of total
exposure for all three exposure scenarios considered.
Several PFAS have been monitored in the human population as part of the National Health
and Nutrition Examination Survey [NHANES; CDC (2019)]. but PFBA was not among those
This document is a draft for review purposes only and does not constitute Agency policy.
1-5 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Toxicological Review ofPFBA and Related Salts
measured. PFBA has also been detected in breastmilk and baby food products, indicating a
potential additional route of exposure for infants. Antignac et al. T20131 reports that PFBA was
detected in 17% (8 of 48) of breastmilk samples in a population of French mothers, with a mean
concentration of 0.081 ng/L. Lorenzo etal. T20161 further reported that PFBA was detected in
breastmilk, infant formulas, dry cereal baby food, and processed baby food in Valencia, Spain.
Although PFBA-specific exposure information is sparse, populations that might experience
exposures greater than those of the general population could include individuals in occupations
that require frequent contact with materials containing PFAS that break down into PFBA, such as
individuals working with stain-resistant fabrics, paper food packaging, ski wax, and carpets (see
Section 1.1.2). For example, Nilsson et al. f20101 observed a significant correlation between the
number of years individuals had worked as ski wax technicians and their blood levels ofPFBA.
Populations living near fluorochemical facilities where environmental contamination to PFAS that
can break down into PFBA has occurred might also be more highly exposed.
1.2. SUMMARY OF ASSESSMENT METHODS
Section 1.2 summarizes the methods used for developing this assessment A more detailed
description of the methods for each step of the assessment development process is provided in the
systematic review protocol (see Appendix A). The protocol includes additional problem
formulation details, including the specific aims and key science issues identified for this assessment
1.2.1. Literature Search and Screening
The detailed search approach, including the query strings and Populations, Exposures,
Comparators, and Outcomes (PECO) criteria (Table 1-3), are provided in Appendix A, Section 4 and
Appendix B, respectively. The results of the current literature search and screening efforts are
documented below. Briefly, a literature search was first conducted in 2017 and regular updates are
performed (the literature searches will continue to be updated until shortly before release of the
document for public comment). The literature search queries the following databases (no date or
language restrictions were applied):
• PubMed fNational Library of Medicine 1
• Web of Science fThomson Reuters!
• Toxline (National Library of Medicine)7
• TSCATS (Toxic Substances Control Act Test Submissions)
7 Toxline has recently been moved into PubMed as part of a broad National Library of Medicine
reorganization. Toxline searches can now be conducted within PubMed.
This document is a draft for review purposes only and does not constitute Agency policy.
1-6 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Toxicological Review ofPFBA and Related Salts
In addition, relevant literature not found through database searching was identified by:
• Review of studies cited in any PFBA PECO-relevant studies and published journal reviews;
finalized or draft U.S. state, U.S. federal, and international assessments (e.g., the draft
Agency for Toxic Substances and Disease Registry [ATSDR] assessment released publicly in
2018).
• Review of studies submitted to federal regulatory agencies and brought to the attention of
EPA. For example, studies submitted to EPA by the manufacturers in support of
requirements under the Toxic Substances Control Act (TSCA).
• Identification of studies during screening for other PFAS. For example, epidemiological
studies relevant to PFBA sometimes were identified by searches focused on one of the other
four PFAS currently being assessed by the Integrated Risk Information System (IRIS)
Program.
• Other gray literature (e.g., primary studies not indexed in typical databases, such as
technical reports from government agencies or scientific research groups; unpublished
laboratory studies conducted by industry; or working reports/white papers from research
groups or committees) brought to the attention of EPA.
All literature is tracked in the U.S. EPA Health and Environmental Research Online (HERO)
database fhttps://hero.epa.gov/hero/index.cfm/proiect/page/project id/26321. The PECO criteria
(Table 1-3) identify the evidence that addresses the specific aims of the assessment and to focus the
literature screening, including study inclusion/exclusion.
This document is a draft for review purposes only and does not constitute Agency policy.
1-7 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Table 1-3. Populations, Exposures, Comparators, and Outcomes (PECO)
criteria
PECO
element
Evidence
Populations
Human: Any population and lifestage (occupational or general population, including children and
other sensitive populations). The following study designs will be included: controlled exposure,
cohort, case control, and cross-sectional. (Note: Case reports and case series will be tracked as
potential supplemental material.)
Animal: Nonhuman mammalian animal species (whole organism) of any lifestage (including
preconception, in utero, lactation, peripubertal, and adult stages).
Other: In vitro, in silico, or nonmammalian models of genotoxicity. (Note: Other in vitro, in silico,
or nonmammalian models will be tracked as potential supplemental material.)
Exposures
Human: Studies providing quantitative estimates of PFBA exposure based on administered dose
or concentration, biomonitoring data (e.g., urine, blood, or other specimens), environmental or
occupational-setting measures (e.g., water levels or air concentrations, residential location or
duration, job title, or work title). (Note: Studies that provide qualitative, but not quantitative,
estimates of exposure will be tracked as supplemental material.)
Animal: Oral or inhalation studies including quantified exposure to PFBA based on administered
dose, dietary level, or concentration. (Note: Nonoral and noninhalation studies will be tracked
as potential supplemental material.) PFBA mixture studies are included if they employ an
experimental arm that involves exposure to a single PFBA. (Note: Other PFBA mixture studies
will be tracked as potential supplemental material.)
Studies must address exposure to the following: PFBA (CASRN 375-22-4), or the ammonium salt
NH4+PFB (CASRN 10495-86-0). [Note: Although PFBAs are not metabolized or transformed in
the body, precursor compounds known to be bio-transformed to a PFAS are of interest;
e.g., 6:2 fluorotelomer alcohol is metabolized to PFHxA and PFBA (Russell et al., 2015). Thus,
studies of precursor PFAS that identify and quantify PFBA will be tracked as potential
supplemental material (e.g., for ADME analyses or interpretations).]
Comparators
Human: A comparison or reference population exposed to lower levels (or no
exposure/exposure below detection levels) or for shorter periods of time.
Animal: Includes comparisons to historical controls or a concurrent control group that is
unexposed, exposed to vehicle-only or air-only exposures. (Note: Experiments including
exposure to PFBA across different durations or exposure levels without including one of these
control groups will be tracked as potential supplemental material [e.g., for evaluating key
science issues; Section 2.4 of the protocol].)
Outcomes
All cancer and noncancer health outcomes. (Note: Other than genotoxicity studies, studies
including only molecular endpoints [e.g., gene or protein changes; receptor binding or
activation] or other nonphenotypic endpoints addressing the potential biological or chemical
progression of events contributing toward toxic effects will be tracked as potential supplemental
material [e.g., for evaluating key science issues; Section 2.4 of the protocol].)
This document is a draft for review purposes only and does not constitute Agency policy.
1-8 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Toxicological Review ofPFBA and Related Salts
In addition to those studies meeting the PECO criteria and studies excluded as not relevant
to the assessment, studies containing supplemental material potentially relevant to the specific
aims of the assessment were inventoried during the literature screening process. Although these
studies did not meet PECO criteria, they were not excluded. Rather, they were considered for use in
addressing the identified key science issues (see Appendix A, Section 2.4) and other potential
scientific uncertainties identified during assessment development but unanticipated at the time of
protocol posting. Studies categorized as "potentially relevant supplemental material" included the
following:
• In vivo mechanistic or mode of action studies, including non-PECO routes of exposure
(e.g., intraperitoneal injection) and populations (e.g., nonmammalian models)
• In vitro and in silico models
• Absorption, distribution, metabolism, and excretion (ADME) and toxicokinetic studies
(excluding models)8
• Exposure assessment or characterization (no health outcome) studies
• Human case reports or case series studies
• Studies of other PFAS (e.g., perfluorooctanoic acid [PFOA] and perfluorooctane sulfonate
[PFOS])
The literature was screened by two independent reviewers with a process for conflict
resolution, first at the title and abstract level and subsequently the full-text level, using structured
forms in DistillerSR (Evidence Partners; https://distillercer.com/products/distillersr-systematic-
review-software/). Literature inventories for PECO-relevant studies and studies tagged as
"potentially relevant supplemental material" during screening were created to facilitate subsequent
review of individual studies or sets of studies by topic-specific experts.
1.2.2. Evaluation of Individual Studies
The detailed approaches used for the evaluation of epidemiological and animal toxicological
studies used in the PFBA assessment are provided in the systematic review protocol (see
Appendix A, Section 6). The general approach for evaluating PECO-relevant health effect studies is
the same for epidemiological and animal toxicological studies, although the specifics of applying the
approach differ; thus, they are described in detail in Appendices A, Sections 6.2 and 6.3,
8Given the known importance of ADME data, this supplemental tagging was used as the starting point for a
separate screening and review of toxicokinetics data (see Appendix A, Section 9.2 for details).
This document is a draft for review purposes only and does not constitute Agency policy.
1-9 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofPFBA and Related Salts
respectively. Approaches for evaluating mechanistic evidence are described in detail in Appendix
A, Section 6.5.
The key concerns for the review of epidemiological and animal toxicological studies are
potential bias (systematic errors or deviations from the truth related to internal validity that affect
the magnitude or direction of an effect in either direction) and insensitivity (factors that limit the
ability of a study to detect a true effect; low sensitivity is a bias toward the null when an effect
exists). In evaluating individual studies, two or more reviewers independently arrived at
judgments regarding the reliability of the study results (reflected as study confidence
determinations; see below) with regard to each outcome or outcome grouping of interest; thus,
different judgments were possible for different outcomes within the same study. The results of
these reviews were tracked within EPA's version of the Health Assessment Workplace
Collaboration (HAWC). To develop these judgments, each reviewer assigned a category of good,
adequate, deficient (or not reported, which generally carried the same functional interpretation as
deficient), or critically deficient (listed from best to worst methodological conduct; see Appendix A,
Section 6 for definitions) related to each evaluation domain representing the different
characteristics of the study methods that were evaluated on the basis of the criteria outlined in
HAWC.
Once all evaluation domains were evaluated, the identified strengths and limitations were
collectively considered by the reviewers to reach a final study confidence classification:
• High confidence: No notable deficiencies or concerns were identified; the potential for bias
is unlikely or minimal, and the study used sensitive methodology.
• Medium confidence: Possible deficiencies or concerns were noted, but the limitations are
unlikely to be of a notable degree or to have a notable impact on the results.
• Low confidence: Deficiencies or concerns were noted, and the potential for bias or
inadequate sensitivity could have a significant impact on the study results or their
interpretation. Low confidence results were given less weight than high or medium
confidence results during evidence synthesis and integration (see Sections 1.2.4 and 1.2.5).
• Uninformative: Serious flaw(s) were identified that make the study results unusable.
Uninformative studies were not considered further, except to highlight possible research
gaps.
Using the HAWC platform (and conflict resolution by an additional reviewer, as needed), the
reviewers reached a consensus judgment regarding each evaluation domain and overall
(confidence) determination. The specific limitations identified during study evaluation were
This document is a draft for review purposes only and does not constitute Agency policy.
1-10 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofPFBA and Related Salts
carried forward to inform the synthesis (see Section 1.2.4) within each body of evidence for a given
health effect (i.e., study confidence determinations were not used to inform judgments in isolation).
1.2.3. Data Extraction
The detailed data extraction approach is provided in Appendix A, Section 8. Briefly, data
extraction and content management were carried out using HAWC. Data extraction elements that
were collected from epidemiological, controlled human exposure, animal toxicological, and in vitro
studies are described in HAWC (https: //hawcprd.epa.gov/about/). Not all studies that meet the
PECO criteria went through data extraction: studies evaluated as being uninformative were not
considered further and therefore did not undergo data extraction, and outcomes determined to be
less relevant during PECO refinement did not go through data extraction. The same was true for
low confidence studies when medium and high confidence studies (e.g., on an outcome) were
available. All findings are considered for extraction, regardless of the statistical significance of their
findings. The level of extraction for specific outcomes within a study could differ (i.e., ranging from
a narrative to full extraction of dose-response effect size information). For quality control, data
extraction was performed by one member of the evaluation team and independently verified by at
least one other member. Discrepancies in data extraction were resolved by discussion or
consultation within the evaluation team.
1.2.4. Evidence Synthesis and Integration
For the purposes of this assessment, evidence synthesis and integration are considered
distinct but related processes (see Appendix A, Sections 9 and 10 for full details). For each assessed
health effect, the evidence syntheses provide a summary discussion of each body of evidence
considered in the review that directly informs the integration across evidence to draw an overall
judgment for each health effect The available human and animal evidence pertaining to the
potential health effects are synthesized separately, with each synthesis providing a summary
discussion of the available evidence that addresses considerations regarding causation that are
adapted from Hill (1965). Mechanistic evidence is also synthesized as necessary to help inform key
decisions regarding the human and animal evidence; processes for synthesizing mechanistic
information are covered in detail in Appendix A, Section 9.2.
The syntheses of the human and animal health effects evidence focus on describing aspects
of the evidence that best inform causal interpretations, including the exposure context examined in
the sets of studies. The evidence synthesis is based primarily on studies of high and medium
confidence. Low confidence studies could be used if few or no studies with higher confidence are
available to help evaluate consistency, or if the study designs of the low confidence studies address
notable uncertainties in the set of high or medium confidence studies on a given health effect. If low
confidence studies are used, a careful examination of the study evaluation and sensitivity with
potential effects on the evidence synthesis conclusions will be included in the narrative. When
possible, results across studies are compared using graphs and charts or other data visualization
This document is a draft for review purposes only and does not constitute Agency policy.
1-11 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofPFBA and Related Salts
strategies. The synthesis of mechanistic information informs the integration of health effects
evidence for both hazard identification (e.g., biological plausibility or coherence of the available
human or animal evidence; inferences regarding human relevance, or the identification of
susceptible populations and lifestages across the human and animal evidence) and dose-response
evaluation (e.g., selection of benchmark response levels, selection of uncertainty factors).
Evaluations of mechanistic information typically differ from evaluations of phenotypic evidence
(e.g., from routine toxicological studies). This is primarily because mechanistic data evaluations
consider the support for and involvement of specific events or sets of events within the context of a
broader research question (e.g., support for a hypothesized mode of action; consistency with
known biological processes), rather than evaluations of individual apical endpoints considered in
relative isolation.
Following the synthesis of human and animal health effects data, and mechanistic data,
integrated judgments are drawn across all lines of evidence for each assessed health effect. During
evidence integration, a structured and documented two-step process is used, as follows:
Building from the separate syntheses of the human and animal evidence, the strength of the
evidence from the available human and animal health effect studies are summarized in
parallel, but separately, using a structured evaluation of an adapted set of considerations
first introduced by Sir Bradford Hill (Hill. 1965). This process is similar to that used by the
Grading of Recommendations Assessment, Development, and Evaluation (GRADE) (Morgan
etal.. 2016: Guvatt etal.. 2011: Schiinemann et al.. 2011). which arrives at an overall
integration conclusion based on consideration of the body of evidence. These summaries
incorporate the relevant mechanistic evidence (or mode-of-action [MOA] understanding)
that informs the biological plausibility and coherence within the available human or animal
health effect studies. The terms associated with the different strength of evidence
judgments within evidence streams are robust, moderate, slight, indeterminate, and
compelling evidence of no effect.
The animal, human, and mechanistic evidence judgments are then combined to draw an
overall judgment that incorporates inferences across evidence streams. Specifically, the inferences
considered during this integration include the human relevance of the animal and mechanistic
evidence, coherence across the separate bodies of evidence, and other important information
(e.g., judgments regarding susceptibility). Note that without evidence to the contrary, the human
relevance of animal findings is assumed. The final output is a summary judgment of the evidence
base for each potential human health effect across evidence streams. The terms associated with
these summary judgments are evidence demonstrates, evidence indicates (likely), evidence suggests,
evidence inadequate, and strong evidence of no effect. The decision points within the structured
evidence integration process are summarized in an evidence profile table for each considered
health effect
This document is a draft for review purposes only and does not constitute Agency policy.
1-12 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofPFBA and Related Salts
As discussed in the protocol (Appendix A), the methods for evaluating the potential
carcinogenicity of PFAS follow processes laid out in the EPA cancer guidelines fU.S. EPA. 20051 and
that the judgements described here for different cancer types are used to inform the evidence
integration narrative for carcinogenicity and selection of one of EPA's standardized cancer
descriptions. These are: (1) carcinogenic to humans, (2) likely to be carcinogenic to humans, (3)
suggestive evidence of carcinogenic potential, (4) inadequate information to assess carcinogenic
potential, or (5) not likely to be carcinogenic to humans. However, for PFBA, data relevant to cancer
were sparse and did not allow for such an evaluation (see Section 3.3).
1.2.5. Dose-Response Analysis
The details for the dose-response employed in this assessment can be found in Appendix A,
Section 11. Briefly, a dose-response assessment was performed for noncancer health hazards,
following exposure to PFBA via the oral route, as supported by existing data. For oral noncancer
hazards, oral reference doses (RfDs) are derived when possible. An RfD is an estimate, with
uncertainty spanning perhaps an order of magnitude, of an exposure to the human population
(including susceptible subgroups) that is likely to be without an appreciable risk of deleterious
health effects over a lifetime (U.S. EPA. 2002). The derivation of a reference value like the RfD
depends on the nature of the health hazard conclusions drawn during evidence integration. For
noncancer outcomes, a dose-response assessment was conducted for evidence integration
conclusions of evidence demonstrates or evidence indicates (likely). In general, toxicity values are
not developed for noncancer hazards with evidence suggests conclusions (see Appendix A, Section
10.2 for exceptions).
Consistent with EPA practice, the PFBA assessment applied a two-step approach for
dose-response assessment that distinguishes analysis of the dose-response data in the range of
observation from any inferences about responses at lower environmentally relevant exposure
levels flJ.S. EPA. 2012. 20051:
• Within the observed dose range, the preferred approach was to use dose-response
modeling to incorporate as much of the data set as possible into the analysis. This modeling
to derive a point of departure (POD) ideally includes an exposure level near the lower end
of the range of observation, without significant extrapolation to lower exposure levels.
• As derivation of cancer risk estimates and reference values nearly always involves
extrapolation to exposures lower than the POD; the approaches to be applied in these
assessments are described in more detail in Appendix A, Section 11.2.
When sufficient and appropriate human and laboratory animal data are available for the
same outcome, human data are generally preferred for the dose-response assessment because use
This document is a draft for review purposes only and does not constitute Agency policy.
1-13 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Toxicological Review ofPFBA and Related Salts
of human data eliminates the need to perform interspecies extrapolations. For reference values,
this assessment will derive a candidate value from each suitable data set Evaluation of these
candidate values will yield a single organ/system-specific value for each organ/system under
consideration from which a single overall reference value will be selected to cover all health
outcomes across all organs/systems. Although this overall reference value represents the focus of
these dose-response assessments, the organ/system-specific values can be useful for subsequent
cumulative risk assessments that consider the combined effect of multiple PFAS (or other agents)
acting at a common organ/system. For noncancer toxicity values, uncertainties in these estimates
are characterized and discussed.
For dose-response purposes, EPA has developed a standard set of models
fhttp: //www.epa.gov/bmds] that can be applied to typical data sets, including those that are
nonlinear. In situations where alternative models with significant biological support are available
(e.g., toxicodynamic models), those models are included as alternatives in the assessment(s) along
with a discussion of the models' strengths and uncertainties. EPA has developed guidance on
modeling dose-response data, assessing model fit, selecting suitable models, and reporting
modeling results [see the EPA Benchmark Dose Technical Guidance fU.S. EPA. 20121], Additional
judgment or alternative analyses are used if the procedure fails to yield reliable results; for
example, if the fit is poor, modeling might be restricted to the lower doses, especially if competing
toxicity at higher doses occurs. For each modeled response, a POD from the observed data was
estimated to mark the beginning of extrapolation to lower doses. The POD is an estimated dose
(expressed in human-equivalent terms) near the lower end of the observed range without
significant extrapolation to lower doses. The POD is used as the starting point for subsequent
extrapolations and analyses. For noncancer effects, the POD is used in calculating the RfD.
This document is a draft for review purposes only and does not constitute Agency policy.
1-14 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
Toxicological Review ofPFBA and Related Salts
2. LITERATURE SEARCH AND STUDY EVALUATION
RESULTS
2.1. LITERATURE SEARCH AND SCREENING RESULTS
The database searches yielded 610 unique records, with 4 records identified from
additional sources, such as Toxic Substances Control Act (TSCA) submissions, posted National
Toxicology Program (NTP) study tables, and review of reference lists from other authoritative
sources (AT SDR. 2018b) (see Figure 2-1). Of the 610 identified, 552 were excluded during title and
abstract screening, and 58 were reviewed at the full-text level. Of the 58 screened at the full-text
level, 17 were considered to meet the Populations, Exposures, Comparators, and Outcomes (PECO)
eligibility criteria (see Table 8, Appendix A). The studies meeting PECO criteria at the full-text level
included six epidemiological studies, nine animal studies, and one in vivo genotoxicity study. No
high-throughput screening data on perfluorobutanoic acid (PFBA) are currently available from
ToxCastor Tox21.
This document is a draft for review purposes only and does not constitute Agency policy.
2-1 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
PubMed
(n = 461)
PFBA
Literature Searches (through 2020)
WOS
(n =456)
ToxLine
(n = 28)
TSCATS
(n = 0)
Other
ATSDR assessment (n = 1)
Submitted to EPA (n = 3)
TITLE AND ABSTRACT SCREENING
Figure 2-1. Literature search and screening flow diagram for
perfluorobutanoic acid (PFBA).
2.2. STUDY EVALUATION RESULTS
1 Human and animal studies have evaluated potential effects to the thyroid, reproductive
2 systems, developing fetus, liver, urinary, and other organ systems (e.g., hematological) following
3 exposure to PFBA. The evidence base for these outcomes is presented in Sections 3.2.1-3.2.5.
4 The database of all repeated-dose oral toxicity studies for PFBA and the related compound
5 ammonium perfluorobutanoate (NH4+PFB) that are potentially relevant for deriving oral reference
6 dose (RfD) values includes four short-term studies in rats and mice fPermadi et al.. 1993: Permadi
7 et al.. 1992: lust et al.. 1989: Ikeda etal.. 19851. two 28-day studies in rats and mice fButenhoff et
8 al.. 2012: Foreman et al.. 2009: van Otterdiik. 2007al. one subchronic-duration study in rats
This document is a draft for review purposes only and does not constitute Agency policy,
2-2 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Toxicological Review ofPFBA and Related Salts
(Bute nhoff etal.. 2012: van Otterdijk. 2007b). and one gestational exposure study in mice (Das et
al.. 20081. In addition, eight epidemiological studies were identified that report on the association
between PFBA and human health effects fNian etal.. 2019: Wang etal.. 2019: Song etal.. 2018: Bao
etal.. 2017: Li etal.. 2017a: Li etal.. 2017b: Kim etal.. 2016: Fu etal.. 20141. The available animal
studies were generally well conducted and rigorous (i.e., medium or high confidence; see
Figure 2-2); thus, specific study limitations identified during evaluation are primarily discussed for
studies interpreted as low confidence, or when a limitation affects a specific inference for drawing
conclusions (e.g., in relation to a specific assessed endpoint within the health effects synthesis
sections below). No animal studies were considered uninformative. Thus, all animal studies
meeting PECO criteria during literature screening are included in the evidence synthesis and
dose-response analysis.
The study evaluations of the available epidemiological studies are summarized in
Figure 2-3, and rationales for each domain and overall confidence rating are available in Health
Assessment Workspace Collaborative (HAWC; see link in Figure 2-3). Based on the study
evaluations, one human epidemiological study was considered uninformative due to critical
deficiencies in exposure measurement fKim etal.. 20161: this study is not discussed further in this
assessment except to point out in more detail its critical deficiencies in the relevant health effects
section.
This document is a draft for review purposes only and does not constitute Agency policy.
2-3 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
ft <*? *** >> *9 aft ^ v^-
Repoitnff quality -
Allocation
Gbscrvabonai bias/blinding
Contou.rHlm^j'vJoabSf control
Selective reporting and altnlion
Chemtcal sdmnnstrabeni and thar attention
Exposure faming frequency and dmralion
Endpoinl jeitsibvrty »r confidence (overall)
Adequate (metric) or Medium confidence (overs!)
Deficient (unelrwj or Low confidence (ovcra*)
B
Crrhcalty dtficsenil (fflelnc) Or Uranformalive (eroeratt)
[nr
Noi reported
Figure 2-2. Evaluation results for animal studies assessing effects of
perfluorobutanoic acid (PFBA) exposure (see interactive data graphic for
rating rationales).
The following health outcome categories were investigated by the studies listed in Figure 2-2: thyroid effects
(Butenhoff et al., 2012), liver effects (Butenhoff et al.. 2012; Foreman et al„ 2009; Das et al., 2008; Permadi et al.,
1993; Permadi et al.. 1992; Just et al., 1989; Ikeda et al., 1985), developmental effects (Das et al., 2008), and
reproductive effects (Butenhoff et al., 2012).
This document is a draft for review purposes only and does not constitute Agency policy,
2-4 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
^' ^ X>^ 9^'
1 , 1 . 1 . '
Legend
Good (metric) or Hign confidence (over*#)
Adequate (me1ric|i or Medium confidence (overall i
Dericienl (metrtc) or Low confidence (overall)
Crittcaly deficient (metric) or Unmlormatrve (overall)
Not reported
Figure 2-3. Evaluation results for epidemiological studies assessing effects of
perfluorobutanoic acid (PFBA; interactive data graphic for rating rationales).
The following health outcome categories were investigated by the studies listed in Figure 2-3: thyroid effects (Li et
al„ 2017b; Kim et al.. 2016), liver effects (Nian et al.. 2019), developmental effects (Li et al., 2017a), reproductive
effects (Song et al., 2018), blood lipids (Fu et al., 2014), hypertension/blood pressure (Bao et al., 2017), and renal
function (Wang et al., 2019).
This document is a draft for review purposes only and does not constitute Agency policy,
2-5 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Toxicological Review ofPFBA and Related Salts
3. TOXICOKINETICS, EVIDENCE SYNTHESIS, AND
EVIDENCE INTEGRATION
3.1. TOXICOKINETICS
Toxicokinetic studies have been conducted with dosing solutions prepared from
Perfluorobutanoic acid (PFBA) (e.g., Burkemper et al. (201711 and the ammonium and potassium
salts (e.g., Chang etal. f200811. Some of the results evaluated below are semi-qualitative (e.g., that
distribution is to all tissues of the body), hence are described with reference to the acidic form,
since the salts immediately dissociate after dissolution and analytic measurements are of the acid.
These results are applicable independent of the form used to prepare dosing solutions.
The one study for which quantitation of the pharmacokinetic parameters might depend on
the form is Chang etal. f20081. Chang etal. f20081 were careful to identify the form used for each
part of their study so it is also clear that the chemical analysis used to measure concentrations in
serum used to determine pharmacokinetic parameters is that of the acid, PFBA. However,
calculation of the volume of distribution and clearance also involves the administered dose and
Chang etal. (2008) does not specify whether or not the doses were converted to dose of the acid
form. In a subsequent paper by the same research group evaluating the toxicokinetics of PFHxS,
Sundstrom etal. (2012b) explicitly state, "concentrations in serum, liver, urine and feces are
reported as PFHxS anion, and percent recoveries of administered dose in those matrices are
corrected for the potassium salt" Hence, we will presume that Chang et al. f20081 similarly
corrected either the applied dose or the serum concentrations to consistent units before reporting
their toxicokinetic parameters. Since the key parameters, volume of distribution and clearance,
effectively involve the ratio of dose to serum concentration (or the area-under-the-concentration
curve), resulting in measures of volume per kg BW or volume per time that are independent of the
molecular weight, these results can be applied to analysis ofPFBA per se, i.e., the acid or anion.
Conversion to corresponding doses of a given salt is applied before or after toxicokinetic analysis
then provides the appropriate human equivalent doses for each form.
Animal evidence has shown that perfluorobutanoic acid (PFBA), like other perfluorinated
chemicals, is well absorbed following oral administration and distributes to all tissues of the body
(Burkemper etal.. 2017). A study evaluating the volume of distribution concluded, however, that
distribution is predominantly extracellular (Chang etal.. 2008). Because of its chemical resistance
to metabolic degradation, PFBA appears to be primarily eliminated unchanged in urine and feces.
Toxicokinetic studies ofPFBA in rats, mice, and monkeys have been performed, providing
information on the absorption, distribution, metabolism, and excretion (ADME) ofPFBA
fBurkemper etal.. 2017: Chang etal.. 20081. Also, Russell etal. f20151 evaluated the metabolism of
This document is a draft for review purposes only and does not constitute Agency policy.
3-1 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofPFBA and Related Salts
6:2 fluorotelomer alcohol (6:2 FTOH) in mouse, rat, and human hepatocytes, showing that PFBA is a
metabolite of 6:2 FTOH, and evaluated PFBA toxicokinetics (TK) after inhalation and oral exposure
of rats to 6:2 FTOH. The distribution ofPFBA in human tissues also has been investigated fPerez et
al.. 20131. Information on the absorption and distribution of PFBA to the serum and liver
specifically has been investigated in several toxicological studies fGomis etal.. 2018: Butenhoff et
al.. 2012: Foreman et al.. 2009: Das etal.. 2008).
3.1.1. Absorption
Chang etal. f20081 conducted a set of toxicokinetic experiments in which Sprague-Dawley
(S-D) rats (3 male and 3 female) were given either a single intravenous (i.v.) or oral dose (30 mg/kg
body weight via gavage) of ammonium perfluorobutanoate (NH4+PFB). The serum area-under-the-
concentration-curve (AUC) was 1,090 ± 78 and 239 ± 5 ((ig-h/mL) in male and female rats,
respectively, after i.v. dosing and 1,911 ±114 and 443 ± 42 in males and females, respectively, after
oral dosing. That the AUC after oral dosing was almost two times higher than after i.v. dosing is
theoretically impossible but might be a statistical result from the small sample size (n = 3/group) or
due to a problem in dosing. The result, however, indicates 100% oral absorption.
In other experiments, Chang etal. (2008) orally administered 3-300 mg/kg to male and
female S-D rats via gavage. As expected, the concentration ofPFBA in the serum increased with
dose in a fairly linear fashion up to 100 mg/kg PFBA; however, the serum concentration ofPFBA in
rats dosed orally to 300 mg was approximately 60% the concentration at 100 mg/kg. Maximum
concentration (Cmax) values were similar in males and females following oral exposures to
30 mg/kg PFBA (131 ± 5 and 136 ± 12 [ig/mL, respectively), butthe time to peak concentration
(Tmax) differed between sexes: 1.25 ± 0.12 hours for males and 0.63 ± 0.23 hours for females. Both
values, however, indicate that absorption to the serum was fairly rapid in rats.
Cmax values for male and female mice exposed to PFBA via oral gavage also were similar at
lower doses (10 mg/kg; 50.50 ± 5.81 and 52.86 ± 2.08 [ig/mL), but differed at 30 mg/kg
(119.46 ± 13.86 and 151.20 ± 6.92 ^g/mL) and 100 mg/kg (278.08 ± 20.38 and
187.97 ±15.90 [ig/mL). Cmax and Tmax values for rats and mice at 30 mg/kg appear similar;
however, the Tmax was higher in female mice than in male mice (the opposite relationship compared
to rats).
3.1.2. Distribution
Burke mper etal. (2017) investigated the distribution ofPFBA in male CD-I mice (n = 4)
given a single i.v. dose of radiolabeled [18F]-PFBA (~0.074 MBq/|iL). At 4 hours postinjection, the
[18F]-PFBA was detected in every tissue investigated, with most of the dose found in the stomach
(~7.5% injected dose/g). All concentrations in the blood, lung, liver, kidney, intestines, and skin
were similar (~2-3%). Compared with perfluorooctanoic acid (PFOA) and perfluorohexanoic acid
(PFHxA), the concentration of PFBA was much lower in the liver (~27 and ~20%, respectively).
Chang etal. f20081 estimated volumes of distribution (Vd, mL/kg) for NH4+PFB in male and female
This document is a draft for review purposes only and does not constitute Agency policy.
3-2 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Toxicological Review ofPFBA and Related Salts
rats (209 ± 10 and 173 ± 21 at 30 mg/kg orally), mice (152 and 107 at 10 mg/kg orally; 296 and
134 at 30 mg/kg orally), and cynomolgus monkeys (526 ± 68 and 443 ± 59 at 10 mg/kg i.v.) (N = 3
animals/sex/dose group for all species); these values indicate thatNH4+PFB is primarily distributed
in the extracellular space.
Distribution in rats and mice was also examined in multiple toxicological studies ofPFBA
(see Table 3-1). Although limited in scope (i.e., PFBA was measured only in the liver and blood
serum), these studies demonstrated consistently that PFBA does distribute to the liver
compartment in both species. Butenhoff et al. (2012) observed that liver concentrations ofPFBA
([ig/g) were higher in male and female S-D rats exposed to PFBA for 28 days vs. rats exposed for
90 days. The ratio between liver concentrations (|J.g/g) and serum concentrations (|ig/mL) ranged
from 26% to 47% in the 28-day rats and 16% to 31% in the 90-day rats. In both exposure groups,
the concentration ofPFBA in the serum or liver was drastically reduced following a 3-week
recovery period. Das etal. (2008) investigated the distribution of PFBA to the liver in both
pregnant and nonpregnant rats and in postnatal day (PND) 1 and PND 10 pups. Serum levels and
liver levels ofPFBA differed between pregnant and nonpregnant rats in the lowest two dose groups.
Serum concentrations were approximately twofold higher in pregnant mice compared to
nonpregnant mice in the 35 mg/kg-day and 175 mg/kg-day dose groups. This pattern also was
observed for liver concentrations where pregnant animals had approximately two to three times
the liver concentration ofPFBA compared to nonpregnant animals in the 35 mg/kg-day and
175 mg/kg-day dose groups. Differences between pregnant and nonpregnant mice in serum and
liver concentrations of PFBA were attenuated in high-dose (350 mg/kg) animals. As would be
expected, both the serum and liver concentrations in PND 1 pups were much greater than those in
PND 10 pups. Das etal. f20081 corroborated the observations by Butenhoff et al. T20121 and Chang
etal. f20081 that serum PFBA concentrations are higher than liver concentrations. The ratios of
liver to serum PFBA concentration observed in Chang etal. f20081 were 22%-27% in male rats,
20%-23% in male mice, and 15%-17% in female mice. These differences in liver/serum
concentrations also were observed in various genetic strains of mice exposed to 35-350 mg/kg
PFBA: 38%-73% in wild-type mice, 13%-35% in peroxisome proliferator-activated receptor alpha
(PPARa) null mice, and 20%-33% in humanized PPARa mice (Foreman et al.. 2009).
Table 3-1. Serum and liver concentrations of perfluorobutanoic acid (PFBA)
following subchronic or gestational exposure
Dose group
(mg/kg-d)
Serum (pg/mL)
Liver (pg/g)
Serum (pg/mL)
Liver (pg/g)
Pregnant dams (Das et al., 2008)
Nonpregnant female mice (Das et al., 2008)
0
0.002 ± 0.001
0.003 ± 0.002
0.006 ± 0.003
0.038 ±0.017
35
3.78 ± 1.01
1.41 ±0.42
1.96 ± 1.0
0.51 ±0.20
175
4.44 ± 0.65
1.60 ±0.25
2.41 ± 1.65
0.86 ±0.55
This document is a draft for review purposes only and does not constitute Agency policy.
3-3 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Toxicological Review ofPFBA and Related Salts
Dose group
(mg/kg-d)
Serum (ng/mL)
Liver (ng/g)
Serum (ng/mL)
Liver (ng/g)
Pregnant dams (Das et al., 2008)
Nonpregnant female mice (Das et al., 2008)
350
2.49 ±0.60
0.96 ±0.18
2.67 ± 1.2
0.89 ±0.38
PD1 male and female neonates (Das et al., 2008)
PD10 male and female neonates (Das et al., 2008)
0
Not detected
0.004 ± 0.001
0.002 ± 0.002
0.003 ± 0.001
35
0.56 ±0.15
0.22 ±0.05
0.11 ±0.03
0.04 ± 0.01
175
0.61 ±0.39
0.29 ±0.14
0.14 ±0.07
0.04 ± 0.02
350
0.37 ±0.14
0.24 ± 0.08
0.12 ±0.05
0.04 ± 0.02
28-d male rats (Butenhoff et al., 2012)
90-d male rats (Butenhoff et al., 2012)
0
0.04 ± 0.05
<0.05
<0.01
<0.05
1.2
-
-
6.10 ±5.22
1.34 ± 1.24
6
24.65 ± 17.63
7.49 ± 4.46
13.63 ±9.12
3.07 ± 2.03
30
38.04 ±23.15
17.42 ±8.15
52.22 ±24.89
16.09 ± 9.06
150
82.20 ±31.83
37.44 ± 18.12
-
-
28-d female rats (Butenhoff et al., 2012)
90-d female rats (Butenhoff et al., 2012)
0
0.01 ±0.01
0.05 ± 0.03
0.07 ± 0.06
<0.05
1.2
-
-
0.23 ±0.14
0.05 ± 0.02
6
0.34 ±0.13
0.16 ±0.04
0.92 ±0.52
0.15 ±0.08
30
1.72 ±0.88
0.434 ±0.174
5.15 ±3.29
0.91 ±0.55
150
10.30 ± 4.50
2.70 ± 1.47
-
-
Perez etal. T20131 investigated the distribution ofPFBA in multiple tissues in cadavers in
Tarragona County, Spain. PFBA was detected in liver, brain, lung, and kidney samples, but was
below the level of detection in bone. Lung and kidney samples by far had higher PFBA
concentrations (304 and 464 ng/g, respectively) than brain or liver samples (14 and 13 ng/g,
respectively). For both the lungs and kidneys, PFBA was detected in greater quantities than any of
the other 20 per- and polyfluoroalkyl substances (PFAS) compounds analyzed. The observation
that PFBA was observed in the greatest quantities in kidney samples could be related to kidney
reabsorption. Chang etal. f20081 observed that rats given 300 mg/kg PFBA orally excreted
substantially greater amounts ofPFBA in the urine than did rats given 100 mg/kg (90.16% ± 2.75%
vs. 50.99% ± 4.35%), and the authors suggested this as evidence of saturation of a renal tubular
reabsorption process.
Data are not available that can be used reliably to estimate the volume of distribution (Kd) in
humans, which effectively provides the total body burden based on observed blood or serum
concentrations. An estimation of human body distribution for other PFAS is provided by the PBPK
models for PFOA and PFOS of fLoccisano etal.. 20111. which assume identical tissue:blood partition
This document is a draft for review purposes only and does not constitute Agency policy.
3-4 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofPFBA and Related Salts
coefficients (PCs) in humans and monkeys, equal to the values measured using tissues from rats
(PFOA) and mice (PFOS). This assumption is common to many PBPK models, based on the
expectation that the biochemical properties of a given tissue, muscle for example, which determines
the relative affinity of a chemical for that tissue compared to blood, are similar across mammalian
species: mouse, rat, monkey, and human muscle are all similar in composition and the difference in
chemical distribution to muscle as a whole is determined by the difference in the volume of muscle
per kg BW between species.
PCs are the effective tissue specific V& values because they determine the ratio of the
amount in a tissue vs. blood concentration at equilibrium. Based on this PBPK model fLoccisano et
al.. 20111. the Vd for PFOA predicted in monkeys and humans is 0.210 and 0.195 L/kg, respectively,
and for PFOS is 0.333 and 0.322 L/kg, respectively. These predictions are obtained by summing the
tissue fractions (ratios of tissue volumes/BW) multiplied by the corresponding PCs. In comparison,
based on the Loccisano etal. (2012) model for adult rats, the corresponding Vd values in that
species, for PFOA and PFOS, are 0.290 and 0.398, respectively. The difference between these rat
values and the human and monkey values is primarily due to the difference in physiology,
specifically the proportion of BW that is liver, kidney, and other tissues. Because of the
physiological similarities between humans and monkeys (more similar tissue fractions), the
predicted Vd values are within 7% of each other, although the difference between human and rat Vd
values is predicted to be 49% for PFOA and 24% for PFOS. They are much more similar between
humans and monkeys than between humans and rats, but the difference between humans and rats
is still less than a factor of 1.5.
Based on this analysis for PFOA and PFOS, the most reasonable choice for estimation of Vd
for PFBA in humans is to assume that it is similar to the Vd estimated for PFBA in monkeys, rather
than values estimated for mice or rats.
3.1.3. Metabolism
PFBA has been shown to be a product of the metabolism of 6:2 FTOH in mice, rats, and
humans (Russell etal.. 2015: Ruan etal.. 2014). No evidence of biotransformation for PFBA,
however, was found. PFBA, a short-chain (C4) of perfluoroalkyl acids (PFAAs), is expected to be
metabolically inert because its chemical stability is the same as longer chain PFAA chemicals,
including perfluorohexane sulfonate (PFHxS, C6), perfluorooctane sulfonate (PFOS, C8), and PFOA,
C8.
3.1.4. Excretion
In an overview of the toxicology of perfluorinated compounds, Lau (2015) briefly
summarized the excretion half-lives of seven compounds, including PFBA. All supporting data for
that review pertinent to PFBA are included in this analysis.
Chang etal. f20081 investigated the excretion ofPFBA in S-D rats, CD-I mice, cynomolgus
monkeys, and workers occupationally exposed to PFBA or compounds metabolized to PFBA. For
This document is a draft for review purposes only and does not constitute Agency policy.
3-5 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
rats and monkeys, three animals per sex were used (rats: three animals each for i.v. and oral
dosing) at the single dose given to each. For mice, three animals per sex per time point were used at
each dose, or 15-18 animals/dose. OECD guidelines state that a minimum of four animals per sex
per dose should be used fOECD. 20101. Thus, the rat and monkey studies fall short of this standard.
For rats, however, the average clearance from the two routes of exposure is proposed to best
represent males and females of that species (details below), which is then based on data from six
animals per sex. For monkeys, the average volume of distribution for both males and females is
used as an estimate for that value in humans, again incorporating data from six animals. Therefore,
these data are presumed sufficient for the specific parameters being estimated. In S-D rats exposed
orally to 30 mg/kg PFBA, a marked difference was noted in the serum PFBA excretion constants (A)
between males and females, 0.075/h and 0.393/h, respectively, for oral exposure and 0.109/h and
0.673/h, respectively, for intravenous exposure (see Appendix C for a complete discussion on
whether the calculated elimination constants in various species are mono- or biphasic). The
difference in oral A resulted in half-lives (ti/2) of 9.22 and 1.76 hours, respectively, for males and
females.
Russell etal. f20151 attempted to evaluate the excretion ofPFBA, formed as a metabolite of
6:2 FTOH, after inhalation exposures in rats (strain not stated). In single-day studies, the animals
were exposed by inhalation for 6 hours and their blood levels monitored for 24 hours after start of
exposure. The decline in PFBA blood concentration was negligible, however, after 0.5 and 5 ppm
6:2 FTOH exposures in male rats and after 0.5 ppm exposure in female rats, precluding estimation
of half-life. An excretion half-life of 19 hours was estimated from the 5-ppm single-day data for
5 ppm in female rats. After a 23-day inhalation exposure to male rats, use of a TK model resulted in
estimation of a 27.7-hour half-life for that sex, which could explain the inability to estimate a half-
life from the single-day exposures. Both estimates depend on the estimated yield (percent of 6:2
FTOH metabolized to PFBA), however, which was 0.2% for male rats and 0.02% for female rats.
Given the low yields, small errors in the estimate of that parameter could result in significant errors
in the estimated half-life. Thus, the results of Chang etal. (2008) is used to represent excretion in
rats.
In male CD-I mice, the clearance was similar in mice exposed to 10 mg/kg
(0.35 ± 0.09 mL/h) and 30 mg/kg PFBA (0.37 ± 0.80 mL/h); however, clearance at 100 mg/kg was
much higher (0.98 ± 0.14 mL/h) f Chang etal.. 20081. Although the fit of the simple one-
compartment model used to describe the kinetic data appeared adequate for the two lower doses, it
underpredicted the data at 24 and 48 hours for the 100 mg/kg dose, indicating it was not sufficient
for this highest exposure. In female mice clearance showed a similar, but less strong pattern, with
values of 0.76 ± 0.03, 0.87 ± 0.04, and 1.67 ± 0.08 mL/h at 10, 30 and 100 mg/kg doses, respectively
(Chang etal.. 20081. Unlike the data for male mice, the female mouse data were fit well by the one-
compartment pharmacokinetic (PK) model. For female data, the possible dose-dependence can be
resolved by using the average clearance for the lower two doses, which are closer to the doses
This document is a draft for review purposes only and does not constitute Agency policy.
3-6 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
evaluated for point-of-departure (POD) determination. Because male mouse endpoints are not
considered for POD determination, an alternative PK analysis of these data is not supported.
Cynomolgus monkeys (N = 3/sex) displayed a clear biphasic excretion pattern, with a rapid
decline in the initial (a) phase and a slower decline in the second ((3) phase fChang etal.. 20081.
Notably, the (3 phase began at around 24 hours and was observed because samples also were taken
at 2, 4, 7, and 10 days, while in rodents, samples were reported only to 24 hours (rats and female
mice) or 48 hours (male mice). Whereas serum levels in female rats and mice dropped to less than
3% of peak concentration by 24 hours, indicating minimal longer-term elimination, the levels in
male mice and rats did not drop as quickly and are more suggestive of a (3 phase. Also noted is that
the mouse and rat PK plots in Chang etal. f20081 use a linear y-axis, while the monkey PK plots use
a log y-axis. That a (3 phase would have been clearly observed in male mice and rats is possible had
serum sampling been continued for a longer duration, and possibly in female mice and rats had the
data simply been plotted with a log y-axis. Serum excretion half-lives for the a and (3 phases in male
monkeys exposed to 10 mg/kg PFBA via i.v. injection were 1.61 ± 0.06 hours and
40.32 ± 2.36 hours, respectively; ti/2 values in female monkeys were 2.28 ± 0.14 hours and
41.04 ± 4.71 hours, respectively.
Excretion ofPFBA from the serum in humans also was investigated by Chang etal. f20081.
In the initial occupational study, baseline PFBA serum concentration was determined in male
workers (n = 3) exposed to either PFBA or related fluorinated compounds. Following voluntary
removal from the workplace, workers had blood samples taken over 8 days to estimate half-lives of
excretion. Given the small sample size of the initial occupational study, a second study was
conducted in which seven male and two female workers had blood samples taken immediately
before a vacation and upon returning to the production facility (minimum elapsed time was 7 d).
For the male workers in the initial study, ti/2 of excretion from the serum ranged from 28.6 to
109.7 hours (1.2 to 4.6 d). For the nine workers in the second study, the ti/2 ranged from 44 to
152 hours (1.9 to 6.3 d), with an average value of 72 hours (95% confidence interval
[CI]: 1.8-4.2 d). Because these workers had been exposed previously for a significant duration and
the PK study was conducted over periods ranging from 7 to 11 days, the observed elimination is
reasonably presumed to represent (3-phase elimination, rather than the initial distribution phase.
Although only two female subjects were included in the second study (and their final PFBA serum
concentrations fell below the limit of detection), their estimated ti/2 values (118 h and 56 h) fell
within the range of ti/2 values reported for males (44-152 h). Therefore, although sex differences
in serum excretion in rodent species appear strong, the data in cynomolgus monkeys and humans
do not indicate such a difference.
Using an assumed BW°75 scaling and standard species BWs of 0.25 kg in rats and 80 kg in
humans, the half-life in humans is predicted to be 4.2 times greater than in rats. Given half-lives of
9.22 and 1.76 hours, respectively, in male and female rats (oral dose values), one would then
predict half-lives of 37.8 hours in men and 7.2 hours in women. Although the value for men based
This document is a draft for review purposes only and does not constitute Agency policy.
3-7 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofPFBA and Related Salts
on the BW°75 scaling approach is within a factor of 2 of the value determined by Chang et al. (20081.
BW°75 scaling is not based on data for this class of chemicals (i.e., serum binding and clearance
mechanisms are known to occur for PFAS). For example, EPA's Recommended Use of Body Weight
3/4 as the Default Method in Derivation of the Oral Reference Dose fU.S. EPA. 20111 does not mention
serum binding; it does include references related to VOCs, drugs, and overall metabolism (with
metabolism a significant component in the clearance of many other toxic chemicals) but does it cite
papers evaluating the pharmacokinetics of PFAS. These results for PFBA indicate that BW°75
scaling would lead to a lower prediction of human health risk at a given exposure than dosimetric
scaling based on the empirical data. Further, although only two women participated in the Chang et
al. f20081 study, that the observed elimination for them was 8 and 16 times slower than predicted
by BW°75 is an unlikely occurrence—even given the small sample size—and using of BW0-75 scaling
(applied to the half-life in female rats) could underpredict the risk of exposure by an order of
magnitude. Therefore, use of BW0 75 as an alternative means of extrapolation is not considered
further here.
Excretion in the urine appears to be the major route by which PFBA is excreted from the
body. Female rodents (rats: 100.68%-l 12.37%; mice: 65.44%-67.98%) are observed to have
higher percentages of the dose excreted in urine at 24 hours compared to male rodents (rats:
50.99%-90.16%; mice: 34.58%-35.16%). This is consistent with evidence that organic anion
transporters (OAT) expressed in the kidneys of rodents reabsorb PFAS (Weaver etal.. 2010: Yang et
al.. 2009) and are more highly expressed in male rodents (Cerrutti etal.. 2002: Kato etal.. 2002)
(Liuboievic etal.. 2007: Liuboievic etal.. 2004: Buist etal.. 2002). Both Yang etal. (2009) and
Weaver etal. (2010). however, observe that PFBA is not an active substrate of organic anion
transporters OAT1, OAT2, or OATPlal. Therefore, although the observed sex difference in urinary
excretion ofPFBA is consistent with the literature for reabsorption of PFAS in general in the kidney
in male rodents, the mechanism for this reabsorption for PFBA specifically is not currently known.
Sex differences in urinary excretion rates are not observed in primates, with both female and male
cynomolgus monkeys having rates similar to those of male mice (36.2% and 41.69%, respectively)
Chang etal. (2008). The excretion of PFBA in feces in rats and mice was very low compared with
the excretion in urine, but higher in mice than in rats (4.10%-10.92% and 0.16%-2.99%,
respectively).
3.1.5. Summary
PFBA clearance (CL) data, which can be used to estimate the average blood concentration
for a given dose, are available for mice and rats. For mice, the average CL from PK experiments at
10 and 30 mg/kg is suggested for use in animal-human extrapolation. For rats, the average of
values estimated from i.v. and oral exposure to 30 mg/kg is suggested.
Direct comparison of animal and human data requires consideration of observed half-lives,
because such data are available in humans, but CL cannot be directly estimated in humans.
Collectively, although the PFBA excretion half-lives for male and female rats appear shorter than for
This document is a draft for review purposes only and does not constitute Agency policy.
3-8 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Toxicological Review ofPFBA and Related Salts
male and female mice, respectively, data suggest a strong sex-specific toxicokinetic difference for
both species (i.e., females appear to have a much faster excretion rate than males). Humans have a
longer serum excretion half-life (~d) than rodents (~h). Although data in male mice and rats might
indicate a longer (3 phase elimination, the lower dose data in male mice are reasonably fit using a
single half-life (one-compartment model) as are the i.v. and oral data at the single dose given to rats
(30 mg/kg); the female mouse and rat data are likewise fit well by a one-compartment model
(Chang etal.. 2008). Therefore, although a longer elimination phase might be evident if additional
data were available, the estimated total clearance is unlikely to differ substantially from the
estimates provided here. The a-phase half-lives in monkeys (1.6-2.3 h) are similar to the half-life
obtained for female mice (2.8-3.1 h) and female rats (1-1.8 h) but are substantially shorter than the
half-life observed in male mice (13-16 h at lower doses) and male rats (6-9 h). The (3-phase half-
life in monkeys (1.7 d) is considerably longer than any of these rodent values but is comparable to
the lower end of the range for human subjects (1.8-2 d), although roughly one-half the average
among humans (3 d). As noted above, these human half-lives are expected to represent (3-phase,
considering the period of observation vs. exposure.
Human CL can be estimated using the PK relationship, CL = Vd ¦ ln(2)/to s. Because
human data do provide a value of ti/2, only a value of Vd is needed to determine CL. As discussed
above, however, one can reasonably anticipate that Vd in humans is similar to that in other primates
based on the similarity in physiology and assumptions common to PBPK modeling. This similarity
is illustrated on the basis of PBPK models for PFOA and PFOS (Loccisano etal.. 2011). from which
Vd in humans is predicted to be within 7% of the value for monkeys for those two PFAS. Thus, this
choice seems appropriate for estimating human clearance ofPFBA.
Table 3-2 provides a summary of PFBA toxicokinetics.
This document is a draft for review purposes only and does not constitute Agency policy.
3-9 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Table 3-2. Summary of toxicokinetics of serum perfluorobutanoic acid (PFBA)
(mean ± standard error)
Species/
sex
Study design
Excretion
half-life (h)
AUC
(Hg-h/mL)
Clearance
(mL/h)
Clearance
(mL/kg-h)a
Volume of
distribution
(mL/kg)
Rats
Male
30 mg/kg i.v. dose
6.38 ±0.53
1,090 ± 78
7.98 ±0.57
27.52
253 ±6
30 mg/kg oral dose
9.22 ±0.75
1,911 ± 114
4.63 ±0.28
15.70
209 ± 10
Female
30 mg/kg i.v. dose
1.03 ± 0.03
239 ±5
27.65 ±0.55
125.52
187 ±3
30 mg/kg oral dose
1.76 ±0.26
443 ± 42
14.32 ± 1.36
67.72
173 ±21
Mice
Male
10 mg/kg oral dose
13.34 ±4.55
1,026 ± 248
0.35 ±0.09
9.75
152
30 mg/kg oral dose
16.25 ±7.19
2,869 ± 6,116
0.37 ±0.80
10.46
296
100 mg/kg oral dose
5.22 ±2.27
3,630 ± 530
0.98 ±0.14
27.55
207
Female
10 mg/kg oral dose
2.87 ±0.30
387 ± 14
0.76 ±0.03
25.84
107
30 mg/kg oral dose
3.08 ±0.26
999 ±42
0.87 ± 0.04
30.03
134
100 mg/kg oral dose
2.79 ±0.30
1,760 ± 88
1.67 ±0.08
59.82
207
Monkeys
Male
10 mg/kg i.v. dose
1.61 ±0.06 (a)
40.32 ± 2.36 (P)
112 ±6
494 ±61
89.3
526 ±68
Female
10 mg/kg i.v. dose
2.28 ±0.14 (a)
41.04 ± 4.71 (P)
159 ±8
224 ± 19
62.9
443 ± 59
Humans
Males and
females
NV
Study 1:
28.6-109.71
Study 2: 72
(mean)
NV
NV
NV
NV
AUC = area-under-the-concentration-curve, NV = not available.
All data from Chang et al. (2008).
Calculated as dose (mg/kg) x (1000 ng/mg) / (AUC ng-h/mL).
1
2
3
4
9In this review, "statistical significance" indicates a p-value < 0.05, unless otherwise noted.
This document is a draft for review purposes only and does not constitute Agency policy.
3-10 DRAFT-DO NOT CITE OR QUOTE
3.2. NONCANCER EVIDENCE SYNTHESIS AND INTEGRATION
For each potential health effect discussed below, the synthesis describes the database of
available studies and the array of the experimental animal study results (the primary evidence
available for this PFAS) across studies. Effect levels presented in these arrays are based on
statistical significance9 or biological significance, or both. Examples relevant to interpretations of
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Toxicological Review ofPFBA and Related Salts
biological significance include directionality of effect (e.g., statistically significantly decreased
cholesterol/triglycerides is of unclear toxicological relevance) and tissue-specific considerations for
magnitude of effect (e.g., statistically nonsignificant increase of >10% in liver weight might be
considered biologically significant). A significant finding at a single, lower dose level but not at
multiple, higher dose levels might be interpreted as potentially spurious. For this section, evidence
to inform organ/system-specific effects ofPFBA in animals following developmental exposure is
discussed in the individual organ/system-specific sections (e.g., liver effects after developmental
exposure are discussed in the liver effects sections). Evidence of other effects informing potential
developmental effects (e.g., vaginal opening, eyes opening) is discussed in the "Developmental
Effects" section.
3.2.1. Thyroid Effects
Human Studies
Two studies reported on the association between PFBA exposure and thyroid hormones or
disease. One study on congenital hypothyroidism was considered uninformative10 due to concerns
with participant selection, confounding, and exposure measurement (Kim etal.. 2016). In one low
confidence study fLi etal.. 2017bl examining thyroid hormones among participants without
thyroid disease, inverse associations with thyroxine (T4), free triiodothyronine (T3), and
thyroid-stimulating hormone (TSH) were reported. Among the thyroid hormones measured, only
TSH demonstrated a statistically significant association (Pearson correlation coefficient = -0.348,
p < 0.01).
Animal Studies
Two high confidence studies reported in two unpublished reports and one publication from
the same research group evaluated the effects ofPFBA exposure on the thyroid, specifically
hormone levels, histopathology, and organ weight (Butenhoff etal.. 2012: van Otterdijk. 2007a. b)
following oral exposure (via gavage) of SD rats.11 Some outcome-specific considerations for study
evaluations were influential on the overall study rating for thyroid effects, but none of these
10Clicking on the hyperlinked study evaluation determination will take users to the HAWC visualization for
that study evaluation review. From there, users can click on individual domains to see the basis for that
decision. In the subsequent hazard sections, hyperlinked endpoint names will take users to the HAWC
visualization for that endpoint, from which users can click on the endpoint or studies to see the response data
from which the visualization is derived.
1 'The Butenhoff et al. (2012) study reported the findings of two unpublished industry reports: a 28-day and
90-day gavage study fully reported in van Otterdijk et al. (2007a, b). These industry reports were conducted
at the same facility and largely by the same staff but independently of one another and at different times: July
26, 2006 through September 15,2006 for the 28-day study and April 5, 2007 through August 6, 2007 for the
90-day study. Throughout the Toxicological Review, both Butenhoff et al. (2012) and the relevant industry
report are cited when discussing effects observed in these reports. Although only one study evaluation was
performed for this group of citations in HAWC, the overall confidence level of high applies to both the 28-day
and 90-day reports.
This document is a draft for review purposes oniy and does not constitute Agency poiicy.
3-11 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Toxicological Review ofPFBA and Related Salts
individual domain-specific limitations were judged likely to be severe or to have a notable impact
on the study results; all studies considered further in this section were rated as high or medium
confidence (see Figure 3-1). For more information on outcome-specific considerations for study
evaluations, please refer to the study evaluations in the HAWC PFBA database.
V2P"
Reporting quality
Allocation
Observational bias/blinding
Confounding/variable control -I
Selective reporting and attrition -I
Chemical administration and characterization -I
Exposure timing, frequency and duration -I
Endpoint sensitivity and specificity -
Results presentation -I
Overall confidence -I
! Legend
9 Good (metric) or High confidence (overall)
+ Adequate (metric) or Medium confidence (overall)
- Deficient (metric) or Low confidence (overall)
9 Critically deficient (metric) or Uninformative (overall)
Nr] Not reported
Figure 3-1. Evaluation results for animal studies assessing effects of
perfluorobutanoic acid (PFBA) exposure on the thyroid (see interactive data
graphic for rating rationales! .
Organ weight
Absolute and relative thyroid weights were statistically significantly (p < 0.01) increased
(—twofold) at the end of treatment in male rats exposed to 6 or 3 0 mg/kg-day via oral gavage for
28 days compared with controls. Organ weights, however, were increased only ~50% at
150 mg/kg-day, and this difference was not statistically significant (Butenhoffetal.. 2012: van
Otterdiik. 2007al. Thyroid weights were not significantly increased in male rats following the
recovery period or in female rats following the treatment or recovery period. Thyroid weight was
not measured in the rats exposed to NH4+PFB for 90 days fButenhoff etal.. 2012: van Otterdiik.
2007b).
Thyroid hormones
Male rats exposed to NH4+PFB for 28 days via gavage exhibited significantly decreased total
thyroxine (T4) and free T4 (fT4) levels compared with controls (see Table 3-3 and Figure 3-2).
Total T4 was reduced 59%, 66%, and 79% and free T4 was reduced 46%, 50%, and 66% at 6, 30,
and 150 mg/kg-day, respectively fButenhoff etal.. 2012: van Otterdiik. 2007al. Free T4
This document is a draft for review purposes only and does not constitute Agency policy.
3-12 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
1 concentrations had returned to control levels at all doses 21 days after exposure ended, but total T4
2 levels remained decreased in the 150 mg/kg-day group (-23%). TSH levels were not affected by
3 NH4+PFB at any exposure level. No treatment-related effects on any of the thyroid hormone
4 measures were observed in female rats exposed for 28 days fButenhoffetal.. 2012: van Otterdiik.
5 2007a).
Table 3-3. Percent change in thyroid hormones due to perfluorobutanoic acid
(PFBA) exposure in short-term and subchronic oral toxicity studies
Animal group
Dose (mg/kg-d)
1.2
6
30
150
Free T4
28 d; male S-D rats
(Butenhoff et al., 2012)
-46
-50
-66
28 d; female S-D rats
(Butenhoff et al., 2012)
-0.5
+18
-25
90 d; male S-D rats
(Butenhoff et al., 2012)
a
-9b
-30b
90 d; female S-D rats
(Butenhoff et al., 2012)
-6
+27
-15
Total T4
28 d; male S-D rats
(Butenhoff et al., 2012)
-59
-66
-79
28 d; female S-D rats
(Butenhoff et al., 2012)
-8
+27
-31
90 d; male S-D rats
(Butenhoff et al., 2012)
+13
-15
-39
90 d; female S-D rats
(Butenhoff et al., 2012)
+16
+14
-21
Bolded cells indicate statistically significant changes compared to controls (except for the 6 mg/kg-day and
30 mg/kg-day dose groups for free T4 in male rats exposed for 90 days, tests for statistical significance in those
cases were made to the 1.2 mg/kg-day group [see footnote b]); shaded cells represent doses not investigated in
the individual studies.
aNo sample for the control group was available due to insufficient sample volume for assay.
bComparison is made to the 1.2 mg/kg-day dose group.
6 Decreased total T4 and free T4 levels also were observed in male rats exposed to NH4+PFB
7 viagavage for 90 days fButenhoffetal.. 2012: van Otterdiik. 2007b). Total T4 increased 13% and
8 decreased 15% following 1.2 and 6 mg/kg-day, respectively. In male rats exposed to the highest
9 dose tested (30 mg/kg-day NH4+PFB), total T4 was significantly reduced by 39%. Free T4 was also
10 reduced in the 30-mg/kg-day dose group, but comparison to a control group was not possible due
11 to insufficient sample volume in the control group. The decrease in free T4, however, appeared to
This document is a draft for review purposes only and does not constitute Agency policy.
3-13 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Toxicological Review ofPFBA and Related Salts
be monotonic with increasing dose, and the decrease in the 30-mg/kg-day group (30%} was
statistically significant compared with the free T4 concentration in the 1.2 mg/kg-day group. No
statistically significant treatment-related effects were observed in female rats exposed to NHU+PFB
for 90 days, although total T4 was nonsignificantly decreased at the highest dose [30 mg/kg-day;
Butenhoff et al. (20121: van Otterdiik f2007bl].
Sludy Name Endpoint Name Study Type Animal Description Observation Time PFBA Thyroid Hormone Effects
Bulenhoff. 2012, 1289835 Total Thyroxine
-------
Toxicological Review ofPFBA and Related Salts
1 females and 110 effects were noted in the 6-mg/kg-day group (although the thyroid of only one
2 animal was available for testing in this group). No treatment-related effects were observed in the
3 recovery groups. In contrast to the histopathological examination, the histomorphometric analysis
4 reported no effects on thyroid cell height or colloidal area in either the treatment or recovery
5 groups. Follicular hypertrophy/hyperplasia also was observed to increase in male rats exposed to
6 30 mg/kg-day (9/10) for 90 days compared to controls when considering all lesions
7 (9/10 vs. 4/10; Cochran Armitage trend p = 0.0108) and lesions were graded "slight"
8 (5/10 vs. 0/10; Cochran Armitage trend p < 0.0001).
Table 3-4. Incidence and severity of thyroid follicular
hypertrophy/hyperplasia due to perfluorobutanoic acid (PFBA) exposure in
short-term and subchronic oral toxicity studies
Animal group (n = 10 in
all groups)
Dose (mg/kg-d)
0
1.2
6
30
150
28 d; male S-D rats
(Butenhoff et al., 2012)
3 (min)
3 (min)
9 (min)
7
(4 min, 3 mild)
90 d; male S-D rats
(Butenhoff et al., 2012)
4 (min)
6 (min)
4 (min)
9
(4 min, 5 mild)
Bolded cells indicate statistically significant changes compared with controls; shaded cells represent doses not
investigated in the individual studies. Severity normalized to four point scaled as follows: min = minimal severity;
mild = mild/slight severity; mod = moderate severity; sev = marked severity.
Study Name
Endpoint Name
Butenhoff, 2012, 1289835 Follicular Hypertrophy/Hyperplasia 28 Day Ora
90 Day Ora
Thyroid Follicular Colloidal Area 28 Day Ora
90 Day Ora
Thyroid Follicular Epithelial Cell Height 28 Day Ora
90 Day Ora
Study Type Animal Description
Rat, Sprague-Dawley (c
Rat. Sprague-Dawley (c
Rat. Sprague-Dawley (£
Rat, Sprague-Dawley {£
Rat, Sprague-Dawley (£
Rat. Sprague-Dawley (£
Observation Time
28.0 days
49.0 days
90 0 days
111.0 days
28.0 days
49.0 days
90.0 days
111.0 days
28.0 days
49.0 days
90.0 days
111.0 days
PFBA Other Thyroid Effects
-•
Thyroid Weight, Absolute
Thyroid Weight, Relative
28 Day Oral Rat. Sprague-Dawley (;
Rat. Sprague-Dawley (;
28 Day Oral Rat, Sprague-Dawley (;
Rat, Sprague-Dawley (5
28 0 days
49 0 days
28.0 days
49 0 days
28.0 days
49.0 days
28.0 days
49.0 days
0 No significant change
A Treatment-Related increase
^ Treatment-Related Decrease
-•
-•
-•
-•
-•
-•
-i—m n mi
mg/kg-day
Figure 3-3. Thyroid histopathology and organ-weight responses to
ammonium perfluorobutanoate (NH4+PFB) exposure (see interactive data
graphic and rationale for study evaluations for other thvroid effects in Health
Assessment Workspace Collaborative [HAWC]).
This document is a draft for review purposes only and does not constitute Agency policy,
3-15' DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofPFBA and Related Salts
Mechanistic Evidence and Supplemental Information
Thyroid effects observed in the PFBA database consist of increased thyroid weight,
increased incidence of follicular hypertrophy/hyperplasia, and decreased levels of thyroxine (total
and free T4). Overall, this pattern of decreased hormone levels with corresponding alterations in
tissue weight and histopathology in the absence of an increase in TSH is consistent with the human
clinical condition referred to as "hypothyroxinemia" (Alexander etal.. 2017: Choksi etal.. 2003).
The PFBA database is limited to two adult exposure studies (28- and 90-d) in rats (Bute nhoff etal..
2012: van Otterdiik. 2007a. b) but supplemental information from structurally related PFAS (PFBS
and PFHxA) is available. Decreases in thyroid hormones (total T3, total T4, and free T4) were
observed in PFBS-exposed pregnant mice and gestationally exposed female mouse offspring at
>200 mg/kg-d fFeng etal.. 20171 and in adult female and male rats following short-term exposures
of >62.6 mg/kg-d (NTP. 2019). Increased TSH was reported in mouse dams and in pubertal (PND
30) offspring following gestational exposure (Feng etal.. 2017). but no changes were noted in rats
exposed to PFBS as adults (NTP. 2019). a pattern consistent with the hypothyroxinemia observed
following adult PFBA exposure. Thyroid weight and histopathology were not changed after
short-term exposure to PFBS in adult male or female rats fNTP. 20191. Although the available
evidence for PFHxA provides weaker support for endocrine effects than studies on PFBA or PFBS,
the only study in the PFHxA database of animal toxicity studies to examine thyroid hormone levels
observed that short-term oral exposure to PFHxA altered thyroid hormone levels in male but not
female rats (NTP. 2018). Statistically significant, dose-dependent decreases in free and total T4
(25-73% and 20-58%, respectively) and to a lesser degree T3 (18-29%) were observed with no
concomitant increase in TSH (NTP. 2018).
Decreased serum T4 or T3 is a key event preceded by disrupted TH synthesis (via multiple
possible mechanisms, including thyroid stimulating hormone receptor [TSHR] binding and thyroid
peroxidase [TPO] or sodium-iodide symporter [NIS] inhibition) and results in a myriad of
downstream neurodevelopmental outcomes, including altered hippocampal anatomy/function and
hearing deficit. Thyroid hormones are critically important for proper brain development (Bernal.
2015: Miller etal.. 2009: Williams. 2008: Crofton. 2004: Morreale de Escobar etal.. 2004: Zoeller
and Rovet. 2004: Howdeshell. 2002) because they directly influence neurodevelopmental
processes, such as neurogenesis, synaptogenesis, and myelination fPuig-Domingo and Vila. 2013:
Stenzel and Huttner. 2013: Patel etal.. 20111. Early in gestation, TH is delivered to the developing
fetal brain via placental transfer from the mother to the fetus (Calvo etal.. 1990). The mother
imparts TH as its sole source until the fetal thyroid gland begins functioning. The fetal gland is
completely nonfunctional until late gestation (gestation day [GD] 17), having only minimal
functionality until near parturition (GD 22) (Bernal. 2015: Obregon etal.. 2007: Morreale de
Escobar et al.. 2004) at this point, in rats, approximately 17% of fetal T4 is still derived from the
maternal source despite the presence of a newly functioning thyroid gland fMorreale De Escobar et
This document is a draft for review purposes only and does not constitute Agency policy.
3-16 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofPFBA and Related Salts
al.. 1990). In humans, these maternal-derived fetal T4 estimates range from 30% to 50% (Obregon
etal.. 2007: Morreale de Escobar et al.. 2004: Vulsma etal.. 19891.
Cases of severe maternal and fetal hypothyroidism, which results from iodine deficiency,
Hashimoto's disease, or premature birth, further underscore the importance of maintaining thyroid
hormone homeostasis during pregnancy. Several human epidemiological studies have
demonstrated key relationships between decreased circulating levels of thyroid hormones, such as
T4 in pregnant women and in utero and early postnatal life neurodevelopmental status. For
example, neurodevelopmental and cognitive deficits have been observed in children who
experienced a 25% decrease in maternal T4 during the second trimester in utero fHaddowetal..
19991. Children born euthyroid but exposed to thyroid hormone insufficiency in utero (e.g., <10th
percentile free T4), present with cognitive impairments (e.g., decreased intelligence quotient [IQ],
increased risk of expressive language) or concomitant abnormalities in brain imaging fKorevaar et
al.. 2016: Henrichs etal.. 2010: Lavado-Autric etal.. 2003: Mirabella et al.. 2000). This level of T4
insufficiency (<10th percentile), defined as mild-to-moderate thyroid insufficiency, has been shown
to correspond to a 15%-30% decrease in T4 serum levels compared to median levels (Finken etal..
2013: Tulvez etal.. 2013: Roman etal.. 2013: Henrichs etal.. 20101. Animal toxicity studies also have
shown that decreases in mean maternal T4 levels of ~10%-17% during pregnancy and lactation
elicit neurodevelopmental toxicity in rat offspring f Gilbert etal.. 2016: Gilbert. 20111.
There are data gaps in the PFBA developmental toxicity database, including a lack of
information on the thyroid and nervous system following gestational exposure. Although short-
term PFBA exposure did not appear to alter thyroid hormone levels in nonpregnant adult female
rats, thyroid hormone levels fluctuate throughout normal gestation (O'Shaughnessv etal.. 2018:
Hassan etal.. 2017: Perez etal.. 2013: Calvo etal.. 1992: Calvo etal.. 1990: Fukuda et al.. 19801 as
maternal demands to provide the fetus with adequate thyroid hormones. Specifically, serum T4 and
T3 normally decline over the course of pregnancy and then rise during the postnatal period
(O'Shaughnessv etal.. 2018). Thus, although no changes in thyroid hormone levels occurred in
nonpregnant rats, that PFBA influences hormone homeostasis differently in pregnant rats during
the perinatal period is possible as maternal and fetal hormone demands fluctuate.
Overall, animal studies specific to PFBA and other potentially relevant PFAS provide
support for thyroid hormone disruptions by PFBA consistent with the human clinical condition of
hypothyroxinemia and for these alterations to potentially lead to other effects of concern (e.g.,
neurodevelopmental effects).
Evidence Integration Summary
Inverse associations between PFBA exposure and thyroid hormone levels were observed in
the one available informative human study (Li etal.. 2017b). Given the low confidence in the study
methods and the lack of biological coherence across the hormone changes, however, the available
human evidence did not notably contribute to the evidence integration judgment on PFBA-induced
thyroid effects (i.e., indeterminate evidence).
This document is a draft for review purposes only and does not constitute Agency policy.
3-17 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
The animal evidence comes from two high confidence experiments conducted by the same
laboratory fButenhoffetal.. 2012: van Otterdiik. 2007a. b), which reported PFBA-induced
perturbation of the thyroid in one species and sex (male S-D rats) across two different exposure
durations. The reported PFBA exposure-induced effects across thyroid hormone measures
(i.e., adult males, reductions in total or free T4; T3 was not measured) were consistent, dose
dependent, and associated with increasing absolute and relative thyroid weights and
histopathology (follicular hypertrophy/hyperplasia). These decreases were large in magnitude
(>50% in some PFBA exposure groups), and perturbations in total T4 were shown to persist at
least 21 days after the termination of 90-day exposure to the highest dose (150 mg/kg-day) but not
lower doses (in fact, total T4 was increased at 30 mg/kg-day). No effects (e.g., increases) on TSH in
exposed rats were observed. The observed pattern of effects on the thyroid (i.e., decreased total
and free T4 without a compensatory increase in TSH) after PFBA exposure is consistent with
thyroid perturbations following exposure to other PFAS, including the structurally related
compound perfluorobutane sulfonate (U.S. EPA. 2018b). Taken together, the consistent changes in
total and free T4, thyroid weights, and histopathology across the two available oral PFBA exposure
experiments are biologically coherent and plausible.
Several aspects of the animal evidence base decrease the strength or certainty of the
evidence. Although there is coherence across different measures of thyroid toxicity in male rats,
some effects across durations of exposure are inconsistent: some effects occur in the 28-day study
but not in the 90-day study, and the magnitude of change of some effects is larger in the short-term
than in the subchronic study. Also, in male rats, for free T4 only, the lack of a control group in
animals exposed for 98 days complicates the interpretation of that endpoint The overall pattern of
decreased thyroid hormones in the absence of a coordinated increase in TSH and commensurate
alterations in thyroid tissue weight and histopathology, however, is consistent with
hypothyroxinemia. Hypothyroxinemia has been defined in humans as a low percentile value of
serum free T4 (ranging from the 2.5th percentile to the 10th percentile of free T4), with a TSH level
within the normal reference range (Alexander etal.. 2017).
Although the organ-weight increases and histopathological effects (follicular hypertrophy)
observed in Butenhoff et al. (2012) are consistent with hypothyroxinemia, the mechanism by which
these changes occurred is unclear. Rodents are more sensitive to these histopathological changes
(follicular hypertrophy), which then can develop into follicular tumors fU.S. EPA. 19981. Increased
thyroid follicular hypertrophy supports the finding that the thyroid hormone economy is
perturbed. That the observed hypothyroxinemia was due to increased metabolism or competitive
displacement of T4 is likely (Butenhoff etal.. 2012). That no thyroid effects (e.g., hormone or
histopathological changes) were observed in females at any dose or treatment duration might be
related to PFBA toxicokinetics because clearance rates in rats are faster in females (compared to
males, see Section 3.1.4). Taken together, the available animal studies provided moderate evidence
for thyroid effects.
This document is a draft for review purposes only and does not constitute Agency policy.
3-18 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
Rodents and humans share many similarities in the production, regulation, and functioning
of thyroid hormones. Although differences exist, including the timing of in utero thyroid
development and hormone turnover rates, rodents are considered a good model for evaluating the
potential for thyroid effects in humans fZoeller et al.. 20071. More specifically, the observed
decreases in total or free T4 in the absence of increases in TSH are considered biologically relevant
to humans (Crofton. 2004: Lau etal.. 2003). TSH is an indicator the thyroid system has been
perturbed, but it does not always change when serum T4 is decreased (Hood etal.. 19991. Adverse
neurological outcomes have been demonstrated following hypothyroxinemia during the early
neonatal period with no changes in T3 or TSH fCrofton. 20041. The typical compensatory feedback
loop involves microsomal enzymes that induce uridine 5'-diphospho-glucuronosyltransferase
(UDP-GT), affecting the thyroid gland by increasing T4 glucuronidation, which in turn reduces
serum T4. In this case, the typical response to reduced serum free T4 is an increased production of
TSH (Hood and Klaassen. 2000). which can lead to thyroid hyperplasia or rat follicular tumors. In
that way, observation of thyroid histopathology can be an indication of perturbations in TSH levels
over time even in situations where increased TSH is not observed at the time histopathology is
measured fHood etal.. 19991. Rodents have been shown to have a unique sensitivity to thyroid
follicular hyperplasia (leading to development of follicular tumors), however, that is considered
less relevant to humans fU.S. EPA. 19981. Nevertheless, the coherent and consistent perturbations
to thyroid hormone economy and the resultant increased thyroid histopathology indicates that
PFBA is exerting some effect on the thyroid of exposed male rats. Even considering the increased
sensitivity of rodents to thyroid follicular hyperplasia compared to humans, thyroid hormone
perturbations are considered relevant to humans and might be even more sensitive to change in
humans compared to rodents fU.S. EPA. 19981.
A notable data gap, however, exists: Studies evaluating PFBA effects on neurodevelopment
or thyroid measures after developmental exposure (see Section 3.2.3 "Developmental Effects")
were not identified, thus leaving uncertainty on the potential for more sensitive developmental
effects ofPFBA exposure on the thyroid and nervous systems. During developmental lifestages,
such as gestational/fetal and postnatal/early newborn, thyroid hormones are critical in myriad
physiological processes associated with somatic growth and maturation and survival mechanisms,
such as thermogenesis, pulmonary gas exchange, and cardiac development fSferruzzi-Perri et al..
2013: Hillman etal.. 20121. That thyroid hormones are at sufficient levels is essential during times
critical to brain development and functioning and in the growth, development, and functioning of
numerous organ system processes, including basal metabolism and reproductive, hepatic, sensory
(auditory, visual) and immune systems (Forhead and Fowden. 2014: Gilbert and Zoeller. 2010:
Hulbert. 2000) (see Mechanistic Evidence and Supplemental Information subsection above).
Mammals are more susceptible during perinatal and postnatal lifestages because their
compensatory feedback responses are absent or not fully developed and they have low thyroid
hormone reserves fMorreale de Escobar et al.. 2004: Zoeller and Rovet. 20041. Further, thyroid
This document is a draft for review purposes only and does not constitute Agency policy.
3-19 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Toxicological Review ofPFBA and Related Salts
hormones are critically important in early neurodevelopment as they directly influence
neurogenesis, synaptogenesis, and myelination fPuig-Domingo and Vila. 2013: Stenzel and Huttner.
20131. Although the PFBA database lacks information on thyroid hormone levels in exposed
pregnant animals or offspring exposed during gestation, these effects have been observed following
exposure of mice to the structurally related PFAS, PFBS fU.S. EPA. 2018bl. Decreases in total T4
and T3 were observed in dams at GD 20 and offspring at PND 1, 30, and 60, clearly indicating that
thyroid hormone levels were perturbed during periods of neurological development Further,
given the evidence consistent with hypothyroxinemia, the PFBS assessment identifies
developmental neurotoxicity as a database limitation due to the known association between
thyroid hormone insufficiency during gestation and developmental neurotoxicity outcomes fU.S.
EPA. 2018b], Accordingly, given that developmental neurotoxicity (due to thyroid hormone
insufficiency) is a concern following exposure to PFBS, it follows that this concern is relevant to
exposure to PFBA during development because of the similarities in thyroid effects across the two
PFAS.
Taken together, the evidence indicates that PFBA exposure is likely to cause thyroid
toxicity in humans, given relevant exposure circumstances (see Table 3-5). This judgment is based
primarily on consistent and biologically coherent results from two high confidence studies (short-
term and subchronic study design) in male rats that indicate effects on thyroid hormone levels (T4
without compensatory effects on TSH). These effects on thyroid hormone levels generally occurred
at PFBA exposure levels >30 mg/kg-day, although some notable effects were observed after
exposure to 6 mg/kg-day.
This document is a draft for review purposes only and does not constitute Agency policy.
3-20 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Table 3-5. Evidence profile table for thyroid effects
Evidence Stream Summary and Interpretation
Inferences and Summary
Judgment
Evidence from studies of exposed humans (see Section 3.2.1: Human Studies)
Studies and
confidence
Summary of key
findings
Factors that increase
certainty
Factors that decrease
certainty
Judgments and
rationale
©©O
Evidence indicates (likely)
Thyroid Hormones
1 low confidence
study
• Single study
reporting inverse
associations with
free T4, free T3, and
TSH; only TSH was
statistically
significant
• No factors noted
• Lack of coherent
associations across
hormones
• Imprecision
ooo
Indeterminate
Primary basis:
Two high confidence studies in rats
ranging from short-term to
subchronic exposure; effects
observed at >6 mg/kg-d PFBA;
similar effects for related PFAS
Human relevance:
Effects in rats are considered
Evidence from in vivo animal studies (see Section 3.2.1: Animal Studies)
Studies and
confidence
Summary of key
findings
Factors that increase
certainty
Factors that decrease
certainty
Judgments and
rationale
potentially relevant to humans
based on conserved biological
processes, and the observed
Thyroid Hormones
2 hiqh confidence
studies in adult rats:
• 28-d
• 90-d
• Decrease in free
and total T4 in male
rats at >6 mg/kg-d
• Decrease in T4 with
no increase in TSH
is consistent with
hypothyroxinemia
• Consistent
increases in males
across all studies
• Dose-response
gradient
• Coherence of
decreased T4 with
histopathology
• Magnitude of
effect, up to 79%
• High confidence
studies
• Potential lack of
expected coherence
(no compensatory
TSH increase to T4
decrease)
®©o
Moderate
Findings considered
adverse based on
consistent and
biologically coherent
results for thyroid
hormone levels, organ
weights, and
pattern of changes is consistent
with hypothyroxinemia (see
Section 3.2.1: Mechanistic
Evidence and Supplemental
Information)
Cross-stream coherence:
N/A (human evidence
indeterminate)
Susceptible populations and
lifestages: The developing fetus
This document is a draft for review purposes only and does not constitute Agency policy.
3-21 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Evidence Stream Summary and Interpretation
Inferences and Summary
Judgment
Histopathology
2 hiqh confidence
studies in adult rats:
• 28-d
• 90-d
• Follicular
hypertrophy/hyper
plasia observed in
male rates at
30 mg/kg-d
• No histopathological
effects at
150 mg/kg-d (after
short-term
exposure)
• Consistent follicular
hypertrophy/hyper-
plasia in male rats
across studies
• Coherence of
hypertrophy with
T4 decreases
• High confidence
studies
• Potential lack of
expected coherence
(no change in TSH
levels)
• Unexplained lack of
significant effects at
highest tested dose
histopathology. The
observation of effects
only in males might be
explained by
toxicokinetics.
Uncertainties remain
as to how organ
weights and
histopathology are
affected in the
absence of TSH
increases.
and children are susceptible to
altered thyroid hormone status;
the lack of data on thyroid or
nervous system effects following
gestational exposure is a data gap.
Organ Weight
1 hiqh confidence
study in adult rats:
• 28-d
• Increase in thyroid
weight (absolute
and relative) at 6
and 30 mg/kg-d
• No change in
thyroid weight at
150 mg/kg-d
• Magnitude of
effect, >2-fold
increases
• High confidence
study
• Potential lack of
expected coherence
(no change in TSH
levels)
• Unexplained lack of
significant effects at
highest tested dose
Mechanistic evidence and supplemental information (see subsection above)
Summary of key findings, interpretation, and limitations
Evidence stream judgment
Key findings and interpretation:
• Pattern of effects consistent with human condition of hypothyroxenemia
• PFBA-induced thyroid changes similar to those for related PFAS (i.e., PFBS and,
although the evidence is weaker, PFHxA)
• Findings for PFBS indicate the potential for effects of concern during development
Limitations: No PFBA-specific mechanistic evidence informing thyroid effects
Findings for related PFAS
support the plausibility of
findings for PFBA, and the
potential for effects of concern
with PFBA exposure during
development
This document is a draft for review purposes only and does not constitute Agency policy.
3-22 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Toxicological Review ofPFBA and Related Salts
3.2.2. Hepatic Effects
Human Studies
One epidemiological study reported on the relationship between PFBA exposure and serum
biomarkers of liver injury. This study (Nian etal.. 2019) was cross-sectional and was classified as
medium confidence given minor concerns over participant selection, outcome ascertainment, and
confounding. Sensitivity was considered deficient due to low exposure levels and narrow contrast
for PFBA (detected in 52%, median [interquartile range (IQR)] = 0.03 ng/mL [0.01-1.6 ng/mL]),
which likely reduced the study's ability to detect an effect. The study found no association between
serum levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), total protein,
alkaline phosphatase (ALP), y-glutamyl transferase (GGT), total bilirubin, or cholinesterase with
PFBA exposure, but given the sensitivity concerns, this is difficult to interpret.
In addition, one low confidence cross-sectional study (Fu etal.. 2014) examined the
association between PFBA exposure and blood lipids. No association was reported; however, the
exposure levels in the study population were very low with narrow contrast (median [IQR] = 0.1
[0.03-0.2] ng/mL), so the study had poor sensitivity to detect an effect
Animal Studies
Hepatic effects were evaluated in multiple high and medium confidence, short-term and
subchronic studies in rats and mice (Bute nhoff etal.. 2012: Foreman et al.. 2009: van Otterdijk.
2007a. b; Permadi etal.. 1993: Permadi etal.. 1992) and in one high confidence developmental
toxicity study in mice (Das etal.. 2008). Some outcome-specific considerations for study
evaluations were influential on the overall study rating for liver effects, but none of these individual
domain-specific limitations were judged as likely to be severe or have a notable impact on the study
results, and all studies considered further in this section were rated as high or medium confidence
(see Figure 3-4). For more information on outcome-specific considerations for study evaluations,
please refer to the study evaluations in the HAWC PFBA database.
This document is a draft for review purposes only and does not constitute Agency policy.
3-23 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Toxicological Review ofPFBA and Related Salts
rt\\0*' qCP° -A\> ~
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Toxicological Review ofPFBA and Related Salts
Table 3-6. Percent increase in relative liver weight due to perfluorobutanoic
acid (PFBA) exposure in short-term and subchronic oral toxicity studies
Animal group
Dose (mg/kg-d)
1.2
6
30
35
150
175
350
28 d; male S-D rats
(Butenhoff et al., 2012; van Otterdiik, 2007a)
5
24
48
28 d; female S-D rats
(Butenhoff et al., 2012; van Otterdiik, 2007a)
-1
0
-3
90 d; male S-D rats
(Butenhoff et al., 2012; van Otterdiik, 2007b)
9
7
33
90 d; female S-D rats
(Butenhoff et al., 2012; van Otterdiik, 2007a)
0
-3
3
28 d; PPARa wild-type male SV/129 mice
(Foreman et al., 2009)
61
101
112
28 d; humanized PPARa male SV/129 mice
(Foreman et al., 2009)
38
63
81
28 d; PPARa null male SV/129 mice
(Foreman et al., 2009)
3
1
7
Pregnant P0 female CD-I mice on GD 18
(Das et al., 2008)
9
28
32
Nonpregnant P0 female CD-I mice on GD 18
(Das et al., 2008)
14
32
29
Fi male and female CD-I mice on PND 1
(Das et al., 2008)
9
30
41
Bolded cells indicate statistically significant changes compared with controls; shaded cells represent doses not
investigated in the individual studies.
The only null study (Ikeda etal.. 1985) reported that relative liver weight was not increased
over controls in male S-D rats exposed to 0.02% PFBA in the diet for 2 weeks (approximately
20 mg/kg-day). This study was judged low confidence, however, on the basis of concerns over
reporting, exposure characterization, and endpoint sensitivity/selectivity. Conversely, following
10 days of dietary exposure to 0.02% PFBA, relative liver weight was increased 38% in male
C57B1/6 mice in a medium confidence study fPermadi et al.. 19931. Twenty-eight days of daily
gavage exposure to >35 mg/kg-day PFBA significantly increased relative liver weights in adult male
wild-type (+/+) or humanized PPARa (hPPARa) Sv/129 male mice (Foreman et al.. 2009). The
relative liver weight of wild-type male mice was increased by 61%, 101%, and 112% at 35,175, and
350 mg/kg-day, respectively. Increased relative liver weight was also observed in these same dose
groups in humanized PPARa (hPPARa) male mice, although they were somewhat less than those
observed in wild-type mice: 38%, 63%, and 81%. Relative liver weight was not changed in PPARa
null (-/-) mice fForeman et al.. 20091. A similar profile of increased relative liver weight also was
This document is a draft for review purposes only and does not constitute Agency policy.
3-25 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
observed in male S-D rats exposed to >30 mg/kg-day NH4+PFB via oral gavage for 28 days
fButenhoffetal.. 2012: van Otterdiik. 2007al: Relative liver weights were increased 24% and 48%
at 30 and 150 mg/kg-day. Relative liver weights in both dose groups were observed to return to
control levels following a 21-day recovery period. Female rats exposed at the same dose levels
experienced no increases in relative liver weights (1-3% decrease).
Similar to increases following 28-day exposures, relative liver weights also were observed
to increase in male S-D rats exposed to NH4+PFB via oral gavage for 90 days fButenhoffetal.. 2012:
van Otterdiik. 2007b). with relative liver weights increased 33% at 30 mg/kg-day. As with the
short-term exposure, relative liver weights returned to control values following a 21-day recovery
period after termination of subchronic exposure. As observed in the short-term study, exposure to
NH4+PFB for 90 days did not increase liver weights in female rats (3% decreases to 3% increases).
In a developmental toxicity study in CD-I mice, exposure to NH4+PFB via oral gavage increased
relative (to body weight) liver weights in pregnant (measured on GD 18) and nonpregnant Po
females (Das etal.. 2008) at >175 mg/kg-day. Relative liver weights were increased by 28% and
32% at 175 and 350 mg/kg-day (respectively) in pregnant mice, whereas relative liver weights
were increased 32% and 29% in nonpregnant mice at the same dose levels. No effect on liver
weights was observed in the subset of dams followed until after weaning (PND 22). Similar
magnitudes of relative liver weight increases also were observed in Fi animals at PND 1: 30% and
41% at 175 and 350 mg/kg-day, respectively. In animals at PND 10, however, no change in relative
liver weights was observed. The lack of an effect on PND 10 in Fi or Po animals on PND 22 could be
because these animals were not exposed during lactation and therefore had a 10- or 22-day
recovery period compared with offspring or dams whose liver weights were measured on PND 1
and GD 17. This observation of no effect following a recovery period is consistent with the findings
of the subchronic and short-term exposures in adult animals fButenhoff et al.. 2 012: van Otterdiik.
2007a. hi.
In conclusion, effects on relative liver weights in adult male rats and mice were observed at
>30 or 35 mg/kg-day following subchronic or short-term exposures (respectively), whereas effects
in adult pregnant and nonpregnant female mice (exposed during pregnancy) and their offspring
were observed only at higher doses (>175 mg/kg-day). Adult female rats were only exposed up to
150 mg/kg-day in the subchronic study fButenhoffetal.. 2012: van Otterdiik. 2007bl. so whether
these animals would exhibit the same effects at the exposure levels used in the developmental
toxicity study fDas etal.. 20081 is unclear. Regardless, the data for relative liver weight seem to
indicate that male animals are more susceptible to this effect than female animals, possibly because
females have a much faster (5-6 times greater) excretion rate than males (see Section 3.1.4 for
details).
Changes in absolute liver weight across all studies were generally consistent with those
observed for relative liver weight. Following 10 days of dietary exposure to 0.02% (w/w) PFBA,
absolute liver weights were observed to be increased 64% in male C57B1/6 mice fPermadi etal..
This document is a draft for review purposes only and does not constitute Agency policy.
3-26 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Toxicological Review ofPFBA and Related Salts
1993: Permadi et ai. 19921. Absolute liver weights were also increased 27% and 45% following
28 days of exposure to 30 or 150 mg/kg-day NH-i+PFB, respectively fButenhoff etal.. 2012: van
Otterdiik. 2007a], No effects were observed in female rats following exposure or in male rats
following a 21-day recovery. Similar to increases following 28-day exposures, liver weights were
also observed to increase due to treatment in male S-D rats exposed to NH4+PFB for 90 days
(Butenhoffetal.. 2012: van Otterdijk. 2007b). with absolute liver weights increased by 23%. Liver
weights returned to control levels following a 21-day recovery period. As observed in the short-
term study, exposure to NH4+PFB for 90 days did not increase liver weights in female rats
(~3%-8% increases). In a developmental toxicity study in CD-I mice, exposure to NH4+PFB
increased absolute liver weights in pregnant and nonpregnant Po females fDas etal.. 20081 at
>175 mg/kg-day. Absolute liver weights were increased by 24% and 35% at 175 and
350 mg/kg-day, respectively, in pregnant mice, whereas absolute liver weights were increased 34%
and 21% at those same doses in nonpregnant Po females. Similar magnitudes of absolute liver
weights increases (27% and 32%) also were observed in Fi animals atPND 1 at 175 and
350 mg/kg-day fDas etal.. 2008). As with relative liver weights, no effect was observed in offspring
at PND 10 or in pregnant Po animals atpostweaning (PND 22),
Study Name
Endpoint Name
Study Type
Animal Description
Observation Time
PFBA Liver Weight Effects
Permadi 1993.1332452
Liver Wetghl Relative
10 Day Oral
Mouse C57BI/6 (£)
10.0 days
•-A
Ikeda 1985 2325571
Liver Wetghl Relative
14 Day Oral
Rat. Sprague-Dawley (£)
14 0 days
• •
Foreman 2009, 2325387
Relative Liver Weight
28 Day Oral
Mouse 129/SV(-')
280 days
•—A—
*
—A
Mouse 129/SV PPARo null (£)
280 days
•
—•
Mouse 129/Sv humanized PPARo {£)
28 0 days
• A
A
—A
Butenhoff, 2012. 1289835
Ltver Weight Relative
28 Day Oral
Rat Sprague-Dawley (~)
280 days
mA
~
490 days
—
•
Rat. Sprague-Dawley <2)
280 days
—
•
49 0 days
(••—
#
90 Day Oral
Rat, Sprague-Dawley (£)
90 0 days
«A
111.0 days
Rat. Sprague-Dawley (2)
Ill.Odays
<••
90 0 days
Das 2008. 1290825
Relative Liver Weight (Pregnant)
17 Day Oral
P0 Mouse CD-I (2)
18 0 days
•—•—
£
—A
Relative Liver Weight (Non-pregnant)
17 Day Oral
P0 Mouse CD-I (2)
18 0 days
A
—~
Relative Liver Weight (PND 1)
17 Day Oral
FI Mouse CD-I (=2)
20 0 days
*
—A
Relative Uver Weight (PND 10)
17 Day Oral
FI Mouse CD-1 (= = )
290 days
#
—•
Permadi 1993. 1332452
Liver Weight. Absolute
10 Day Oral
Mouse C57BM6 (£)
10 0 days
•-A
Butenhoff 2012. 1289835
Liver Weight Absolute
28 Day Oral
Rat, Sprague-Dawley (£)
28.0 days
« A—
~
49 0 days
ttt—
•
Rat Sprague-Dawley (2)
28 0 days
tt-t—
•
49 0 days
—
•
90 Day Oral
Rat. Sprague-Dawley (£)
90 0 days
» A
# Doses
111.0 days
A Treatment-Related Increase
Rat. Sprague-Dawley (2)
90 0 days
(• •
^ Trealment-Reiated Decrease
Ill.Odays
(• •
HH Dose Range
Das 2008. 1290825
Absolute Uver Weight (Pregnant)
17 Day Oral
P0 Mouse. CD-I (2)
180 days
•—
~
—A
Absolute Uver Weight (Non-pregnant)
17 Day Oral
P0 Mouse CD-I (2)
180 days
A
—A
Liver Weight Absolute (PND 1)
17 Day Oral
F1 Mouse CD-1 (£2)
20 0 days
~
—A
Liver Weight Absolute (PND 10)
17 Day Oral
FI Mouse CD-I (=2)
29 0 days
—•
—•
i i 1 i r i > i
-50 0 50 100 150 200 250 300 350 400
Axis label
Figure 3-5. Liver-weight response to ammonium perfluorobutanoate
(NH4+PFB) or perfluorobutanoic acid (PFBA) exposure (see interactive data
graphic and rationale for study evaluations for liver-weight effects in Health
Assessment Workspace Collaborative [HAWC]).
This document is a draft for review purposes only and does not constitute Agency policy,
3-27 ' DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofPFBA and Related Salts
Histopathologv
Histopathological examination of the livers of mice and rats across three separate gavage
studies of 28-day fButenhoff etal.. 2012: Foreman etal.. 2009: van Otterdiik. 2007al or 90-day
fButenhoffetal.. 2012: van Otterdiik. 2007bl exposure duration revealed significant,
dose-dependent alterations and lesions (see Table 3-7 and Figure 3-6).
Both wild-type and hPPARa mice exposed to PFBA for 28 days developed hepatocellular
hypertrophy at doses >35 mg/kg-day, whereas PPARa null mice did not develop hypertrophic
lesions at any dose following 28-day exposures (Foreman et al.. 20091. Although the incidence and
severity of the hypertrophic lesions were similar between wild-type and hPPARa mice at higher
doses, hPPARa mice developed more severe lesions at 35 mg/kg-day than did the wild-type mice
(5/10 severe lesions vs. 0/10, respectively). Hepatocellular hypertrophy also was observed in
6/10 S-D rats exposed to 150 mg/kg-day PFBA for 28 days (Bute nhoff etal.. 2012: van Otterdijk.
2007a) and 9/10 rats exposed to 30 mg/kg-day PFBA for 90 days fButenhoffetal.. 2012: van
Otterdiik. 2007b). In both cases, no lesions were observed in animals following a 21-day recovery
period.
hPPARa mice were much less susceptible to the development of hepatic focal necrosis
following a 28-day exposure to PFBA compared to wild-type mice. Wild-type mice developed
hepatic focal necrosis (with inflammatory cell infiltration) at 175 mg/kg-day (6/10) and
350 mg/kg-day (9/10), whereas focal necrosis was observed in only 1/10 and 2/10 hPPARa and
PPARa null mice at 175 and 350 mg/kg-day, respectively (Foreman etal.. 2009). For all strains,
most of the necrotic lesions were judged mild in severity. By comparison, in rats exposed to PFBA
for 28 days, no increase in hepatocellular coagulative necrosis fButenhoffetal.. 2012: van Otterdiik.
2007a) was observed. No effects on hepatocellular necrosis in rats were observed following 90-day
exposures to PFBA fButenhoffetal.. 2012: van Otterdiik. 2007bl.
Following exposure to 350 mg/kg-day for 28 days, centrilobular and periportal vacuolation
was observed in PPARa null and humanized mice, respectively, while no vacuolation was reported
for wild-type mice (Foreman etal.. 2009). No quantitative data were reported for these effects, so
examining the dose-response or magnitude of effect across doses was not possible. The lack of
vacuolation in wild-type animals is consistent with the lack of vacuolation in rats exposed to PFBA
for 90 days in fButenhoffetal.. 2012: van Otterdiik. 2007bl. where 4/10 control animals were
reported to exhibit vacuolation, but incidence dropped to 1/10 in the low-dose group and no
vacuolation was observed at higher doses. Although the number of studies was small, mice did
seem more sensitive to development of hepatocellular lesions compared to rats, possibly owing to
the observed differences in toxicokinetics between the two species: Mice are observed to have
serum excretion half-lives approximately two times longer than rats at similar exposure levels (see
Section 3.14 and Table 3-2 for details).
This document is a draft for review purposes only and does not constitute Agency policy.
3-28 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Table 3-7. Incidence and severity of liver histopathological lesions due to perfluorobutanoic acid (PFBA)
exposure in short-term and subchronic oral toxicity studies
Animal group (n = 10 in all groups)
Dose (mg/kg-d)
0
1.2
6
30
35
150
175
350
Hypertrophy
28 d; male rats
(Butenhoff et al., 2012; van Otterdiik, 2007a)
0
0
0
6 (min)
90 d; male rats
(Butenhoff et al., 2012; van Otterdiik, 2007b)
0
0
0
9 (5 min, 4
mild)
28 d; PPARa wild-type male mice
(Foreman et al., 2009)
0
10 (4 mild, 6
mod)
10 (1 mild, 1
mod, 8 sev)
10 (sev)
28 d; hPPARa male mice
(Foreman et al., 2009)
0
10 (1 mild, 4
mod, 5 sev)
10 (2 mod, 8
sev)
10 (sev)
28 d; PPARa null male mice
(Foreman et al., 2009)
0
0
0
0
Coagulative necrosis
90 d; male rats
(Butenhoff et al., 2012; van Otterdijk, 2007b)
0
0
0
0
Focal necrosis3
28 d; PPARa wild-type male mice
(Foreman et al., 2009)
0
1 (mild)
6 (2 min, 4
mild)
9 (8 mild, 1
mod)
28 d; hPPARa male mice
(Foreman et al., 2009)
0
1 (min)
1 (min)
2 (min)
28 d; PPARa null male mice
(Foreman et al., 2009)
0
0
1 (min)
2 (min)
Vacuolation
28 d; PPARa wild-type male mice
(Foreman et al., 2009)
None reported
28 d; hPPARa male mice
(Foreman et al., 2009)
Periportal vacuolation reported to increase at 350 mg/kg-d, compared to controls
28 d; PPARa null male mice
(Foreman et al., 2009)
Centrilobular vacuolation reported to increase at 350 mg/kg-d, compared to controls
Bolded cells indicate statistically significant changes compared to controls; shaded cells represent doses not investigated in the individual studies. Severity
normalized to four point scaled as follows: min = minimal severity; mild = mild/slight severity; mod = moderate severity; sev = marked severity,
incidence of focal necrosis for the positive control of Wy-14,643 (a known PPARa activator) was 3 total (1 minimal, 2 mild) at 50 mg/kg-day exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
3-29 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Toxicological Review ofPFBA and Related Salts
Study Name
Endpoint Name
Study Type
Animal Description
Observation Time
PFBA Liver Histopathology Effects
Foreman, 2009, 2325387
Hepatocellular Hypertrophy
28 Day Oral
Mouse. 129/SV (^)
28.0 days
•—A-
A
—A
Mouse. 129/SV PPARa null (^)
28.0 days
•—•—
4
—•
Mouse. 129/Sv humanized PPARa
28.0 days
•—A-
A
—A
Butenhoff, 2012, 1289835
Hepatocellular Hypertrophy
28 Day Oral
Rat, Sprague-Dawley (£)
28.0 days
€#-•—
# Doses
49.0 days
•
A Treatment-Related Increase
90 Day Oral
Rat, Sprague-Dawley (^)
90.0 days
<• A
Y Treatment-Related Decrease
111.0 days
I—I Dose Range
Foreman, 2009, 2325387
Hepatic Focal Necrosis
28 Day Oral
Mouse. 129/SV (;?)
28.0 days
4
A
Mouse. 129/SV PPARa null (^)
28.0 days
•
—•
Mouse. 129/Sv humanized PPARa (*',
28.0 days
0
—•
Butenhoff, 2012, 1289835
Hepatocellular Coagulative Necrosis
28 Day Oral
Rat. Sprague-Dawley (J)
28.0 days
•
49.0 days
<#-#—
•
0 0 50
100 150
200 250 300
350
4
0
mgrtcg-day
Figure 3-6. Liver histopathology response to ammonium perfluorobutanoate
(NH4+PFB) or perfluorobutanoic acid (PFBA) exposure (see interactive data
graphic and rationale for study evaluation for liver histopathology effects in
Health Assessment Workspace Collaborative [HAWC]).
Serum biomarkers
Serum biomarkers associated with altered liver function or injury including ALT, AST, ALP,
total protein, albumin, and total bilirubin were not significantly changed in male or female S-D rats
exposed to up to 150 mg/kg-day PFBA for 28 days (Butenhoffetal.. 2012: van Otterdijk. 2007a).
However, prothrombin time (a measure of clotting time induced by the liver-produced
prothrombin protein) was decreased at 150 mg/kg-day in males and at 6 and 30 mg/kg-day in
females (but not at 150 mg/kg-day), although decreases were small (~5-9% relative to control)
and were reported to be within the concurrent reference range for S-D rats. Prothrombin time,
however, was statistically significantly decreased (p < 0.01) in all dose groups in females after the
21-day recovery period. Some alterations in serum biomarkers were also observed in rats exposed
to PFBA for 90 days: ALP was increased 32% in male rats exposed to 30 mg/kg-day and bilirubin
was decreased 21% and 13% in male and female rats (respectively) exposed to 30 mg/kg-day
(Butenhoffetal.. 2012: van Otterdiik. 2007a). ALT was not affected by PFBA exposure in wild-type,
PPARa null, or hPPARa mice (Foreman et al.. 2009). Cholesterol levels were significantly (p < 0.01)
decreased 20% and 27% in male rats exposed to 30 and 150 mg/kg-day PFBA, respectively, for
28 days fButenhoffetal.. 2012: van Otterdiik. 2007al. Cholesterol levels returned to control values
following recovery, and no effects on cholesterol were observed in male rats exposed to PFBA for
90 days. No clear explanation exists to describe why cholesterol levels might be changed after 28,
but not 90, days ofPFBA exposure.
Mechanistic Evidence and Supplemental Information
The liver effects observed in the PFBA database consist of increased liver weight, increased
incidence of hepatocellular hypertrophy, and (to a lesser degree) hepatocellular necrosis.
Increased liver weight and hepatocellular hypertrophy can be associated with changes that are
adaptive in nature (Hall etal.. 2012). and not necessarily indicative of adverse effects unless
This document is a draft for review purposes only and does not constitute Agency policy.
3-30 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Toxicological Review ofPFBA and Related Salts
observed in concordance with other clinical, pathological markers of overt liver toxicity (see PFBA
Protocol; Appendix A). The IRIS PFAS Assessment Protocol (which addresses PFBA) states the
panel recommendations from Hall etal. f20121 can be used to judge whether observed hepatic
effects are adverse or adaptive in nature. Given that Hall etal. f20121 was focused on framing
noncancer liver effects in the context of progression to liver tumors, however, the protocol further
indicates that "...consultation of additional relevant information will be considered to interpret the
adversity of noncancer liver effects over a lifetime exposure, taking into account that effects
perceived as adaptive can progress into more severe responses and lead to cell injury." For PFBA,
the "additional relevant information" consists of multiple in vitro mechanistic studies, an in vivo
study investigating PFBA-induced liver effects in wild-type humanized PPARa mice, and PPARa-
null mice (Foreman), as well as evidence from other PFAS that help elucidate possible MOAs of
PFBA liver toxicity.
Many of the hepatic effects caused by exposure to perfluorinated compounds such as PFBA
have been attributed to activation of the peroxisome proliferator-activated receptor alpha
(PPARa12) (Rosenmai et al.. 2018: Biork and Wallace. 2009: Foreman etal.. 2009: Wolf etal.. 2008).
Due to reported cross-species differences in PPAR signaling potency and dynamics, the potential
human relevance of some hepatic effects has been questioned, particularly as it relates to
differences in PPARa activation and activity across species. The goal of the qualitative analysis
described in this section is to evaluate the available mechanistic evidence for PFBA-induced liver
effects and to assess the biological relevance of effects observed in animal models to possible effects
in humans.
Although the database is smaller for PFBA than for some other PFAS, in vitro studies
demonstrate that PFBA activates PPARa in both rodent and human cell lines. Studies using rodent
cell lines or COS-1 cells transfected to express rodent PPARa generally report that exposure to
PFBA consistently results in activation of PPARa and increased expression of PPARa-responsive
genes (Rosen etal.. 2013: Wolf etal.. 2012: Bjork and Wallace. 2009: Wolf etal.. 2008). Although
PFAS generally have been shown to activate PPARa, however, shorter chain PFAS such as PFBA
appear to be weak activators. For example, Biork and Wallace (2009) showed PFBA is a weaker
activator of PPARa in primary rat and human hepatocytes than is either the six-carbon PFHxA or
the eight-carbon PFOA. PFBA is also one of the weakest mouse and human PPARa activators
compared with other longer chain PFAS [i.e., C5-C12; Rosen etal. f20131: Wolf etal. f20121: Wolf et
al. f20081]. These studies also observed diminished effects and transcription levels in human cell
lines (primary hepatocytes) or COS-1 cells transfected with human PPARa compared to mice
(primary hepatocytes or transfected COS-1 cells). One study using the human hepatoma cell line
HepG2 also reported activation of PPARa after exposure to PFBA for 24 hours, further
12PPARa is a member of the nuclear receptor superfamily that can be activated endogenously by free fatty
acid derivatives. PPARa plays a role in lipid homeostasis but is also associated with cell proliferation,
oxidative stress, and inflammation fNIDWOI. 2017: Angrish et al.. 2016: Mellor etal.. 2016: Hall et al.. 2012).
This document is a draft for review purposes only and does not constitute Agency policy.
3-31 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
demonstrating that the human PPARa can be activated by PFBA (Rosenmai et al.. 20181.
Interestingly, when modeling the slope of PPARa activation in human hepatoma cells for various
PFAS, Rosenmai et al. f20181 observed PFBA (slope = 7.4 x 10~3) was a stronger activator than
PFOA (slope = 4.9 x 10-3). Foreman etal. f20091 investigated PPARa activation in the liver of mice
following in vivo exposure to PFBA. The PPARa-responsive gene CYP4A10 was activated to a
greater degree in wild-type mice than in humanized mice, but acyl-CoA oxidase (ACO, active in
(3-oxidation and lipid metabolism) appeared to be activated to a similar magnitude in both
wild-type and humanized mice. The known PPAR a/y activator Wy-14,643 activated CYP4A10 and
ACO to a similar magnitude in humanized PPARa mice compared to PFBA but to a lesser degree in
wild-type mice. Neither gene was activated following exposure to PFBA or Wy-14,643 in PPARa
null mice.
One in vivo study fForeman et al.. 20091 provided evidence that oral PFBA exposure elicits
apical, toxicological effects in humanized PPARa mice. This study showed that increased liver
weight and hepatocellular hypertrophy were induced following exposure to >35 mg/kg-day PFBA
in wild-type and hPPARa mice. Although magnitude of liver-weight increases was larger for
wild-type mice, the effect on hypertrophy was the same for wild-type and hPPARa mice at higher
exposures. Conversely, hPPARa mice had more severe lesions at lower doses compared with
wild-type mice. Increased liver weight and hypertrophy also occurred in positive controls treated
with Wy-14,643.
Liver enlargement is one of the most common observations associated with chemical
exposures via the oral route in laboratory animals and humans. In addition to measured increases
in the mass of liver tissue, histological evaluation typically reveals isolated or multifocal areas of
hepatocellular hypertrophy. The swelling of hepatocytes could include accumulation of lipid
material (e.g., micro- or macrovesicular steatosis), organellar growth and proliferation
(e.g., peroxisomes, endoplasmic reticulum), increased intracellular protein levels (e.g., Phase I and
II enzymes), and altered regulation of gene expression (e.g., stress response, nuclear receptors) (for
review see: Batt and Ferrari (1995)). Importantly, hepatocellular hypertrophy alone is
morphologically indistinguishable between an adaptive or toxic response in the absence of
additional indicators of cell status (Williams and Iatropoulos. 2002). such as reduced glutathione
(GSH) levels, mitochondrial integrity, receptor-dependent or independent signal transduction
pathway activity (e.g., pro-survival vs. pro-cell death balance), or redox state, for example. Although
hepatocellular hypertrophy is commonly attributed to receptor-dependent organellar growth and
proliferation (e.g., PPAR mediated), the milieu of pathways involved in modulating hepatocyte
structural and functional response to chemicals are diverse (Williams and Iatropoulos. 2002). For
example, hepatocyte swelling also has been associated with cell death processes, in particular
oncosis or oncotic necrosis (Kleiner etal.. 2012). Several liver diseases or conditions, such as
ischemia-reperfusion injury, drug-induced liver toxicity, and partial hepatectomy, have noted
oncosis (oncotic necrosis) upon cellular/tissue examination (for review see: Kass f20061: Taeschke
This document is a draft for review purposes only and does not constitute Agency policy.
3-32 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
and Lemasters (2003)1 and are not dependent on peroxisome proliferation or PPAR signaling.
Rather, cellular alterations such as a transition in mitochondrial membrane permeability and
caspase activation (especially Caspase-8) have been identified as key mediators or tipping points
for a shift from a hypertrophic (oncotic) hepatocellular phenotype to apoptotic or primary necrotic
cell death fMalhi etal.. 2006: Van Cruchten and Van Den Broeck. 20021. As such, an assumption that
chemical-induced hepatocellular hypertrophy is by default a distinctly proliferative/growth
response associated exclusively with PPAR signaling might not be accurate.
One study investigated the activation of PPARa and pregnane X receptor (PXR) in the livers
of exposed neonatal mice fDas etal.. 20081. This study showed the expression of genes associated
with either PPARa or PXR was not increased in the livers of neonatal male and female mice,
possibly indicating that the increased liver weights in these animals were associated with a non-
PPARa or PXR MOA. No other PFBA-specific studies investigated activation of other isoforms of
PPAR (e.g., PPARy) or additional pathways (e.g., constitutive androstane receptor [CAR] or
pregnane X receptor [PXR]); however, evidence from human cell culture experiments involving
PFOS and PFOA, two of the most heavily studied PFAS, suggest the possibility of other non-PPARa
MOAs operational in liver toxicity. For example, PFOA and PFOS exposure is associated with PPARy
activation fBeggs etal.. 2016: Buhrke etal.. 20151. and increased mRNA levels of CAR and PXR
responsive genes fAbe etal.. 2017: Zhang etal.. 20171. Activation of these hepatic nuclear
receptors plays an important role in regulating responses to xenobiotics and in energy and nutrient
homeostasis (di Masi et al.. 2009). Animal studies of other PFAS also provide some evidence
suggesting that nuclear receptor pathways other than PPARa might be involved in PFAS-induced
liver effects. For example, two separate in vivo studies using PPARa null animal models report
increases in absolute and relative liver weight fDas etal.. 2017: Rosen etal.. 20171 and in
hepatocellular hypertrophy and lipid accumulation fDas etal.. 20171 following PFHxS or PFNA
exposure. Multiple in vivo studies have also evaluated activation of CAR and PXR in rodents
exposed to PFDA: PFDA exposure in wild-type C57BL6/6J mice led to increased nuclear
translocation of CAR and mRNA levels of CAR/PXR responsive genes [CYP2B10 and CYP3A11; Abe
etal. (2017)]: these effects were not observed in CAR or PXR null mice. PFDA has also been
observed to activate PXR in human HepG2 cells (Zhang etal.. 2017) and increase mRNA levels of
CAR/PXR-regulated genes (CYP2B6 and CYP3A4) in primary human hepatocytes f Rosen etal..
2013).
In addition to hypertrophy, Foreman etal. f20091 also observed additional
histopathological effects. Hepatic focal necrosis was statistically significantly increased following
exposure of wild-type mice to >175 mg/kg-day PFBA. Although no statistically significant increases
in focal necrosis were observed at any dose in PPARa null or humanized mice, necrosis did increase
slightly in the highest dose compared to controls (2/10 vs. 0/10) in hPPARa; that exposure to
higher doses of PFBA would elicit increased necrotic effects in hPPARa mice is possible.
Interestingly, no statistically significant increase in focal necrosis was observed in any mouse strain
This document is a draft for review purposes only and does not constitute Agency policy.
3-33 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
treated with Wy-14,643 in this study. That PFBA exposure resulted in liver necrosis in wild-type
mice, but not PPARa null mice, suggests that PPARa is required for the development of this lesion.
The observation that the positive control for PPARa activation, Wy-14,643, however, also did not
result in this lesion (in this study), as well as suggestive evidence of increased necrosis in hPPARa
mice, supports that a PPARa-independent, complementary or multifaceted MOA could be active in
the observed liver toxicity. Supporting this conclusion is the observation that centrilobular and
periportal vacuolation (i.e., lipid accumulation) was increased compared with controls in PPARa
null and humanized mice after exposure to 350 mg/kg-day PFBA, with greater vacuolation in
PPARa null mice than in humanized mice. Vacuolation was not reported in wild-type mice, and
results for the vacuolation endpoints were provided only for the control and low-dose groups for
the PPARa null and hPPARa mice. This result is consistent with Das etal. f20171 who reported
PFAS increased accumulation and oxidation of lipids in the liver of exposed mice, with
accumulation occurring faster than oxidation. Thus, although vacuolation occurs in humanized
PPARa mice, oxidation is also induced (as evidenced by the upregulation of ACO), limiting lipid
accumulation to a degree. In PPARa null mice, however, accumulation of lipids in the liver of
exposed animals must be occurring through a PPARa-independent mechanism. Thus, PFBA
appears to result in increased lipid accumulation in the liver via a PPARa-independent mechanism,
and although humanized mice do exhibit an increase in (3-oxidation via ACO upregulation, this
increase in lipid catabolism is not sufficient to overcome the increased lipid deposition in the liver.
The observation of increased liver weight, increased incidence of hepatocellular
hypertrophy, vacuolation, and necrosis in wild-type and humanized PPARa mice is important when
considered in the context of the recommendations of the Hall etal. (2012) paper. In interpreting
"histological changes caused by an increase in liver weight"—exactly the situation observed in
PFBA-exposed hPPARa mice in Foreman etal. f20091— Hall etal. f20121 suggests that coincident
histological evidence of liver injury/damage can be used to support the conclusion that the liver
weight increases/histological changes (i.e., hypertrophy) are adverse. Among the histological
changes that Hall etal. (2012) identifies as sufficient supporting evidence is necrosis and steatotic
vacuolar degeneration, with the study authors further differentiating between macrovesicular
vacuolation (considered nonadverse) and microvesicular vacuolation. Microvesicular vacuolation
is described by the presence of hepatocytes partially or completely filled with multiple small
vacuoles without displacement of the nucleus fKleiner and Makhlouf. 20161. This pattern of
vacuolation is precisely what Foreman et al. f20091 observed in hPPARa mice exposed to PFBA.
Additionally, focal necrosis is observed in wild-type mice in Foreman etal. (2009). Thus, according
to the Hall recommendations, observation of liver weight increases, hypertrophy, microvesicular
vacuolation, and necrosis across wild-type and hPPARa mice is consistent with a determination
that these interconnected PFBA-induced liver effects meet the criteria for adversity.
Accumulation of lipids in the liver is an apical key event (decreased fatty acid efflux
resulting in lipid accumulation) leading to hepatic steatosis fAngrish etal.. 2016: Kaiser etal.. 20121
This document is a draft for review purposes only and does not constitute Agency policy.
3-34 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Toxicological Review ofPFBA and Related Salts
and has been observed in animal toxicological studies following exposure to numerous
environmental agents that ultimately cause steatosis floshi-Barve etal.. 2015: Wahlang et al..
20131. Sustained steatosis can progress to steatohepatitis and other adverse liver diseases such as
fibrosis and cirrhosis fAngrish etal.. 20161. Therefore, that vacuolation occurring in null PPARa
mice indicates a PPARa-independent mechanism for lipid accumulation in the liver, possibly as a
precursor to more severe forms of liver injury. The occurrence of vacuolation in humanized mice
further supports the human relevance of the observed hepatic toxicity.
Overall, evidence specific to PFBA and from other potentially relevant PFAS provides
support for both PPARa dependent and independent pathway contributions to hepatic toxicity, and
further, that activation of humanized PPARa by PFBA can likewise result in hepatic effects of
concern. Additionally, application of the recommendations from Hall etal. f20121 clearly supports
the conclusion that the multiple and interconnected effects observed in the livers of exposed
animals meet the criteria for adversity.
Evidence Integration Summary
No association between PFBA and circulating levels of multiple serum biomarkers of
hepatic injury were observed in the only available, medium confidence epidemiological study with
reduced sensitivity fNian etal.. 20191. These null findings from a single study with low sensitivity
did not influence the evidence integration judgments, providing indeterminate evidence.
Hepatic effects associated with oral exposures to PFBA have been consistently observed in
high or medium confidence short-term and subchronic studies in adult mice or rats of both sexes
(Bute nhoff etal.. 2012: Foreman et al.. 2009: van Otterdiik. 2007a. b; Permadi etal.. 1993: Permadi
etal.. 19921 and in a developmental toxicity study in mice fDas etal.. 20081. Overall, changes in
liver weights and histopathology (hepatocellular hypertrophy) were consistently observed across
two species, with effects occurring in male adult rats and mice, female pregnant or nonpregnant
adult mice, and in male and female neonatal mice. In particular, increases in liver weight and
hepatocellular hypertrophy incidence occurred at similar dose levels across species, occurred at
multiple doses, and appeared to be dose related (i.e., increasing magnitude of effect with increasing
dose). Although uncertainties remain, given the consistency, coherence, and inferred adversity (see
below) of these findings, there is moderate animal evidence for hepatic effects ofPFBA exposure.
Increased liver weights were consistently observed in male, but not female, adult rats
following 28- or 90-day exposures fButenhoff et al.. 2012: van Otterdiik. 2007a. b) and in male
wild-type and hPPARa mice, pregnant and nonpregnant female mice, and neonatal male and female
mice on PND 1 (Foreman etal.. 2009: Das etal.. 2008: Permadi etal.. 1993: Permadi et al.. 1992).
For male rodents, the doses at which effects occurred appeared to differ appreciably across species,
but wild-type PPARa mice seemed to exhibit greater magnitudes of effect vs. humanized PPARa
mice or rats. As noted above, female pregnant and nonpregnant mice, along with their offspring,
exhibited effects only at higher doses compared with adult male rats and mice, possibly relating to
This document is a draft for review purposes only and does not constitute Agency policy.
3-35 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
the observation that female rodents eliminate PFBA much more rapidly than males (see Section
3.1.4).
Liver histopathology was also consistently observed across PFBA studies fButenhoffetal..
2012: Foreman et al.. 2009: van Otterdiik. 2007a. b), although differences in the type or severity of
lesions differed somewhat across species and durations of exposure. Wild-type and hPPARa mice
were both observed to develop hepatocellular hypertrophy following 28 days of oral exposure to
PFBA, whereas only wild-type mice developed hepatic focal necrosis (Foreman et al.. 20091. PPARa
null mice developed neither of these lesions in response to exposure. Adult male rats also were
observed to develop hepatocellular hypertrophy, but not coagulative necrosis, following 28 or
90 days of exposure fButenhoffetal.. 2012: van Otterdiik. 2007a. b). Again, differences in
toxicokinetics might explain somewhat the differences in lesion incidence across species, with rats
eliminating PFBA much more rapidly than mice. Interestingly, PPARa null and hPPARa mice were
observed to develop centrilobular and periportal vacuolation, whereas wild-type mice did not This
possibly indicates the accumulation of lipids within the liver. Increased liver weights were
concurrently observed at all doses with hepatocellular hypertrophy in wild-type and hPPARa mice
following short-term exposure fForeman et al.. 20091. In wild-type mice, however, liver weight
increases occurred at lower doses than did focal necrosis in the same study fForeman et al.. 20091.
although focal necrosis was not observed in hPPARa mice in the presence of liver weight changes at
any dose. In male rats, changes in liver weight occurred at lower doses than hepatocellular
hypertrophy following 28-day exposures, whereas both effects were observed at the same dose
following 90-day exposures fButenhoffetal.. 2012: van Otterdiik. 2007a. b).
Changes in serum biomarkers of liver function or injury were not consistently observed
across exposure durations or concurrently with hepatocellular lesions. In the 28-day study in rats,
prothrombin time alterations were observed only at 150 mg/kg-day; no changes in ALT, AST, or
ALP were observed. Although increased ALP and increased hepatocellular hypertrophy were both
observed in male rats exposed to 30 mg/kg-day for 90 days in the subchronic study, no concurrent
increase in ALT and AST was observed at this exposure level. Further, the observed decreased
bilirubin is inconsistent with what would be expected as a marker of liver injury (i.e., an increase in
bilirubin); therefore, this observation is of unclear toxicological significance. Lastly, cholesterol
levels were decreased in a dose-dependent manner following the 28-day, but not the 90-day,
exposure. As a whole, the various clinical chemistry endpoints, as measurements of liver toxicity,
are inconsistent across endpoints and durations of exposure, and thus did not influence the
evidence integration judgments.
One characteristic of the evidence base for PFBA is the sparsity of chemical-specific
mechanistic data to inform the human relevance of the observed increases in liver weight and
hypertrophic lesions in rats and mice. In the one study that does provide chemical-specific
information, PFBA exposure to wild-type and hPPARa mice increased both liver weights and
hepatocellular hypertrophy. Only wild-type mice were observed to develop focal necrosis, possibly
This document is a draft for review purposes only and does not constitute Agency policy.
3-36 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofPFBA and Related Salts
indicating that activation of PPARa was a necessary step in the MOA for developing this lesion.
Hepatic focal necrosis, however, was not observed in any group (wild-type, hPPARa, or PPARa null
mice) exposed to the positive control (the PPARa activator Wy-14,643) in wild-type mice. Further,
increased vacuolation was reported only in PPARa-null and hPPARa mice, an observation
consistent with in vivo evidence for longer chain PFAS fDas etal.. 20171. This observation
(increased vacuolation) in PPARa-null and humanized mice indicates that lipid accumulation in the
liver occurs, at least in part, through a PPARa-independent mechanism, and that either the lack, or
attenuated activity, of PPARa-induced lipid catabolism is not sufficient to overcome the increased
accumulation. This strongly suggests a complementary or multifaceted MOA for development of
PFBA-induced hepatic effects. Indeed, based on evidence from other PFAS chemicals, non-PPARa
mechanisms relevant to hepatic effects are apparent. In vivo and in vitro studies of PFOA, PFOS,
PFDA, and PFNA demonstrate that PFAS exposure can activate PPARy, CAR, and PXR fAbe etal..
2017: Das etal.. 2017: Zhang etal.. 2017: Beggs etal.. 2016: Buhrke etal.. 2015: Rosen etal.. 2013)
and that activation of these receptors results in the hepatic effects observed in PPARa null mice.
Thus, multiple lines of evidence, taken as a whole, indicate that the liver toxicity observed in
rodents due to PFBA exposure is likely adverse, relevant to humans, and dependent on multiple
biological pathways (i.e., both PPARa-dependent and independent pathways). Even considering a
PPARa-only MOA, human PPARa is observed to be activated by PFBA exposure in vitro, and
evidence in humanized PPARa mice (increased liver weight and increased hepatocellular
hypertrophy, which is observed to be more severe than that in wild-type mice) indicates the
PPARa-mediated components of the undefined MOA(s) appear relevant to human toxicity, given the
effects are observed in animals with human PPARa receptors. Further, the existing evidence base
also supports the operation of PPARa-independent pathways for other hepatotoxic effects, given
the direct observation of increased vacuolation in PPARa null mice in response to PFBA exposure,
an observation also occurring in humanized PPARa mice. Even in the absence of PPARa activity,
hepatic toxicity occurs that is possibly the precursor to more clearly adverse liver disease
(e.g., steatohepatitis, fibrosis, and cirrhosis). Thus, although there is uncertainty in relating the
sensitivity of hepatic changes observed in rodents to humans given the generally decreased
sensitivity of human responses to PPARa agonism, evidence from PFBA studies and studies on
other PFAS indicates that PPARa alone cannot be identified as the exclusive MOA for PFBA-induced
liver effects. Lastly, independent of conclusions regarding PPARa as the MOA, consideration of the
recommendations from Hall etal. f20121 also support a determination that the observed hepatic
effects in rodents are relevant to humans. Hall etal. (2012) indicates coincident histological
evidence of liver injury/damage can be used to support the conclusion that liver
weight/hypertrophic effects are adverse. That PFBA induces a constellation of effects in the liver,
including increased liver weight, hypertrophy, vacuolation, and necrosis is clear from the in vivo
evidence in rodents. Therefore, according to Hall etal. f20121. these coincident effects are
This document is a draft for review purposes only and does not constitute Agency policy.
3-37 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
Toxicological Review ofPFBA and Related Salts
consistent with the conclusion that PFBA-induced liver effects in rodents meet the criteria for
adversity.
The available animal evidence for effects on the liver includes multiple high and medium
confidence studies with consistent effects across multiple species, sexes, exposure durations, and
study designs (e.g., exposures during pregnancy); it exhibits coherence between the effects on liver
weights and histopathology and a clear biological gradient (increasing effect with increasing dose);
and the evidence is interpreted to be relevant to humans. Taken together, the available evidence
indicates that PFBA exposure is likely to cause hepatic toxicity in humans (see Table 3-8), given
relevant exposure circumstances. This judgment is based primarily on a series of short-term,
subchronic, and developmental studies in rats and mice, generally exhibiting effects at PFBA
exposure levels >30 mg/kg-day.
This document is a draft for review purposes only and does not constitute Agency policy.
3-38 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Table 3-8. Evidence profile table for hepatic effects
Evidence Stream Summary and Interpretation
Evidence Integration
Summary Judgment
Evidence from studies of exposed humans (see Section 3.2.2: Human Studies)
Studies, outcomes,
and confidence
Summary of key
findings
Factors that increase
certainty
Factors that decrease
certainty
Judgments and rationale
©©O
Evidence indicates (likely)
Serum Biomarkers
1 medium confidence
study;
1 low confidence study
• No association
between PFBA and
liver biomarkers or
blood lipids in studies
with poor sensitivity
• No factors noted
• No factors noted
ooo
Indeterminate
Primary basis:
Three high and one medium
confidence studies in male
adult rats and mice and
maternal and neonatal mice
(short-term, subchronic, and
gestational exposures) at
>30 mg/kg-d PFHxA
Evidence from in vivo animal studies (see Section 3.2.2: Animal Studies)
Studies, outcomes,
and confidence
Summary of key
findings
Factors that increase
certainty
Factors that decrease
certainty
Judgments and rationale
Human relevance:
Effects in rats are considered
relevant to humans (see
Section 3.2.2: Mechanistic
Evidence and Supplemental
Information)
Cross-stream coherence:
N/A (human evidence
indeterminate)
Susceptible populations and
lifestages:
None identified, although
those with preexisting liver
disease could be at greater
risk
Organ Weight
4 hiah, 2 medium, and
1 low confidence
studies in adult rats
and maternal and
neonatal mice:
• 14-d (x3)
• 28-d (x2)
• 90-d
• Gestational
• Increased liver
weight observed in:
o male adult rats at
>30 mg/kg-d
o female mice and
PND1 neonates at
>175 mg/kg-d
o male wild-type
PPARa and hPPARa
mice at >35 mg/kg-d
(no effects in PPARa
null mice)
• Reduced effects in
female rats could be
attributable to
toxicokinetics
• Consistent
increases, across
most studies (one
null study)
• Dose-response in
most studies (one
null study)
• Coherence with
histopathology in
male rats and mice
(especially at high
dose)
• Magnitude of
effect, up to 112%
• High and medium
confidence studies
• No factors noted
®©o
Moderate
Findings were considered
adverse, consistent, dose
dependent, and biologically
coherent across multiple
measures of hepatic
toxicity. PPARa-
dependence appears likely
for some effects (focal
necrosis) but not others
(vacuolation)
This document is a draft for review purposes only and does not constitute Agency policy.
3-39 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Evidence Stream Summary and Interpretation
Evidence Integration
Summary Judgment
Histopathology
2 hiah and 1 medium
confidence studies in
adult rats and mice:
• 28-d (x 2)
• 90-d
• Hepatocellular
hypertrophy
observed in:
o male adult rats at
30 mg/kg-d
(subchronic)
o male wild-type
PPARa and hPPARa
mice at >35 mg/kg-d
(short-term)
• Focal necrosis
observed in male
wild-type PPARa
mice exposed to
>175 mg/kg-d (short-
term)
• Vacuolation
observed in male
PPARa-null and
hPPARa mice at
350 mg/kg-day
(short-term)
• Reduced effects in
female rats could be
attributable to
toxicokinetics
• Consistent cellular
hypertrophy or
focal necrosis
across studies and
species
• Coherence with
liver weight effects
(especially at high
doses)
• Dose-response
• High and medium
confidence studies
• No factors noted
Other inferences: the MOA
for liver effects is not fully
established, although
available evidence indicates
that multiple pathways are
likely involved
Serum Biomarkers
2 high confidence
studies in adult rats:
• 28-d
• 90-d
• Increased ALP and
decreased bilirubin in
male or male and
female rats,
respectively, at
30 mg/kg-day
• High confidence
studies
• Incoherent
observations (e.g.,
increased ALP but not
ALT or AST; bilirubin
increase not decreased
as expected)
Mechanistic evidence and supplemental information (see subsection above)
This document is a draft for review purposes only and does not constitute Agency policy.
3-40 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Evidence Stream Summary and Interpretation
Evidence Integration
Summary Judgment
Biological events or
pathways
Summary of key findings, interpretation, and limitations
Evidence stream judgment
Molecular Initiating
Events—PPARa
Key findings and interpretation:
• In vitro increased expression of PPARa-responsive genes in primary
rata and human hepatocytes and cells transfected with rat or human
PPARa.
• In vivo increased expression of PPARa-responsive genes in wild-type
and hPPARa mice.
Limitations: small database investigating PPARa activation, some
inconsistencies regarding the strength of activation or interspecies
differences.
Overall, studies in rodent
and human in vitro and in
vivo models suggest that
PFBA induces hepatic
effects, at least in part,
through PPARa. The
evidence also suggests a
role for PPARa-
independent pathways in
the MOA for noncancer
liver effects of PFBA.
Molecular Initiating
Events—Other
Pathways
Key findings and interpretation:
• Indirect evidence of alternative pathways following observation of
effects in humanized PPARa mice exposed to PFBA.
• Direct evidence from other PFAS (PFOA, PFOS, PFDA, PFHxA, PFHxS)
that multiple non-PPARa pathways (PPARy, CAR, PXR) activated
following exposure.
Limitations: No PFBA-specific in vitro data; only one in vivo study providing
indirect evidence.
Organ Level Effects
Key findings and interpretation:
• Observation of increased liver weight and increased hepatocellular
hypertrophy/vacuolation in humanized PPARa mice.
• Concurrent observation that a known PPARa activator (Wy-14,643) did
not elicit the same effects (focal necrosis) as PFBA exposure in wild-
type mice.
Limitations: Only one in vivo study.
This document is a draft for review purposes only and does not constitute Agency policy.
3-41 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Toxicological Review ofPFBA and Related Salts
3.2.3. Developmental Effects
This section describes studies ofPFBA exposure and potential early life effects or
developmental delays and effects attributable to developmental exposure. The latter includes all
studies where exposure is limited to gestation or early life. As such, this section has some overlap
with evidence synthesis and integration summaries for other health systems where studies
evaluated the effects of developmental exposure (see Sections 3.2.2 and 3.2.4 on potential "Hepatic
Effects" and "Reproductive Effects," respectively). Synthesis descriptions of studies across sections
can vary in detail, depending on the impact the data have on summarizing the evidence relevant to
that hazard; typically, earlier hazard sections will include a more detailed discussion that is then
cited in later sections.
Human Studies
The one epidemiological study that investigated developmental effects (birth weight,
gestational age) (Li etal.. 2017a) was a cross-sectional study based on umbilical cord PFBA
exposure_deemed low confidence primarily due to concerns over participant selection and
exposure measurement Li etal. f2017al reported a mean birth weight deficit of -46 grams
(95%CI: -111, 19) in the overall population per each unit (ng/mL) PFBA increase; this was driven
by the association in boys (-86 grams; 95%CI: -180, 9) as the results were null in girls. The
exposure range in this study, however, is quite small and a one-unit increase is beyond the bounds
of the exposure range in this population. Thus, when expressed on an IQR unit change, they
reported small birth weight deficits (-4 grams (95%CI: -10, 2) per each PFBA IQR unit change
(0.09 ng/mL) and in boys (-8 grams; 95%CI: -16,1). No association was observed with gestational
age in weeks.
Animal Studies
A standardized suite of potential developmental effects was evaluated in one high
confidence developmental toxicity study in mice (Das etal.. 2008). Some outcome-specific
considerations for study evaluations were influential on the overall study rating for developmental
effects, but none of these individual domain-specific considerations were judged deficient, and the
Das etal. f20081 study considered further in this section was rated as high confidence (see
Figure 3-7). Endpoints evaluated in the study included time to eye opening, full litter resorption,
postnatal survival, vaginal opening, preputial separation, body weights, and morphological
evaluations (see Table 3-9 and Figure 3-8).
This document is a draft for review purposes only and does not constitute Agency policy.
3-42 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
Toxicological Review ofPFBA and Related Salts
Reporting quality -
Allocation -
Observational bias/blinding -
Confounding/variable control -
Selective reporting and attrition -
Chemical administration and characterization
Exposure timing, frequency and duration
Endpoint sensitivity and specificity
Results presentation
Overall confidence
Legend
| Good (metric) or High confidence {overall)
Adequate (metric) or Medium confidence (overall)
Deficient (metric) or Low confidence (overall)
Critically deficient (metric) or Uninformative (overall)
I NR| Not reported
Figure 3-7. Evaluation results for animal studies assessing developmental
effects of perfluorobutanoic acid (PFBA) exposure (see interactive data
graphic for rating rationales!.
Oral exposure via gavage from GD 1 to 17 of CD-I mice (male and female offspring were
evaluated) to NH4+PFB resulted in delayed eye opening by 1.1,1.4, and 1.5 days compared to
controls at 30,175, and 350 mg/kg-day, respectively fDas etal.. 20081. Significantly increased full
litter resorptions also occurred at 350 mg/kg-day (28 vs. 7% in controls), although no effects were
observed on the number of implants or live fetuses. Additionally, although not statistically
significant, postnatal survival was consistently reduced at PNDs 7,14, and 21 by approximately 5%
The male and female pubertal landmarks (preputial separation and vaginal opening, respectively)
were delayed. Preputial separation was delayed by 2.3 days at 350 mg/kg-day although vaginal
opening was delayed 3.3 and 3.6 days (175 and 350 mg/kg-day, respectively). No changes were
observed in neonatal or postweaning body weight. Anatomical changes were observed (renal
dilation, fetal hydronephrosis, and absent testis) but were randomly distributed among the
treatment groups, including controls, and thus were not attributable to PFBA exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
3-43 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Table 3-9. Developmental effects observed following perfluorobutanoic acid
(PFBA) exposure in a developmental toxicity study
Animal group
Dose (mg/kg-d)
0
35
175
350
Full-litter resorptions; pregnant P0 female CD-I mice on GD 18
(Das et al., 2008)
2/29
1/29
4/28
8/29
Survival to PND 1 (%); Fi male and female CD-I mice on PND 1
(Das et al,, 2008)
91.7 ±2.1
90.2 ±2.4
92.9 ± 1.6
87.9 ±2.6
Survival to PND 7 (%); F-: male and female CD-I mice on PND 7
(Das et al., 2008)
90.9 ±2.3
90.0 ±2.3
90.0 ±3.1
86.4 ±2.7
Survival to PND 14 (%); Fi male and female CD-I mice on PND 14
(Das et al., 2008)
90.9 ±2.3
89.7 ±2.4
89.6 ±3.2
85.7 ±3.0
Survival to PND 21 (%); Fi male and female CD-I mice on PND 21
(Das et al., 2008)
90.9 ±2.3
88.7 ±2.4
89.6 ±3.2
85.7 ±3.0
Delayed eye opening (d); Fi male and female CD-I mice
(Das et al., 2008)
16.28 ± 1.19
17.38 ±0.79
17.69 ±0.68
17.8 ±0.83
Delayed vaginal opening (d); Fi female CD-I mice
(Das et al., 2008)
31.25 ±2.62
33.71 ±2.59
34.57 ±2.59
34.92 ±2.23
Delayed preputial separation (d); Fi male CD-I mice
(Das et al., 2008)
29.55 ± 1.14
30.21 ± 1.99
30.56 ± 1.84
31.88 ± 1.72
Study Name Endpoint Name
Study Type
Animal Description
Observation Time
Das, 2008, 1290825 Full Liter Resorption (FLR) Number
17 Day Oral
PO Mouse CD-I (2)
18.0 days
FuJI Lrtler Resorption (FLR) Liller Implants
17 Day Oral
P0 Mouse CD-I <2>
IS O days
Live Fetuses
17 Day Oral
FI Mouse. CD-I (£3)
18,0 days
Fetal Weight
17 Day Oral
FI Mouse CD-1 (='=)
18.0 days
Fetal Renal Dilation
17 Day Oral
FI Mouse CD-1 (£2)
18.0 days
Felal Hydronephrosis
17 Day Oral
FI Mouse, CQ-1 (=2)
18,0 days
Fetal Absent Testis
17 Day Oral
FI Mouse, CD-I (£)
18.0 days
Live biter Implants
17 Day Oral
FI Mouse, CD-I (£2)
19 0 days
Live Births
17 Day Oral
FI Mouse CD-1 (-* = )
19,0 days
Live Births/Jmplanls
17 Day Oral
FI Mouse, CD-1 (£3)
19 0 days
Survival lo PND 1
17 Day Oral
FI Mouse CD-1 (-*3)
20.0 days
Survival lo PNO 7
17 Day Oral
FI Mouse CD-1 (££)
26.0 days
Survival lo PNO 14
17 Day Oral
F1 Mouse CD-1 (£3)
33.0 days
Survival lo PND 21
17 Day Oral
FI Mouse CD-1 (£3)
40.0 days
Eye Opening
17 Day Oral
FI Mouse, CD-I (£3)
29.0 days
Vaginal Opening
17 Day Oral
FI Mouse, CD-1 (3)
45.0 days
Preputial Separation
17 Day Oral
FI Mouse CD-I (£)
45.0 days
PFBA Developmental Effects
# Doses
A Treatmenl-Related Increase
Treatment-Related Decrease
I—\ Dose Range
150 200
Axis label
250 300 350 400
Figure 3-8. Pre- and postnatal developmental responses to gestational
ammonium perfluorobutanoate (NH4+PFB) exposure (see interactive data
graphic and rationale for study evaluations for developmental effects in
Health Assessment Workspace Collaborative [HAWC]).
This document is a draft for review purposes only and does not constitute Agency policy,
3-44 ' DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Toxicological Review ofPFBA and Related Salts
Evidence Integration Summary
One low confidence human study reported lower birth weight in boys with higher PFBA
exposure. No association was observed with gestational age. The lack of additional studies with
lower risk of bias reduces the interpretability of these findings. Overall, the evidence on potential
developmental effects from studies of humans exposed to PFBA was indeterminate.
Coherent effects on developmental maturation were observed in one high confidence study
in mice (Das etal.. 20081 following in utero exposure to PFBA. The developmental effects ofPFBA
exposure in this study included delayed eye opening, full-litter resorption, decreased survival, fetal
absent testis, and delays in vaginal opening and preputial separation, although pup growth and
body weight were unaffected. These effects indicate that PFBA appears to disrupt the normal
gestational and postnatal development of exposed fetuses. One factor increasing the strength of
evidence is that effects on the developing fetus (e.g., delayed eye opening, delays in the
development of the male and female reproductive systems) are seen following exposure to other
PFAS, most notably the structurally related compound perfluorobutane sulfonate (U.S. EPA. 2018b).
but other, longer chain PFAS as well. Following exposure to >200 mg/kg-day PFBS (U.S. EPA.
2018b) or 5 mg/kg-dayperfluorooctanoic acid [PFOA; Lau etal. f20061] or perfluorooctane
sulfonate [PFOS; Lau etal. f2004-1]. similar delays in eye opening (~1.5 d) were observed in mice.
Similarly, following exposure to >200 mg/kg-day PFBS, time to vaginal opening was increased by
>3 days (Feng etal.. 2017) and time to vaginal patency was increased ~3 days in mice exposed to
20 mg/kg-day PFOA (Lau etal.. 2006) and ~2 days in rats exposed to 30 mg/kg-day PFOA
(Bute nhoff etal.. 2004). Time to pubertal milestones was also delayed in male rodents exposed to
PFOA: Preputial separation was delayed ~1.5 days in mice exposed to 20 mg/kg-day (Lau etal..
20061 and ~2 days in rats exposed to 30 mg/kg-day PFOA fButenhoff et al.. 20041. Thus,
qualitatively, a consistent pattern of delayed pubertal milestones is observed following exposure to
related PFAS, increasing certainty in the evidence available for PFBA. Further, the absence of
effects on body weight in PFBA-exposed offspring strengthens the confidence that the observed
developmental delays are biologically significant, adverse effects. Taken together, the available
animal studies provided moderate evidence of potential developmental effects.
Data gaps in the developmental toxicity database include a lack of information on the
thyroid and nervous system following gestational exposure. Given that other PFAS (i.e., PFBS) alter
thyroid hormone levels following gestational exposure and that PFBA induces changes in thyroid
hormone levels in exposed adult animals, PFBA also might alter normal thyroid function in the
developing fetus. As both PFBA and PFBS evidence bases lack studies on developmental
neurotoxicity, a potential consequence of altered thyroid function during development, this
represents an important unknown.
Thus, considering the coherent suite of developmental effects, primarily developmental
delays, observed following PFBA exposure in one high confidence study, and similar effects
observed following exposure to multiple other PFAS (including the structurally similar PFBS), the
This document is a draft for review purposes only and does not constitute Agency policy.
3-45 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
1 evidence indicates PFBA exposure is likely to cause adverse developmental effects in humans (see
2 Table 3-10), given relevant exposure circumstances. The basis for this judgment is a single high
3 confidence gestational exposure study in mice, with multiple adverse effects occurring at PFBA
4 exposure levels >175 mg/kg-day (with delays in eye opening occurring at >35 mg/kg-day).
5 Notably, even in the absence of evidence informing potential similarities of effects between PFBA
6 and other PFAS regarding gestational thyroid function, the available PFBA-specific developmental
7 effects alone support this judgment
This document is a draft for review purposes only and does not constitute Agency policy.
3-46 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Table 3-10. Evidence profile table for developmental effects
Evidence Stream Summary and Interpretation
Evidence Integration
Summary Judgment
Evidence from studies of exposed humans (see Section 3.2.3: Human Studies)
0®Q
Evidence indicates (likely)
Primary basis:
One high confidence gestational
study in mice, with effects
observed at >35 mg/kg-d PFBA
Human relevance:
In the absence of evidence to
the contrary, the
developmental effects observed
in mice are considered relevant
to humans based on conserved
biological processes
Cross-stream coherence:
N/A (human evidence
indeterminate)
Susceptible populations and
lifestages:
Pregnancy and early life
Other inferences:
PFBA-induced developmental
effects are consistent with
effects seen for other PFAS (see
Section 3.2.3: Evidence
Integration Summary
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that decrease
certainty
Judgments and
rationale
Birth Weight
1 low confidence study
• Birth weight deficit with
higher PFBA exposure in
boys (nonstatistically
significant)
• No factors noted
• Low confidence study
• Imprecision
ooo
Indeterminate
Evidence from in vivo animal studies (see Section 3.2.3: Animal Studies)
Studies, outcomes, and
confidence
Summary of key findings
Factors that increase
certainty
Factors that decrease
certainty
Judgments and
rationale
Developmental
Milestones
1 hiqh confidence
gestational study in
mice
• Dose-dependent delays
in developmental
milestones in:
o Eye opening in males
and females at >
35 mg/kg-d
o Preputial separation
in males at 350
mg/kg-d
o Vaginal opening in
females at 175 and
350 mg/kg-d
• Increased full litter
resorption at 350 mg/kg-
d
• No effects on pup weight
• Dose-response
gradient
• Coherence across
developmental
milestones
• Magnitude of
effect, up to 12%
increase in time to
milestone and
4-fold increase in
full litter
resorptions
• High confidence
study
• No factors noted
®©o
Moderate
Coherent delays in
developmental
milestones, with
multiple alterations
observed at
>35 mg/kg-d
This document is a draft for review purposes only and does not constitute Agency policy.
3-47 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Toxicological Review ofPFBA and Related Salts
3.2.4. Reproductive Effects
Human Studies
One low confidence cross-sectional study fSong etal.. 20181 examined the association
between PFBA exposure and semen parameters. No evidence of an association between PFBA
exposure and decreased semen quality was found (correlation coefficients were -0.03 for semen
concentration and 0.2 for progressive motility), although issues were noted during study evaluation
regarding the ability of this study to detect an effect due to the small sample size [n = 58) and risk of
outcome misclassification, which makes the null finding difficult to interpret Other study
deficiencies including the potential for selection bias and confounding were noted in the study
evaluation, but the direction of these biases is unknown.
Animal Studies
Two high confidence studies reported in three publications from the same research group
(Bute nhoff etal.. 2012: van Otterdiik. 2007a. b) evaluated the effects ofPFBA exposure on
reproductive organ weights in rats (see Figure 3-9). In addition, one high confidence
developmental toxicity study fDas et al.. 20081 reported several delays in reproductive system
development (e.g., vaginal opening, preputial separation) after gestational exposure. These latter
results are synthesized and integrated with other studies examining developmental outcomes (see
Section 3.2.3) given the apparent coherence of findings of developmental delays after PFBA
exposure and the general lack of other studies or effects on reproduction, including an absence of
studies on functional measures (see discussion below).
Organ weight
Short-term exposure (28 d) to PFBA in male S-D rats increased absolute epididymis weight
(note: absolute organ weights are typically preferred for these reproductive organ measures) 10%
compared to controls, but only at the lowest dose [6 mg/kg-day; Butenhoff et al. (2012): van
Otterdiik (2007a)]. In a separate cohort, this effect was not observed following a 3-week recovery
period (at 49 d) from exposure at any dose (6, 30, or 150 mg/kg-day). Changes in absolute or
relative testis weight were not observed in rats following either 28 days of exposure or during the
recovery period. Similarly, no changes in absolute or relative ovary weight were observed in rats
following short-term (28 d) PFBA exposure and none arose during the recovery period (Butenhoff
etal.. 2012: van Otterdijk. 2007a).
This document is a draft for review purposes only and does not constitute Agency policy.
3-48 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Toxicological Review ofPFBA and Related Salts
Study Name Eridpoint Name
Study Type
Animal Description
Observation Time
Butenhoff, 2012, 128953S Testes Weight, Absolute
28 Day Oral
Rat Sprague-Davriey (£)
28 0 days
49 0 days
Epididymis Weight. Absolute
28 Day Oral
Rat Sprague-Davriey {£)
28 0 days
49.0 days
Testes Weight. Relative
28 Day Oral
Rat Sprague-Davriey (£)
280 days
49 0 days
Epididyrnidis Weight. Relative
28 Day Oral
Rat Sprague-Davriey {£)
28.0 days
49 0 days
Ovary Weight Absolute
28 Day Oral
Rat Sprague-Davriey (3)
28.0 days
Ovary Weight. Relative
28 Day Oral
Rat Sprague-Davriey (2)
28 0 days
490 days
Ovary Weight Absolute
28 Day Oral
Rat Sprague-Davriey (2)
280 days
PFBA Reproductive Effects
0 Doses
A Treatment-Related Increase
^Treatment-Related Decrease
H Dose Range
40 60 30
Axis label
Figure 3-9. Reproductive responses to ammonium perfluorobutanoate
(NH4+PFB) exposure (see interactive data graphic and rationale for study
evaluations for reproductive effects in Health Assessment Workspace
Collaborative [HAWC]).
Evidence Integration Summary
The database of studies examining the potential for PFBA exposure to elicit effects on
reproductive parameters is limited to one human and one animal study. There is evidence for
delayed development of the reproductive system (i.e., delayed vaginal opening and preputial
separation] following gestational PFBA exposure (Das etal.. 2008). These latter results are
synthesized and integrated in the developmental effects section (see Section 3.2.3) and not
discussed further in this section.
In the only available human study (a low confidence study), no association was observed
between semen quality and PFBA exposure. Null findings in a single study with low sensitivity
(biased toward the null) are not interpreted to influence the evidence integration judgments, and
thus the human evidence was indeterminate.
The available animal evidence is sparse, limited to evaluations of reproductive
organ-weight measurements in a high confidence short-term experiment reported in three
publications from the same research group (ButenhoffetaL 2012: van Otterdiik. 2007a. b).
Specifically, the authors evaluated reproductive organ weights in a cohort of rats immediately after
exposures ended and another cohort 21 days postexposure, both of which were largely null. Given
the limited interpretability of these data, the animal evidence was indeterminate.
Given the sparsity of evidence on potential reproductive effects, the relative insensitivity of
the outcome measures (organ weights) in animals, and the largely null findings, there is
insufficient evidence to determine whether PFBA exposure has the potential to cause reproductive
effects in humans (other than the developmental delays discussed in Section 3.2.3; see Table 3-11).
This document is a draft for review purposes only and does not constitute Agency policy,
3-49 ' DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Table 3-11. Evidence profile table for reproductive effects
Evidence Stream Summary and Interpretation
Evidence Integration Summary
Judgment
Evidence from studies of exposed humans (see Section 3.2.4: Human Studies)
OOO
Insufficient Evidence
Primary basis:
One high confidence study in rats
Human relevance:
Organ weight changes in rats are
considered relevant to humans in
the absence of evidence to the
contrary
Cross-stream coherence:
N/A (human evidence
indeterminate)
Susceptible populations and
life stages:
None identified
Studies, outcomes,
and confidence
Summary of key
findings
Factors that increase
certainty
Factors that decrease
certainty
Judgments and
rationale
Birth Weight
1 low confidence
study
• No association
between PFBA
exposure and
semen quality
• No factors noted
• Low confidence
study
ooo
Indeterminate
Evidence from in vivo animal studies (see Section 3.2.4: Animal Studies)
Studies, outcomes,
and confidence
Summary of key
findings
Factors that increase
certainty
Factors that decrease
certainty
Judgments and
rationale
Organ weights
1 hiqh confidence
28-d study in rats
• Increased
epidydimal weight
in rats at 6 mg/kg-d
but not higher
doses
• No changes in
testis or ovary
weights
• No factors noted
• Lack of dose-
response
• Lack of coherence
across
reproductive organ
weights
OOO
Indeterminate
Largely null findings in
in the only available
study that examined
reproductive organ
weights
This document is a draft for review purposes only and does not constitute Agency policy.
3-50 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofPFBA and Related Salts
3.2.5. Other Noncancer Health Effects
In addition to the potential health effects outlined above, some epidemiological studies have
examined the potential for associations between PFBA exposure and blood pressure and renal
function, although several experiments in rats and mice have examined potential effects ofPFBA
exposure on body weight (note: these data were used to inform interpretation of the health effects
discussed in prior sections), hematological effects, and ocular effects. Given the paucity of studies
available and the lack of consistent or coherent effects of PFBA exposure, there is insufficient
evidence to determine whether any of these evaluated outcomes might represent potential human
health hazards ofPFBA exposure. Additional studies on these health effects could modify these
interpretations.
Human Studies
One medium confidence cross-sectional study (Bao etal.. 2017) examined the association
between PFBA exposure and blood pressure and reported statistically significant increased odds of
hypertension (OR = 1.10 [95%CI: 1.04-1.17 per ln-PFBA, ng/mL]) and increased systolic blood
pressure ((3 = 0.80 mm HG [95%CI: 0.25-1.34 per ln-PFBA, ng/mL]). This is despite narrow
exposure contrast (median 0.16 ng/mL, IQR 0.01-0.54). Although this was a medium confidence
study, potential for bias remains; this includes outcome misclassification resulting from the
volatility of blood pressure and its measurement at a single time point and the cross-sectional
design. In the absence of additional confirmatory epidemiological studies, or other supportive
findings (e.g., from animal studies), the results of this observational study alone are interpreted as
"insufficient evidence."
One low confidence cross-sectional study fWangetal.. 20191 examined the association
between PFBA exposure and renal function. They reported statistically significant lower estimated
glomerular filtration rate ((3: -0.5, 95%CI: -0.8, -0.1 [change in GFR (mL/min/1.73 m2) per
1 ln-serum PFAS (ng/mL)]) and higher, though not significant, odds of chronic kidney disease
(OR: 1.1, 95%CI: 1.0,1.2) despite low exposure levels. There is potential for reverse causation in
this association, however. In essence, as described in Watkins etal. (2013). decreased renal
function (as measured by decreased GFR or other measures) could plausibly lead to higher levels of
PFAS, including PFBA, in the blood. This hypothesis is supported by data presented by Watkins et
al. f20131. although the conclusions are somewhat uncertain because of the use of modeled
exposure data as a negative control and the potential for the causal effect to occur in both
directions. Consequently, there is considerable uncertainty in interpreting the results of studies of
this outcome.
Animal Studies
Body-weight changes were evaluated in multiple high and medium confidence short-term
and subchronic-duration studies in rats and mice f Bute nhoff etal.. 2012: Foreman etal.. 2009: Das
This document is a draft for review purposes only and does not constitute Agency policy.
3-51 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
etal.. 2008: van Otterdijk. 2007a. b). In general, no PFBA-related effects on body weight were
observed in any study. Foreman etal. f20091 reported that body weighs were not affected in any
exposure group of Sv/129 mice. Initial and final body weights were statistically significantly lower
in humanized PPARa (hPPARa) Sv/129 mice exposed to 350 mg/kg-day PFBA compared to all
other groups, but this was explained by random assignment of animals; body weights in this group
actually increased slightly during the study, indicating the lower measured body weights were not
treatment related. The change in body weight across the duration of the study was not changed at
any dose in any group of animals, however, indicating PFBA exposure had no deleterious effect on
adult body weight in mice. Maternal, preweaning, and postweaning body weights were not altered
by PFBA exposure in CD-I mice fDas etal.. 20081. Adult body weights were not altered in S-D rats
exposed to PFBA for either 28 or 90 days f Bute nhoff etal.. 2012: van Otterdiik. 2007a. b). PFBA
appears not to affect body weight across multiple species, exposure durations, or lifestages.
Some evidence of effects on the hematological system was observed in S-D rats exposed to
PFBA. Following 28 days of exposure, no effects other than on prothrombin time (PT; a measure of
clotting potential) were observed (van Otterdiik. 2007a. b). In males, PT was statistically
significantly decreased 6% following exposure to 150 mg/kg-day PFBA, whereas in females,
statistically significant decreases of 4 and 5% were observed in the 6- and 30-mg/kg-day dose
groups, respectively. PT was decreased 4% in the 150-mg/kg-day dose group in females, but the
decrease was not statistically significant. Following the recovery period, no statistically significant
decreases in PT were found in male rats, but consistent statistically significant 7-8% decreases in
PT were observed in all exposed female dose groups (p < 0.01). Hematological effects were more
pronounced following 90-day exposures. In males, red blood cell counts, hemoglobin, and
hematocrit were decreased 4, 6, and 5%, respectively, and red blood cell distribution width was
increased 5% following exposure to 30 mg/kg-day PFBA. Although the number of RBCs and the
RBC distribution width were observed to return to control values following recovery, hemoglobin
and hematocrit remained decreased 5% relative to control. Mean corpuscular hemoglobin and
mean corpuscular hemoglobin concentration were decreased 2-3% in female rats exposed to
30 mg/kg-day PFBA. These effects returned to control levels following recovery. Taken as a whole,
although some hematological effects were observed in exposed rats, the effect sizes were quite
small, they generally returned to control levels following a recovery period, and no consistency of
effects across exposure durations or sexes were found.
Ocular effects also were observed in rats exposed to PFBA for 28 or 90 days fvan Otterdiik.
2007a. b). In male rats exposed for 28 days, a delayed bilateral pupillary reflex was observed at
150 mg/kg-day. Although examination of neuronal tissue (including the optic nerve) revealed no
histopathological effects, ocular histological effects were observed. Outer retinal degeneration,
characterized as a loss of 25-30% of photoreceptors, was observed along with a decrease
(20-35%) in retinal thickness. Ocular effects also were also observed in the 90-day subchronic
study: Delays in pupillary dilation were observed at weeks 8 and 12 in rats exposed to
This document is a draft for review purposes only and does not constitute Agency policy.
3-52 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
1 30 mg/kg-day. These delays were reported to be unilateral, not consistent across the treatment
2 period, and low incidence. No ocular histopathological results were observed in the 90-day
3 subchronic study. Thus, although some ocular effects were observed following PFBA exposure,
4 effects across durations were somewhat inconsistent, with greater effects following short-term
5 exposures than in subchronic exposures. This limited the interpretability of the observed effects.
3.3. CARCINOGENICITY
6 No human or animal studies were available to inform the potential for PFBA exposure to
7 cause genotoxicity or cancer. Only one study fCrebelli et al.. 20191 investigated PFBA-induced
8 genotoxicity: No evidence of DNA damage or micronucleus formation was observed in male mice
9 exposed to PFBA via drinking water for 5 weeks.
This document is a draft for review purposes only and does not constitute Agency policy.
3-53 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Toxicological Review ofPFBA and Related Salts
4. SUMMARY OF HAZARD IDENTIFICATION
CONCLUSIONS
4.1. SUMMARY OF CONCLUSIONS FOR NONCANCER HEALTH EFFECTS
The currently available evidence indicates hazards likely exist with respect to the potential
for thyroid, liver, and developmental effects in humans, given relevant PFBA exposure conditions.
These judgments are based on data from short-term (28-d exposure), subchronic (90-d exposure),
and developmental (17-d gestational exposure) oral-exposure studies in rodents. Further
characterizations of the exposure conditions relevant to the identified hazards are provided in
Section 5. A summary of the justifications for the evidence integration judgments for each of the
main hazard sections is provided below, organized by health effect, and further summarized in
Table 4-1.
The hazard identification judgment that the evidence indicates PFBA exposure is likely to
cause thyroid toxicity in humans (given relevant circumstances) is based primarily on a short-term
and subchronic study in male rats reporting a consistent and coherent pattern of hormonal, organ
weight, and histopathological changes, generally at PFBA exposure levels >30 mg/kg-day, although
some notable effects were observed at 6 mg/kg-day. For effects on the thyroid in exposed animals,
PFBA-induced perturbations were observed in one species and sex (male rats) across two different
exposure durations (short-term and subchronic). Consistent, dose-dependent decreases in total
and free T4 were observed independent of any effect on TSH, which is a pattern of hormone
perturbation consistent with hypothyroxinemia. Additionally, increased thyroid weights and
increases in thyroid follicular hypertrophy were observed. Although the observed thyroid
histopathological changes support the potential for PFBA to disrupt the thyroid hormone economy,
however, rodents are uniquely sensitive to the development of thyroid follicular hypertrophy and
tumor development (U.S. EPA. 1998) compared with humans. Because of the similarities in the
production and regulation of thyroid hormone homeostasis between rodents and humans and the
consistency of the observed pattern of effects with changes observed in humans, the effects in
rodents were considered relevant to humans. A detailed discussion of thyroid effects is included in
Section 3.2.1.
The hazard identification judgment that the evidence indicates PFBA exposure is likely to
cause hepatic toxicity in humans, given relevant exposure circumstances, is based primarily on a
series of short-term, subchronic, and developmental studies in rats and mice, generally exhibiting
effects at PFBA exposure levels >30 mg/kg-day. The PFBA-induced effects were observed in two
species and one sex (male rats and mice) across multiple exposure durations (short-term,
subchronic, and gestational). Consistent, coherent, dose-dependent, and biologically plausible
This document is a draft for review purposes only and does not constitute Agency policy.
4-1 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Toxicological Review ofPFBA and Related Salts
effects were observed for increased liver weights and increased incidences of hepatic
histopathological lesions. Supporting the biological plausibility and human relevance of these
effects is mechanistic information that suggests non-PPARa MOAs could explain some of the
observed effects in exposed rodents and that observed effects might be precursors to clearly
adverse health outcomes such as steatosis. Supporting this conclusion is evidence from other PFAS
that have consistently shown that longer chain PFAS can activate non-PPARa nuclear receptors,
including PPARy, CAR, and PXR, although there is uncertainty in inferring a similar relationship for
the short-chain PFBA.
The hazard identification judgment that the evidence indicates PFBA exposure is likely to
cause developmental effects in humans (given relevant exposure circumstances), including
increased prenatal effects (full-litter resorptions) and delays in developmental milestones (days to
eye opening, vaginal opening, and preputial separation) without effects on fetal (pup) growth is
based on a single study in mice exposed gestationally to PFBA. Although the observed
developmental effects due to PFBA exposure were investigated in only one high confidence study,
they demonstrate a constellation of effects affecting the developing organism that is internally
coherent (within-study) and consistent across related PFAS compounds, including PFBS, PFOA, and
PFOS.
There was insufficient evidence to determine whether PFBA exposure has the potential to
cause reproductive toxicity (in adults), effects on hematological or clinical chemistry markers,
ocular effects, changes in blood pressure, or effects on renal function in humans. Other potential
health outcomes have not been evaluated in the context of PFBA exposure. Most notably, potential
for PFBA exposure to affect the immune system, thyroid or nervous system in developing
organisms, or mammary glands represent important data gaps given the associations observed for
other PFAS, such as PFBS, PFOA and PFOS CMDH. 2019. 2018: U.S. FPA. 2018b: MDH. 2017a: U.S.
EPA. 2016a. hi
This document is a draft for review purposes only and does not constitute Agency policy.
4-2 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Table 4-1. Evidence integration summary for health effects for which evidence
indicates a hazard exists
Evidence stream scenarios
Evidence in
studies of
humans3
Evidence in
animal
studies3
Evidence basis
O
°i_
ro
c
a»
(j
uo
E
ro
a»
s_
+-»
uo
cu
(j
c
Q)
~a
*>
0)
CU)
c
o
+-»
UO
V
Wo Studies, or Low
Confidence or
Conflicting
Evidence
Strong
Mechanistic
Evidence Alone
One High or
Medium
Confidence Apical
Study without
Supporting or
Conflicting
Evidence
Multiple High or
Medium
Confidence Apical
Studies with Some
Inconsistency or
Important
Uncertainties
Multiple High or
Medium
Confidence Apical
Studies with
Strong Support
(e.g., MOA
understanding
supporting
biological
plausibility)
Developmental
Hepatic
Thyroid
Developmental
Thyroid
Hepatic
Developmental
• No human studies
• Coherent observations of delays in
developmental milestones (eye opening, vaginal
opening, preputial separation) and fetal
mortality in one high confidence study of mice
exposed gestationally
• Consistent with findings for related PFAS
• No MOA information
• Human relevance presumed
Thyroid
• Single low confidence study in humans
• Consistent and biologically coherent results for
thyroid hormone levels (T4 without
compensatory changes in TSH), organ weights,
and histopathology from two high confidence
studies (short-term, subchronic) in male rats
• Consistent with findings for related PFAS
• No MOA information
• Human relevance presumed
Hepatic
• Two null studies (one medium and one low
confidence) with poor sensitivity
• Consistent, dose-dependent, and biologically
coherent effects on liver weights and
histopathology from seven high or medium
confidence studies in adult male rats and mice
(short-term and subchronic) and adult and
female mice exposed as adults or gestationally
• PPARa-dependence observed for some effects
(focal necrosis) but other effects (vacuolation)
occur in animals lacking PPARa activity (null
mice) or in animals with human PPARa
(humanized mice)
• Involvement of both PPARa-dependent and
independent mechanisms, including
hypertrophic responses in humanized PPARa
mice
• MOA information supports human relevance
aCan include consideration of studies informing biological plausibility: For studies in humans, this includes studies
of human tissues or cells, and other relevant simulations; for animal studies, this includes ex vivo and in vivo
experiments and other relevant simulations.
This document is a draft for review purposes only and does not constitute Agency policy.
4-3 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Toxicological Review ofPFBA and Related Salts
4.2. SUMMARY OF CONCLUSIONS FOR CARCINOGENICITY
No human or animal studies were available to inform the potential for PFBA exposure to
cause genotoxicity or cancer.
4.3. CONCLUSIONS REGARDING SUSCEPTIBLE POPULATIONS AND
LIFESTAGES
No human studies were available to inform the potential for PFBA exposure to affect
sensitive subpopulations or lifestages.
In adult animals exposed subchronically, PFBA exposure was consistently observed to elicit
stronger responses in male rats compared with female rats. The reason for this sex dependence is
most likely due to differences in toxicokinetics between males and females. The serum half-life of
PFBA following a single oral dose of 30 mg/kg-day is approximately 9 hours, compared to 2 hours
for females (see Table 3-1). Urinary excretion rates are much faster in female rodents compared to
male rodents (approximately 50-90% faster), possibly due to renal reabsorption ofPFBA in male
rats by organic anion transporters. Further, and specifically relevant to hepatic effects, the liver
concentrations ofPFBA following subchronic exposure to 30 mg/kg-day is approximately 16-fold
higher in males than in females [16.09 vs. 0.91 mg/kg-day; Butenhoff et al. (2012): van Otterdiik
f2007a. 2007b}]. No difference in serum half-lives was observed in monkeys exposed to a single i.v.
dose of 10 mg/kg: 1.61 hours for males vs. 2.28 hours in females fChang etal.. 20081. Also,
although quantitative data were not provided, serum excretion half-lives were reported not to
differ between males and females in the one occupational study available (Chang etal.. 2008).
Additionally, effects on liver weight were observed in pregnant and nonpregnant mice (Das etal..
2008). Developmental effects also were observed in female fetuses/neonates (full litter resorption,
delayed eye opening, delayed vaginal opening) and male fetuses/neonates [full litter resorption,
delayed eye opening, delayed preputial separation; Das etal. f20081]. with no clear difference in
sensitivity. Therefore, although there does appear to be a clear sex dependence for some PFBA-
induced health effects in adult rodents, the observed lack of sex-specific sensitivity for other effects
in adult and immature rodents and the apparent lack of toxicokinetic differences between sexes in
primates (and a single human occupational study) preclude the identification of males as a broadly
sensitive subpopulation for PFBA-induced health effects in humans.
Lastly, given the effects observed in pregnant mice (increased liver weights, full-litter
resorptions) and the developing organism (fetal/postnatal death and delays in time to eye opening,
vaginal opening, and preputial separation), that pregnancy and early life represent two sensitive
lifestages to PFBA exposure is possible.
This document is a draft for review purposes only and does not constitute Agency policy.
4-4 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Toxicological Review ofPFBA and Related Salts
5. DERIVATION OF TOXICITY VALUES
5.1. NONCANCER AND CANCER HEALTH EFFECT CATEGORIES
CONSIDERED
The available evidence indicates that oral exposure to PFBA is likely to cause adverse
thyroid, hepatic, and developmental effects in humans based on multiple high and medium
confidence animal toxicity studies (Butenhoffetal.. 2012: Foreman et al.. 2009: van Otterdiik.
2007a. b; Permadi etal.. 1993: Permadi et al.. 19921.
No human or animal toxicity studies are available to inform the potential for PFBA to cause
adverse effects via inhalation. Likewise, no human or animal studies are available to inform the
potential for oral or inhalation exposure to cause genotoxicity or cancer.
5.2. NONCANCER TOXICITY VALUES
The noncancer oral toxicity values (i.e., reference doses) derived in this section are
estimates of an exposure for a given duration to the human population (including susceptible
subgroups and lifestages) that is likely to be without an appreciable risk of adverse health effects
over a lifetime. The RfD derived in Section 5.2.1 corresponds to chronic, lifetime exposure and is
the primary focus of this document. In addition, RfDs specific to each organ or system are provided
(organ/system-specific RfDs), as these toxicity values might be useful in some contexts (e.g., when
assessing the potential cumulative effects of multiple chemical exposures occurring
simultaneously). Less-than-lifetime, subchronic toxicity values (including the subchronic RfD and
organ/system-specific subchronic RfDs), which are derived in Section 5.2.2, correspond to exposure
durations between 30 days and 10% of the life span in humans. These subchronic toxicity values
are presented because they might be useful for certain decision purposes (e.g., site-specific risk
assessments with less-than-lifetime exposures). Section 5.2.3 discusses that no information exists
to inform the potential toxicity of inhaled PFBA.
5.2.1. Oral Reference Dose (RfD) Derivation
Study Selection
Given the identified hazards relating to thyroid, liver, and developmental effects, two high
confidence studies reporting these effects were selected for the purpose of deriving an oral
reference dose (RfD). The subchronic (Butenhoffetal.. 2012) and developmental (Das etal.. 2008)
studies were selected to support RfD derivation given the ability of these study designs to estimate
potential effects of lifetime exposure, as compared to short-term or acute studies. Both studies
used rats or mice as the laboratory animal species and used vehicle-exposed controls. Animals
This document is a draft for review purposes only and does not constitute Agency policy.
5-1 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
were exposed to reagent-grade NFU+PFB (reported as >98% pure or as a 28.9% solution in distilled
water; impurities not reported) via a relevant route (oral administration via gavage) and for a
relevant duration (90 d or GD 1-17) of exposure.
Also available in the PFBA database are two short-term (i.e., 28-d) studies that provide
information on the hepatic and thyroid effects ofPFBA fButenhoff etal.. 2012: Foreman etal.. 2009:
van Otterdijk. 2007a). Although these studies were used for qualitative hazard identification
purposes (they supported the final evidence integration judgments for these endpoints and thus
were critical for identifying these endpoints for dose-response analysis), they ultimately were not
considered for use as the basis for the quantitative dose-response analyses. When developing a
lifetime reference value, chronic or subchronic studies (and studies of developmental exposure) are
generally preferred over short-term or acute studies. Likewise, subchronic and developmental
studies are preferred when developing a subchronic RfD. Although short-term studies were not
used for the identification of points of departure (PODs), however, they were deemed relevant to
decisions regarding the application of uncertainty factors for deriving toxicity values (see
"Derivation of Candidate Toxicity Values" below).
In the liver, a pattern of adverse effects has been observed in mice and rats, with PFBA
exposure resulting in increased liver weights (absolute and relative) in adult exposed animals
fButenhoff etal.. 2012: Das etal.. 2008: van Otterdiik. 2007bl in conjunction with histopathological
lesions [i.e., hepatocellular hypertrophy; Butenhoff et al. (2012): van Otterdijk (2007b)]. As
discussed in Section 3.2.2, the observed effects in the livers of exposed experimental animals are
judged relevant to human health as evidenced by the observation of increased liver weights and
increased hepatocellular hypertrophy in mice expressing human PPARa and increased vacuolation
in humanized-PPARa and PPARa null mice. This strongly suggests a multifaceted mode of action
(MOA) for liver effects consisting, in part, of non-PPARa mechanisms operant in humans (noting
that activation of human PPARa by PFBA also results in hepatic changes). Further, the observation
of vacuolation specifically indicates the observed effects are possible precursors to clearly adverse
downstream effects such as steatohepatitis, fibrosis, and cirrhosis. Thus, the observed pattern of
liver effects in PFBA-exposed animals are judged to be adverse, relevant to human health, and
appropriate to consider for reference value derivation. For the purposes of dose-response
modeling, relative liver weights were chosen over absolute liver weights. Although body weights
were not affected on average in any PFBA study, relative liver weights are still preferred because
this measure of effect accounts for any changes in body weights that occur in individual animals
(changes in body and liver weights are associated). For liver hypertrophy, severity information in
addition to raw incidence was available. Therefore, both total incidence of lesions and incidence of
"slight" severity lesions were considered for dose-response analysis.
A pattern of adverse effects in the thyroid also is observed in exposed rats that consists of
decreased free and total T4 levels and increased incidence of thyroid follicular hypertrophy and
hyperplasia fButenhoff etal.. 2012: van Otterdiik. 2007bl. Decreased thyroid hormone levels are
This document is a draft for review purposes only and does not constitute Agency policy.
5-2 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
judged relevant to human health, given the many similarities in the production, regulation, and
functioning of thyroid hormones between rodents and humans. For effects on T4, total T4 was
chosen for dose-response modeling over free T4, on the basis of lack of data in the control group for
free T4 (given insufficient volume for the assay). In addition, rodents are more sensitive to
increases in thyroid follicular hypertrophy and hyperplasia, and thus changes in thyroid hormone
levels are considered more relevant for deriving human health toxicity values. For this reason, the
increases in thyroid hypertrophy/hyperplasia were not considered further for RfD derivation.
Note, however, that decreased total T4 was observed at 6 mg/kg-day in rats exposed to PFBA for
28 days, but not in rats exposed for 90 days (where it was observed only at 30 mg/kg-day). This
discrepancy can be explained, however, by the difference in serum concentrations following
28- and 90-day exposures. Serum free T4 concentrations were higher in the 6 mg/kg-day dose
group following 28-day exposures (24.7 ng/mL) vs. 90-day exposures (6.1 ng/mL). This difference
was reversed in the 30 mg/kg-day dose group for the 28-day and 90-day animals, being 38.0 ng/mL
vs. 52.2 |ig/mL, respectively. Because serum concentrations following chronic exposures likely will
resemble those following subchronic exposures (more so than serum concentrations following
short-term exposures), the effects on total T4 following subchronic exposure are deemed most
appropriate for deriving lifetime and subchronic toxicity values.
Effects on the developing reproductive system included delays in vaginal opening and
preputial separation (Das etal.. 2008). EPA's Reproductive Toxicity Guidelines (U.S. EPA. 1996)
states that "[significant effects on ... age at puberty, either early or delayed, should be considered
adverse..." and thus supports considering these endpoints for reference value derivation. Delayed
eye opening, also found following PFOA exposure, is identified as a "simple, but reliable" indicator
of impaired postnatal development by Das etal. f20081. Further, a delay of eye opening is a form of
visual deprivation that prevents ocular visual signals from reaching the brain during a critical
period of development fWiesel. 19821. A time-sensitive critical period in the development of the
visual system is when the architecture of the visual cortex is established (Espinosa and Strvker.
2012). and accordingly, any alterations of the visual system during that time is considered adverse.
Evidence in humans further supports the adversity of this endpoint, given that infants born with
congenital cataracts that interfere with the processing of visual signals have permanent visual
defects if the cataracts are removed after the critical window for visual development fWiesel.
19821. Therefore, any delay in the development of sight or development of the visual neurological
system results in permanent functional decrements and is relevant to human health.
Full-litter resorption (FLR), a clear indicator of postimplantation embryo/fetal mortality,
was increased twofold and fourfold in pregnant mice exposed to 175 mg/kg-day or 350 mg/kg-day
(respectively) during pregnancy. In the uteri of dams without full resorptions, there was additional
evidence of fetal resorptions. In addition, in a separate cohort of gestationally exposed dams that
were allowed to deliver litters and were killed after their pups were weaned on lactation day 22,
there was an indication of decreased pre- and postnatal survival of the offspring (as determined by
This document is a draft for review purposes only and does not constitute Agency policy.
5-3 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Toxicological Review ofPFBA and Related Salts
a comparison of the number of maternal implantation sites to the number of pups delivered), the
magnitude of which is considered biologically significant (discussed below). Taken together, the
potential coherence of decreased pre- or postnatal survival with other effects on early fetal
mortality and developmental maturation (i.e., delays in eye opening and pubertal milestones)
supports consideration of all these developmental endpoints for deriving PODs.
Individual animal data were obtained from the study authors, which allowed for a thorough
consideration of pre- and postnatal mortality data. When the FLR data were combined with data
for prenatal mortality from litters without FLR to provide a more complete assessment of
embryo/fetal mortality, the response was statistically significant (p = 0.012) using the Cochran-
Armitage trend test with a Rao-Scott adjustment (CA/RS) method fRao and Scott. 19921. Although
the embryo/fetal mortality observed as FLR is presumed to have occurred much earlier in
pregnancy than fetal mortality in non-FLR litters and could involve different or overlapping
contributing mechanisms, combining these endpoints provides information on pregnancy loss and
fetal mortality over the entire gestational period, corresponding to the period ofPFBA exposure.
This was deemed more appropriate than modeling FLR and non-FLR fetal mortality separately.
Combining the data in this way has the added benefit of allowing the data to be modeled with the
nested dichotomous models and avoids the lower resolution of modeling the FLR data as dam
incidence per dose group.
The individual litter data obtained from the study authors also allowed for consideration of
modeling postnatal mortality (i.e., number of neonatal deaths compared to the number of
implantation sites). Analysis of the individual litter data revealed a nonmonotonic dose-response
for postnatal mortality, with response rates of 0.38%, 1.04%, 2.93%, and 1.2% at 0, 35,175, and
350 mg/kg-day, respectively, and the CA/RS trend test for the dataset was not statistically
significant (p = 0.09). Further, the data for postnatal mortality clearly indicates it is a weaker
response compared to prenatal mortality. Given that postnatal mortality was a weaker response
than prenatal morality, it failed to achieve statistical significance, and prenatal mortality is more
closely aligned with the period of exposure, postnatal mortality was not considered further for POD
derivation.
The studies (excluding the short-term studies) and outcomes relevant to the identified
hazards were selected and advanced for POD derivation as presented in Table 5-1. These selected
datasets were evaluated for toxicity value derivation as described below and in Appendix D.
This document is a draft for review purposes only and does not constitute Agency policy.
5-4 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Table 5-1. Endpoints considered for dose-response modeling and derivation
of points of departure
Endpoint
Reference3
Exposure duration
Species, sex
POD derivation13
Liver
Increased relative liver weight
Butenhoff et
al. (2012)
Subchronic
S-D rat, male
Yes
Gestational
CD-I mouse, female
Yes
Increased absolute liver weight
Subchronic
S-D rat, male
No
Gestational
CD-I mouse, female
No
Increased liver hypertrophy
Subchronic
S-D rat, male
Yes
Thyroid
Decreased total T4
Butenhoff et
al. (2012)
Subchronic
S-D rat, male
Yes
Decreased free T4
Subchronic
S-D rat, male
No
Increased thyroid follicular
hypertrophy
Subchronic
S-D rat, male
No
Developmental
Embryo/fetal mortality
Das et al.
(2008)
Gestational
CD-I mouse, male
and female
Yes
Postnatal mortality
Gestational
CD-I mouse, male
and female
No
Delayed eye opening
Gestational
CD-I mouse, male
and female
Yes
Delayed vaginal opening
Gestational
CD-I mouse, female
Yes
Delayed preputial separation
Gestational
CD-I mouse, male
Yes
aBoth the Butenhoff et al. (2012) and Das et al. (2008) studies were rated as high confidence.
bSee text for rationale for inclusion/exclusion from POD derivation.
Estimation or Selection of Points of Departure (PODs)
1 Consistent with EPA's Benchmark Dose Technical Guidance fU.S. EPA. 20121. the BMD and
2 95% lower confidence limit on the BMD (BMDL) were estimated using a BMR to represent a
3 minimal, biologically significant level of change. The BMD technical guidance (U.S. EPA. 2012) sets
4 up a hierarchy by which BMRs are selected, with the first and preferred approach using a biological
5 or toxicological basis to define what minimal level of response or change is biologically significant.
6 If that biological or toxicological information is lacking, the BMD technical guidance recommends
7 BMRs that can be used instead, specifically a BMR of 1 standard deviation (SD) from the control
8 mean for continuous data or a BMR of 10% extra risk (ER) for dichotomous data. The BMRs
9 selected for dose-response modeling of PFBA-induced health effects are listed in Table 5-2 along
10 with the rationale for their selection.
This document is a draft for review purposes only and does not constitute Agency policy.
5-5 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Table 5-2. Benchmark response levels selected for benchmark dose (BMD)
modeling of perfluorobutanoic acid (PFBA) health outcomes
Endpoint
BMR
Rationale
Liver
Increased relative
liver weight
10%
relative
deviation
A 10% increase in liver weight has generally been considered a minimally biologically
significant response.
Increased liver
hypertrophy
10% extra
risk
A 10% extra risk is a commonlv used BMR for dichotomous endpoints (U.S. EPA,
2012) in the absence of information for a biologically based BMR; the endpoint is not
considered a frank effect and does not support using a lower BMR.
Thyroid
Decreased total T4
1 standard
deviation
Toxicological evidence that would support identification of a minimally biologically
significant response is lacking in adult animals. Further, evidence for the level of
response in thyroid hormones associated with neurodevelopmental effects is
inconsistent, with decreases of 10-25% identified in human and rodent studies
(Gilbert et al., 2016; Gilbert, 2011; Haddow et al., 1999). The BMD technical
guidance (U.S. EPA, 2012) recommends a BMR equal to 1 standard deviation for
continuous endpoints when biological information is not sufficient to identify the
BMR. In this case, the BMR based on 1SD from the Butenhoff et al. (2012) studv
corresponds to a ~13% decrease, consistent with the levels of decreased T4
associated with neurodevelopmental decrements, thus strengthening the rationale
for using a BMR = 1 SD for this endpoint.
Developmental
Embryo/fetal
morality
1% extra
risk
For quantal endpoints, the BMG Technical Guidance states "[f]rom a statistical
standpoint, most reproductive and developmental studies with nested study designs
support a BMR of 5%" and "[b]iological considerations may warrant the use of a
BMR of 5% or lower for some types of effects (e.g., frank effects)...". As increased
treatment-related embryo/fetal mortality is clearly a frank effect, BMRs of 5% and
1% were considered. Given that the study employed a nested design with individual
animal data available that allow the use of the nested dichotomous models (to
account for intra-litter similarity), and the effect of interest was a frank effect
(supporting a BMR 5% or lower), a BMR of 1% extra risk was ultimately selected for
derivation of the POD to account for the biological severity of these endpoints (i.e.,
mortality) and the robust statistical power of the study.
Delayed eye
opening
5% relative
deviations
Biological evidence supports identification of a minimally significant decrease of
visual input (1-d delayed eye opening) due to hypothyroxinemia during a critical
period of retinal development (Espinosa and Stryker, 2012). Delays of 1 d in eye
opening reduces the time available for visual cortex development related to
orientation selectivity bv approximately 20% (Espinosa and Strvker, 2012) and
corresponds to ~6% change in Das et al. (2008). Further, delavs in vaginal opening
greater than or equal to 2 d have been used previously to define biologically relevant
responses (U.S. EPA, 2013), and this magnitude in delav in Das et al. (2008) is also
~6%. Both levels of response are consistent with a 5% relative deviation. Lastly, a
5% change in other markers of growth/development in gestational studies (e.g., fetal
weight) has generally been considered a minimally biologically significant response
level.
Delayed vaginal
opening
Delayed preputial
separation
This document is a draft for review purposes only and does not constitute Agency policy.
5-6 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Toxicological Review ofPFBA and Related Salts
When modeling was feasible, the estimated BMDLs were used as points of departure (PODs,
see Table 5-4). Further details, including the modeling output and graphical results for the model
selected for each endpoint, can be found in Appendix D. When dose-response modeling was not
feasible, or adequate modeling results were not obtained, NOAEL or LOAEL values were identified
based on biological rationales when possible and used as the POD. For example, for liver weight, a
NOAEL would be chosen as the dose below which causes at least a 10% change, consistent with the
rationale for the selecting the BMR for that endpoint. If no biological rationale for selecting the
NOAEL/LOAEL is available, statistical significance was used as the basis for selection. The PODs
(based on BMD modeling or NOAEL/LOAEL selection) for the endpoints advanced for dose-
response analysis are presented in Table 5-4.
Approach for Animal-Human Extrapolation of Perfluorobutanoic Acid (PFBA) Dosimetry
The PFAS protocol (Appendix A) recommends the use of physiologically based
pharmacokinetic (PBPK) models as the preferred approach for dosimetry extrapolation from
animals to humans, while allowing for the consideration of data-informed extrapolations (such as
the ratio of serum clearance values) for PFAS that lack a scientifically sound and sufficiently
validated PBPK model. If chemical-specific information is not available, the protocol then
recommends that doses be scaled alio metrically using body weight (BW)3/4 methods. This
hierarchy of recommended approaches for cross-species dosimetry extrapolation is consistent with
EPA's guidance on using allometric scaling for deriving oral reference doses (U.S. EPA. 2011). This
hierarchy preferentially prioritizes adjustments that result in reduced uncertainty in the dosimetric
adjustments (i.e., preferring chemical-specific values to underpin adjustments vs. use of default
approaches).
No PBPK model is available for PFBA. But, as toxicokinetic data for PFBA exist in relevant
animals (rats, mice, and monkeys) and humans, a data-informed extrapolation approach for
estimating the dosimetric adjustment factor (DAF) can be used. Briefly, the ratio of the clearance
(CL) in humans to animals, CLh:CLa, can be used to convert an oral dose rate in animals
(mg/kg-day) to a human equivalent dose rate. Assuming the exposure being evaluated is low
enough to be in the linear (or first-order) range of clearance, the average blood concentration (CAvg)
that results from a given dose is calculated as:
Cavg (mg/mL) = W°*K"gAe/h) [5.1}
AVO v &/ J CL (mL/kg/h) 1 J
where/abs is the fraction absorbed and dose is average dose rate expressed at an hourly rate.
Assuming equal toxicity given equal Cavg in humans as mice or rats, and that/abS is the same in
humans as animals, the equitoxic dose (i.e., the human dose that should yield the same blood
concentration [Cavg] as the animal dose from which it is being extrapolated) is then calculated as
follows:
This document is a draft for review purposes only and does not constitute Agency policy.
5-7 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Toxicological Review ofPFBA and Related Salts
HED =
POD
(5-2)
CXa/CXh
Thus, the DAF is simply CLh:CLa, the ratio of clearance in humans to clearance in the animal
from which the POD is obtained. Note that although this evaluation of relative internal dose (Cavg)
assumes that internal dose increases linearly with exposure (as does default allometric scaling),
nonlinearity is usually observed only at relative high exposure levels. Further, although clearance of
PFBA could be biphasic, it is still linear: A two-compartment classical PK model still uses all linear
rate equations, and the predicted Cavg from a two-compartment model still increases linearly with
exposure or applied dose.
Clearance values, however, are not reported for humans in the one toxicokinetic study
available for PFBA fChang etal.. 20081. As clearance is a measure of average excretion, to calculate
it, one also needs to evaluate a companion variable, the volume of distribution (Vd), which in turn
requires a measure of total exposure or dose. Chang etal. (2008) did not report the Vd for humans.
Chang etal. (2008) did report Vd for cynomolgus monkeys, however, and as summarized above in
Section 3.1.5, the data suggest a difference in Vd between rodents and monkeys. For comparison, the
Vd values for PFOA and PFOS estimated from the PBPK parameters of (Loccisano etal.. 2011) are
approximately 0.2 and 0.3 L/kg, respectively, although that obtained from monkeys for PFBA is
approximately 0.5 L/kg. This value of Vd for PFBA was obtained from standard analysis of the
empirical PK data, which is not influenced by any preliminary chemical-specific assumptions, but as
stated by the authors, "Volume of distribution estimates indicated primarily extracellular
distribution" (Chang etal.. 2008). The difference between Vd for PFBA and those for PFOA and PFOS
indicates slightly more intracellular distribution by PFBA. As described in Section 3.1.2
Distribution, Vd for humans is expected to be similar to the value for monkeys, thus the average
value for male and female monkeys from Chang etal. f20081 will be used. Human clearance,
normalized to body weight, can be calculated as follows:
Note that in equation (5-3), BW normalization is embedded in the fact that Vd is a volume per kg
BW. For example, the average blood concentration, Cavg (mg/mL), can then be estimated using
equation (5-1) for any given dose (mg/kg/h = (mg/kg/d)/(24 h/d)), independent of specific BW.
As ti/2 is required in the calculation of CL, these values must be determined from the data
presented for humans in Chang etal. (2008). Chang etal. (2008) reported values for human
subjects from two 3M facilities: Cottage Grove, Minnesota and Cordova, Illinois. Cottage Grove had
three subjects, which were not identified by gender. Cordova had nine subjects, two of which were
identified as female. The half-lives for those two women fell among the values of the other subjects
(Cottage Grove and men from Cordova). Considering the minimal difference in ti/2 observed
CLfiuman (mL/kg-h) = ln(2) x
l
^ Kl,monkey
(mL/kg) (5-3)
ti/2,human(h)
This document is a draft for review purposes only and does not constitute Agency policy.
5-8 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Toxicological Review ofPFBA and Related Salts
between male and female monkeys, the available data were assumed insufficient to distinguish
male and female humans. The analytic method used replaced concentration measurements below
the lower limit of quantitation (LLOQ) with LL0Q/V2. For individuals where only two
measurements were made, the resulting half-life estimate was then highly sensitive to this
assumption. The two known female subjects (Cordova), one male subject from Cordova, and one
subject from Cottage Grove fell into this category; half-lives for these four subjects were not used.
Additionally, the last time point for Subject 2 from Cottage Grove was below the LLOQ and was also
excluded from ti/2 estimation. The mean and median ti/2 values estimated from these data (8 total
subjects, 20 observations) were 81.8 and 67.5 hours, respectively. Mixed-effects modeling
confirmed this half-life, estimating an approximate half-life of 67.9 hours when accounting for
clustering (see Appendix C). Other details of the human half-life data are described in Section 3.1.4,
Excretion.
As discussed in Section 3.1.4, using the common assumption of BW°75 scaling of clearance
and standard species BWs of 0.25 kg in rats and 80 kg in humans, the half-life in humans would be
predicted to be 4.2 times greater than rats. Given half-lives of 9.22 and 1.76 hours in male and
female rats, one would then predict half-lives of 38.7 hours in men and 7.4 hours in women.
Although the value for men is in the range of results for humans, the value for women is much less
than that estimated using the human data available from Chang etal. f20081. DAFs based on BW°75
scaling for rats and a standard BW of 0.03 kg for mice are presented in Table 5-3. EPA's guidance on
use of BW°75 as the default method for derivation of an oral reference dose states, however, "EPA
endorses a hierarchy of approaches to derive human equivalent oral exposures from data from
laboratory animal species." It goes on to state that, although use of PBPK models is preferred,
"Other approaches may include using chemical-specific information, without a complete
physiologically-based toxicokinetic model" (i.e., the approach described here, using relative
clearance) and that use of BW0 75 is endorsed, "In lieu of data to support either of these types of
approaches" fU.S. EPA. 20111. Thus, because data are available to support a chemical-specific
approach, it is clearly preferred.
Using a value of 484.5 mL/kg for Vd for humans [average of male and female Vd values in
monkeys, 526 and 443 mL/kg, respectively, Table 4, Chang etal. (2008)] and 67.9 hours for ti/2 in
male humans, CL in humans is estimated to be 4.95 mL/kg-h. See Table 5-3 for the DAFs for
converting rat and mice PODs to human equivalent doses (HEDs).
This document is a draft for review purposes only and does not constitute Agency policy.
5-9 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Toxicological Review ofPFBA and Related Salts
Table 5-3. Rat, mouse, and human clearance values and data-informed
dosimetric adjustment factors
Sex
Species
Animal CL (mL/kg-h)
Human CL (mL/kg-h)
DAF (CIH:CIA)
DAF (BW° 75)d
Male
Rat
21.61a
4.95°
0.229
0.236
Mouse
10.10b
0.490
0.139
Female
Rat
96.62a
0.051
0.236
Mouse
27.93b
0.177
0.139
Data from Tables 2, 3, 5, and 6 of Chang et al. (2008).
aAverage of CL = dose/AUC (area-under-the-concentration-curve) was calculated using values reported for oral and
i.v. exposures reported in Table 2 of Chang et al. (2008); see Table 3-2.
bAverage of CL = dose/AUC was calculated using values reported for the 10- and 30-mg/kg dose groups reported in
Table 3 of Chang et al. (2008); see Table 3-2. CL for the 100-mg/kg dose group was excluded, as it was "threefold
and "twofold higher for males and females, respectively, than the values reported at 10 or 30 mg/kg. This could
be due to saturation of renal absorption or serum binding.
CCL value for humans (male and female) as described above.
dDAFs based on assumption that elimination scales as BW075, hence clearance (elimination/BW) scales as BW"0 25,
using standard BWs of 0.03, 0.25, and 80 kg for mice, rats, and humans, respectively.
Therefore, human equivalent dose (HED) for considered health effects was calculated as
follows, using relative liver weight observed in male rats in the subchronic Butenhoff et al. f20121
study as an example. Note that the concentration of the ammonium salt first needs to be converted
to the concentration of the free acid before HED calculation:
HED PFBA = POD NHf PFB (mg/kg-d) MWPFBA x CLhuman^mL/k9-h') (5_4)
MWNH^ PFB CL animal (mL/kg-h)
. 214 g/mol 4.95 (mL/kg-h)
HED = 9.6 (mg/kg-d) X ^ ^ X n 61 (mL/kg.h) = 2.04 (mg/kg-d)
Uncertainty of Animal-to-Human Extrapolation ofPFBA Dosimetry
There is uncertainty in applying this dosimetric approach given the volume of distribution
(Vd) was not measured in humans and the human Vd was assumed equal to that in monkeys to
estimate clearance in humans. An alternative approach to using the ratio of clearance values for
animal:human dosimetric adjustments is to use the measured serum concentrations from
toxicological studies as BMD modeling inputs and then use the estimated human clearance values to
calculate the HED. This approach, compared to the ratio of the clearance values approach, however,
is interpreted to have even greater uncertainty. First, the measured serum concentrations were
reported to have been taken 24 hours after the last exposure in the developmental toxicity study
fDas etal.. 20081 and likely were similarly taken in the subchronic toxicity study fButenhoffetal..
2012: van Otterdijk. 2007b). Given the relatively short half-life of PFBA measured in mice and rats,
this end-of-exposure measurement of serum concentrations likely did not reflect the average serum
This document is a draft for review purposes only and does not constitute Agency policy.
5-10 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
concentrations exposed animals experienced. For example, the reported serum levels (see
Section 2.1.1) in female mice in the Das etal. f20081 study did not correlate with exposure levels.
Also, to estimate the HED without a validated PBPK model, the resulting POD (in units of serum
concentrations) would need to be multiplied by the estimated human clearance value. Thus, in
addition to the uncertainty in using end-of-exposure serum concentrations not reflective of average
exposures, this approach would be characterized by the same uncertainty as the assumption that
human and monkey volumes of distribution are equal and the uncertainty in the human half-life.
Therefore, the ratio of clearance values is considered to have less uncertainty than either serum
concentration-based BMD modeling or use of default allometric dosimetric adjustments. Thus, the
approach based on clearance values is the one used here.
That only a single study reported PFBA PK data in rats or mice (or monkeys) introduces
qualitative uncertainty, because these results were not validated in independent experiments.
Results from different studies cannot be compared quantitatively. In the Chang etal. (2008) study,
some results have relatively tight standard errors (SEs), indicating high confidence, but others
(especially for mice), indicate high variability/uncertainty. Although the results for AUC in rats have
relatively small SEs, they surprisingly show higher AUC (hence lower clearance) following oral
doses than following i.v. doses (30 mg/kg). Oral absorption or bioavailability can range between
near zero and 100%, but why the blood concentrations after an oral dose are higher than when the
same dose is injected directly into the blood is puzzling. The data and plot of the PK model shown in
Figures 1 and 2 of Chang etal. (2008) indicate the absorption and clearance phases are well
characterized and described by the model, so the uncertainty does not appear to be due to the study
design or analysis method. The almost twofold difference in clearance rates estimated from the oral
vs. i.v. rat data thus indicate a comparable degree of uncertainty.
Compared to the results for rats, the Chang etal. f20081 clearance estimates at the two
lower oral doses in male and female mice are much closer, with only an 8% difference between the
two doses for males and a 16% difference for females. The results for both male and female mice
show a dose-dependent increase in clearance across all dose levels, consistent with the hypothesis
of saturable renal resorption. Although the increase only seems significant with the increase from
30 to 100 mg/kg, the differences between 10 and 30 mg/kg could result from the same mechanism.
Thus, those differences might reflect a biological mechanism as much as experimental or analytic
variability. The lack of i.v. data in mice at the same dose as any of the oral doses, however, means
that one cannot fully compare the apparent self-consistency of the mouse data to the inconsistency
noted above for rats.
If the oral vs. i.v. discrepancy in rats is interpreted as indicating an overall factor of 2
uncertainty in the animal clearance values, that can be considered a moderate degree of
uncertainty. As a rule-of-thumb, PBPK models are expected to match the corresponding data within
a factor of 2, a similar level of uncertainty. Although the human half-life estimates vary just over
fivefold from highest to lowest, this much variability in a human population is not surprising, and
This document is a draft for review purposes only and does not constitute Agency policy.
5-11 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
1 with results from just 12 subjects to characterize the mean, uncertainty in that mean can, again, be
2 considered moderate. Given that the physiological fractions of different tissue types is similar in
3 humans and primates and that the blood serum:tissue portioning is reasonably expected to be
4 similar across mammals, the assumption that the volume of distribution in humans is similar to
5 monkeys is considered to have low uncertainty. Considering all these factors, the overall
6 uncertainty in HED calculations using equation (5-4) with the parameters estimated here is
7 considered moderate, that is, within a factor of 3.
Application of Animal-Human Extrapolation ofPFBA Dosimetry
8 Table 5-4 presents the PODs and estimated PODhed values for the thyroid, liver, and
9 developmental toxicity endpoints.
Table 5-4. Points of departure (PODs) considered for use in deriving
candidate reference values for perfluorobutanoic acid (PFBA)
Endpoint/reference
Species/strain
/sex
POD
type/model
POD NH4+ PFB
(mg/kg-d)
POD PFBA
(mg/kg-
d)a
PODhed
PFBAb
(mg/kg-d)
Increased relative liver weight
(Butenhoff et al., 2012)
S-D rat, male
BMDLiord
Exp3 (LN-CV)
9.6
8.89
2.04
Increased relative liver weight
(Das et al., 2008)
CD-I mouse, P0
female
BMDLiord
Exp4 (CV)
15
13.9
2.46
Increased liver hypertrophy15
(Butenhoff et al., 2012)
S-D rat, male
BMDLioer
Weibull
5.4
5.0
1.15
Decreased total T4
(Butenhoff et al., 2012)
S-D rat, male
NOAEL0
(15% decrease)
6
5.56
1.27
Embryo/fetal mortality
(Das et al., 2008)d
CD-I mouse, Fi
male/female
BMDLier
Nested-Logistic
5.7
5.28
0.93
Delayed eyes openingd
(Das et al., 2008)
CD-I mouse, Fi
male/female
BMDL5RD
Hill (CV)
4.9
4.54
0.80
Delayed vaginal openingd
(Das et al., 2008)
CD-I mouse, Fi
female
BMDL5RD
Hill (CV)
3.8
3.52
0.62
Delayed preputial separationd
(Das et al., 2008)
CD-I mouse, Fi
male
BMDL5RD
Exp3 (CV)
179.1
165.92
29.37
BMDL = 95% lower limit on benchmark dose, RD = relative deviation, LN = log-normal, CV = constant variance,
ER = extra risk, NOAEL = no-observed-adverse-effect level.
a Both of these studies used the ammonium salt of PFBA as the test article. To calculate a POD for the free acid of
PFBA from any PFBA salt, multiply the POD of interest by the ratio of molecular weights of the salt and the free
acid. For example, to convert from the ammonium salt of PFBA to the free acid, multiply the ammonium salt POD
by 0.926: —MWfreeacid— = 111= 092e.
MW ammonium salt 231
See discussion in Section 5.2.1, Approach for Animal-Human Extrapolation of PFBA Dosimetry, for details on HED.
bModeling results for all lesions are used here given greater model uncertainty when modeling only "slight" lesions
(see Appendix D).
This document is a draft for review purposes only and does not constitute Agency policy.
5-12 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
cNo models provided adequate fit to the mean when using constant or nonconstant variance with the normal
distribution or constant variance with the log-normal distribution.
dAII HED calculations used DAF for female mice, given exposures were to pregnant animals.
Derivation of Candidate Toxicity Values for the Oral Reference Dose (RfDJ
1 Under EPA's A Review of the Reference Dose and Reference Concentration Processes (U.S. EPA.
2 2002)and Methods for Derivation of Inhalation Reference Concentrations and Application of
3 Inhalation Dosimetry U.S. EPA (19941. five possible areas of uncertainty and variability were
4 considered in deriving the candidate values for PFBA. An explanation of these five possible areas of
5 uncertainty and variability and the values assigned to each as designated UFs to be applied to the
6 candidate PODhed values are listed in Table 5-5. As discussed below, the short-term studies of
7 thyroid and hepatic effects after PFBA exposure were considered for use in UF selection.
Table 5-5. Uncertainty factors for the development of the candidate values for
perfluorobutanoic acid (PFBA)
UF
Value
Justification
UFa
3
A UFa of 3 (10°5 = 3.16 ~3) is applied to account for uncertainty in characterizing the toxicokinetic
and toxicodynamic differences between mice or rats and humans following oral NH4+PFB/PFBA
exposure. Some aspects of the cross-species extrapolation of toxicokinetic processes have been
accounted for by calculating an HED through application of a DAF based on animal and human
half-lives; however, some residual toxicokinetic uncertainty and uncertainty regarding
toxicodynamics remains. Available chemical-specific data further support the selection of a UF of
3 for PFBA; see text below for further discussion.
UFh
10
A UFH of 10 is applied for interindividual variability in the absence of quantitative information on
the toxicokinetics and toxicodynamics of NH4+PFB/PFBA in humans.
UFS
10
A UFs of 10 is applied to endpoints observed in the subchronic studv (Butenhoff et al., 2012; van
Otterdiik, 2007b) for the purposes of deriving chronic toxicity values. See additional discussion on
this decision below.
1
A UFs of 1 is applied to endpoints observed in the developmental toxicity studv (Das et al., 2008);
the developmental period is recognized as a susceptible lifestage where exposure during certain
time windows (e.g., pregnancy and gestation) is more relevant to the induction of developmental
effects than lifetime exposure (U.S. EPA, 1991).
ufl
1
A UFl of 1 is applied for LOAEL-to-NOAEL extrapolation when the POD is a BMDL or NOAEL
UFd
3
A UFd of 3 is applied because, although the PFBA database is relatively small, high confidence
subchronic and developmental toxicity studies are available in mice and rats. Although these
high confidence studies are available for PFBA, the database has some deficiencies, including the
lack of information on developmental neurotoxicity and other endpoints; see the text below for
further discussion.
UFC
Table
5-7
Composite uncertainty factor = UFA x UFH x UFS x UFL x UFD.
8 As described in EPA's A Review of the Reference Dose and Reference Concentration Processes
9 (U.S. EPA. 20021. the interspecies uncertainty factor (UFA) is applied to account for extrapolation of
This document is a draft for review purposes only and does not constitute Agency policy.
5-13 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
animal data to humans; it accounts for uncertainty regarding the toxicokinetic and toxicodynamic
differences across species. As is usual in the application of this uncertainty factor, the toxicokinetic
uncertainty is mostly addressed through the application of dosimetric approaches for estimating
human equivalent doses (see Section 4.2.2). This leaves some residual uncertainty around the
toxicokinetics and the uncertainty surrounding toxicodynamics. Typically, a threefold UF is applied
for this uncertainty in the absence of chemical-specific information. This is the case for the thyroid
and developmental endpoints. For the liver endpoints, chemical-specific information should be
considered further in determining the most appropriate value for the UFAto account for the
uncertainty.
Foreman et al. f20091 investigated the response to PFBA exposure in PPARa wild-type,
PPARa null, and hPPARa mice for hepatic effects and observed either that effects were generally
equivalent in wild-type vs. humanized mice (liver weight, liver hypertrophy, see Table 3-6 and
Table 3-7), that wild-type mice exhibited effects that humanized mice did not (focal hepatic
necrosis), and that PPARa null mice generally did not exhibit hepatic effects. Additionally, in vitro
studies suggest that human cells or cells transfected with human PPARa were less sensitive to
PPAR activation than rodent cells or rodent PPARa fRosen etal.. 2013: Wolfetal.. 2012: Biork and
Wallace. 2009: Wolf etal.. 20081. If PPARa were the only operant MOA for noncancer effects in the
liver, this observation might support reducing the remaining portion of the UFa to 1, as it could be
argued that humans are not as sensitive as wild-type rats to the hepatic effects of PFBA exposure
(note: without evidence to the contrary, as mentioned in the previous paragraph, the toxicodynamic
portion of this UF is typically assigned a value of 3 assuming responses manifest in humans could
be more sensitive than those observed in animals). Additional evidence presented in Foreman et al.
f20091 and other studies (see Section 2.2.5), however, indicates that non-PPARa MOAs appear to be
active in the livers of exposed rats. Specifically from Foreman et al. f20091. vacuolation is reported
in the livers of PPARa null and humanized mice, but not in wild-type mice, although the degree to
which null or humanized mice are more susceptible to this effect is difficult to characterize given
the results are presented qualitatively. Vacuolation (i.e., the accumulation of lipids) is an important
precursor event in the development of steatosis, which itself is a precursor to other adverse
conditions such as steatohepatitis, fibrosis, and cirrhosis. As discussed in Section 2.2.5, this
observation of PFBA-induced effects independent of PPARa activation is supported by in vitro and
in vivo data that show other PFAS can activate other forms of PPAR (i.e., PPARy) and additional
pathways (i.e., constitutive androstane receptor [CAR] or pregnane X receptor [PXR]). Given the
observation of apical liver effects in humanized PPARa mice and the observation that other MOAs
appear to contribute to potential liver toxicity, the observation that humanized PPARa mice exhibit
diminished responses for some hepatic effects attributable to PPARa activation cannot alone
determine the appropriate value of the toxicodynamic portion of the UFA. Therefore, given the
remaining uncertainty in additional MOAs that appear active in PFBA-induced liver effects, and the
relative contribution of these MOAs to toxicity in humans as compared with rodents, the value of
This document is a draft for review purposes only and does not constitute Agency policy.
5-14 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Toxicological Review ofPFBA and Related Salts
UFa was set to 3 for the purposes of deriving toxicity values for hepatic effects. No MOA information
is available for thyroid or developmental effects; in the absence of information suggesting
otherwise, as noted above, a UFa (3) is also applied to these endpoints to account for any residual
toxicokinetic and toxicodynamic uncertainty.
The short-term studies of Butenhoff et al. f20121. van Otterdiik f2007al. and Foreman etal.
(2009) also were considered for potential use in informing the selection of the UFs. More
specifically, for several outcomes from which PODs were derived, comparisons between short-term
exposure and subchronic exposure appeared possible (i.e., because of the inherent similarities in
study design and experimental conduct). When comparing short-term to subchronic PFBA
exposure for liver weight and thyroid hormone measures, there was no apparent increased
sensitivity with longer exposure duration in terms of the magnitude of the observed effects at the
same tested doses or the lowest doses at which effects were observed. In addition, given the
toxicokinetics of PFBA, steady-state levels in potential target tissues might not substantially
increase with increasing exposure duration (Butenhoff etal.. 2012: van Otterdiik. 2007a. b). In
these studies, the latter conclusion seemed dose dependent, as PFBA levels actually decreased with
longer exposures when comparisons are made at 6 mg/kg-day (~25 to 14 |J.g/mL in serum and
~7.5 to 3.1 |ig/g in liver comparing 28 to 90 days of exposure), whereas levels were either
increased slightly or were similar when comparisons are made at 30 mg/kg-day (~38 to 52 |J.g/mL
in serum and ~17.4 to 16.1 |J.g/mL in liver comparing 28 to 90 days of exposure). This indicates
perhaps that steady-state conditions have been reached in the livers of exposed rats after only
28 days of exposure. Initially, this indicates that increased durations of exposure might not elicit
increased effects in the target tissue, as the LOAEL for liver weights is 30 mg/kg-day for male rats
exposed to either 28 or 90 days. When also considering results from Foreman etal. f20091. and
basing comparisons on human equivalent external concentrations (see Table 5-6 below for
modeling results and application of dosimetric adjustments), liver weight appears affected at
equivalent doses across mice and rats and durations of exposure in the available studies.
Table 5-6. Comparison of liver-weight effects across species and durations of
exposure
Reference
Species/strain/
sex
Duration
POD
type/model
POD NH4+ PFB
(mg/kg-d)
POD PFBA
(mg/kg-d)
PODhed PFBA
(mg/kg-d)
Relative liver weight
(Butenhoff et al., 2012)
S-D rat, male
90 d
NOAEL
6
5.56
1.27
Relative liver weight
(Butenhoff et al., 2012)
S-D rat, male
28 d
BMDLio, Exp4
(NCV)
6.34
5.87
1.3
Relative liver weight
(Foreman et al., 2009)
Sv/129 WT
mouse, male
28 d
LOAEL
35
32.42
1.59a
Relative liver weight
(Foreman et al., 2009)
Sv/129 hPPARa
mouse, male
28 d
BMDLio, Hill
(NCV)
4.41
4.09
2.00
This document is a draft for review purposes only and does not constitute Agency policy.
5-15 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Toxicological Review ofPFBA and Related Salts
aAs this data set only supported identification of a LOAEL, the LOAEL-to-NOAEL uncertainty factor was applied to
facilitate comparison to the other HEDs for liver-weight effects.
This is not the case, however, for all liver effects. Histopathological evaluations of the liver
in male rats exposed to PFBA for 90 days show that hepatocellular hypertrophy occurs at
30 mg/kg-day, whereas hypertrophy occurs only at 150 mg/kg-day in male rats exposed for
28 days (Butenhoffetal.. 2012: van Otterdijk. 2007a. b). Thus, although liver concentrations are
equivalent following 28- or 90-day exposures, that prolonged exposure (i.e., 90 d vs. 28 d) elicits
adverse effects in the liver is readily apparent. Taking into account the increased potential for some
effects in the liver with increasing durations of exposure, and the large uncertainty associated with
the lack of data on whether the effects observed in the subchronic study worsen after chronic
exposure, the UFs were therefore set to 10 for the purposes of the liver endpoints. With regard to
thyroid effects, although no increased sensitivity was observed between short-term and subchronic
exposure durations, chronic exposures could still elicit stronger responses; therefore, the default
UFs was retained for the thyroid endpoints.
As described in EPA's A Review of the Reference Dose and Reference Concentration Processes
(U.S. EPA. 20021. the database uncertainty factor is applied to account for the potential of deriving
an underprotective reference value as a result of incomplete characterization of a chemical's
toxicity. The PFBA database is relatively small but contains high confidence subchronic and
developmental toxicity studies investigating effects in multiple organ systems in male and female
rats and mice.
For PFBA, given the small number of available studies, both a UFd = 10 or a UFd = 3 were
considered due to the limited database (most specifically the lack of a two-generation
developmental/reproductive toxicity study), and a UFd = 3 ultimately was applied. Typically, the
specific study types lacking in a chemical's database that influence the value of the UFd to the
greatest degree are developmental toxicity and multigenerational reproductive toxicity studies.
The PFBA database does include a high confidence Das etal. f20081 developmental toxicity study in
mice. Despite its quality, however, that study fails to cover endpoints related to potential
transgenerational impacts of longer-term exposures evaluated in a two-generation study. The 1994
Reference Concentration Guidance (U.S. EPA. 19941 and 2002 Reference Dose Report (U.S. EPA.
20021 support applying a UFd in situations when such a study is missing. The 2002 Reference Dose
Report fU.S. EPA. 20021 states that "[i]f the RfD/RfC is based on animal data, a factor of 3 is often
applied if either a prenatal toxicity study or a two-generation reproductive study is missing."
Consideration of the PFBA, PFBS (a short-chain perfluoroalkane sulfonic acid with a 4-carbon
backbone like PFBA), PFHxA (a short-chain perfluoroalkyl carboxylic acid),13 and PFHxS (a long-
13The systematic review protocol for PFBA (see Appendix A) defines perfluoroalkyl carboxylic acids with
seven or more perfluorinated carbon groups and perfluoralkane sulfonic acids with six or more
perfluorinated carbon groups as 'long-chain" PFAS. Thus, PFHxA is considered a short-chain PFAS, whereas
PFHxS is considered a long-chain PFAS.
This document is a draft for review purposes only and does not constitute Agency policy.
5-16 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Toxicological Review ofPFBA and Related Salts
chain perfluoroalkane sulfonic acid) databases together, however, diminish the concern that the
availability of a multigenerational reproductive study would result in reference values lower than
those currently derived for PFBA. Although limited in their ability to assess reproductive health or
function, measures of possible reproductive toxicity, including reproductive organ weights
(i.e., epididymis, testis, and ovary weights) were unaffected when measured after exposure to PFBA
for 28 days (Butenhoffet al.. 2012: van Otterdijk. 2007a). Likewise, the available data on
reproductive toxicity in the PFBS database is consistent with this general lack of sensitive
reproductive effects: No biologically significant changes were observed in male mating and fertility
parameters, reproductive organ weights, reproductive hormone levels, or altered sperm
parameters fU.S. EPA. 2018bl The female reproductive effects that were observed (e.g., altered
estrous cyclicity) occurred at doses equal to or higher than those that resulted in effects in other
organ systems (e.g., thyroid, liver), thus indicating they were not more sensitive markers of toxicity.
Further, no notable male or female reproductive effects were observed in epidemiological or
toxicological studies investigating exposure to PFHxA fLuz etal.. 2019: NTP. 2019: Klaunigetal..
2015: Chengelis et al.. 2009) or PFHxS (MDH. 2019). Therefore, when considering the limited
chemical-specific information alongside information gleaned from structurally related compounds,
the lack of a multigenerational reproductive study is not considered a major concern relative to UFd
selection.
Another gap in the PFBA database is the lack of measures of thyroid toxicity in gestationally
exposed offspring and the lack of a developmental neurotoxicity study. Thyroid hormones are
critical in myriad physiological processes and must be maintained at sufficient levels during times
of brain development in utero and after birth. Although no PFBA-specific data on thyroid hormone
levels following gestational exposure are available, total T4 is reduced in both pregnant mice and
their offspring following whole-gestation oral exposure to PFBS, with effects evident in offspring at
PNDs 1, 30, and 60. Therefore, anticipating that effects due to PFBA exposure also could have been
observed had thyroid hormone levels been measured in the Das etal. (2008) developmental study
is reasonable. For PFBS, the PODs for effects in dams and offspring on PND 1 were almost identical,
indicating that thyroid hormone homeostasis is perturbed at equivalent exposure levels in both
pregnant animals and developing offspring. Thus, although some concern remains that thyroid
insufficiency during in utero and perinatal development could be a more sensitive effect ofPFBA
exposure than insufficiency in adults, this concern is mitigated on the basis of data from other PFAS.
Likewise, given that neurodevelopmental effects due to thyroid hormone insufficiency would be
downstream effects, application of a UFd (and derivation of reference values) addressing the
potential for developmental thyroid insufficiency would presumably be protective of any potential
neurodevelopmental endpoints related to that mechanism. The potential for neurodevelopmental
effects independent of a thyroid hormone-related mechanism remains an uncertainty for PFBA.
Lastly, the potential for immunotoxicity and mammary gland effects represents an area of
concern across several constituents of the larger PFAS family (primarily long-chain PFAS). No
This document is a draft for review purposes only and does not constitute Agency policy.
5-17 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Toxicological Review ofPFBA and Related Salts
studies have evaluated these outcomes following PFBA exposure or following exposure to the
structurally related PFBS described above. No chemical-specific information is available to judge
the degree to which the existing endpoints in the PFBA Toxicological Review would be protective of
immunotoxicity or mammary gland effects.
Given the residual concerns for potentially more sensitive effects outlined above, a database
uncertainty factor is considered necessary. Specifically, a value of 3 was selected for the UFd to
account for the uncertainty surrounding the lack of a multigenerational reproductive study,
developmental neurotoxicity study (or information on thyroid hormone perturbation in utero and
postnatally), immunotoxicity, or mammary gland effects. A UFd of 10 was not applied, given that
multiple lines of chemical-specific information or data from structural analogs are available to
partially mitigate the concern that additional study would possibly result in reference values one
order of magnitude lower than the one currently derived. Thus, a UFd value of 3 was applied
because currently available lines of evidence do not fully eliminate this concern.
The candidate values (see Table 5-7) are derived by dividing the PODhed by the composite
uncertainty factor. For example, for relative liver weight in adult rats from Butenhoff et al. (20121.
the candidate value is calculated as:
Candidate value for PFBA = BMDL10 h- UFc (5-5)
Candidate value = 2.04 (m^/kg-d)
Candidate value = 0.002 (m^/kg-d)
Candidate value = 2.0 x 10~3 (m^/kg-d)
This document is a draft for review purposes only and does not constitute Agency policy.
5-18 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Toxicological Review ofPFBA and Related Salts
Table 5-7. Candidate values for perfluorobutanoic acid (PFBA)
Endpoint
PODhed
PFBA
(mg/kg-d)
ufa
UFh
UFs
UFl
UFd
UFC
Candidate
value PFBA
(mg/kg-d)
Candidate
value NH4+
PFB
(mg/kg-d)a
Increased relative liver weight
(Butenhoff et al., 2012)
2.04
3
10
10
1
3
1,000
2.0 x 10"3
2.2 x 10"3
Increased relative liver weight
(Das et al., 2008)
2.46
3
10
10
1
3
1,000
2.5 x 10"3
2.7 x 10"3
Increased liver hypertrophy
(Butenhoff et al., 2012)
1.15
3
10
10
1
3
1,000
1.1 x 10"3
1.2 x 10"3
Decreased total T4
(Butenhoff et al., 2012)
1.27
3
10
10
1
3
1,000
1.3 x 10"3
1.4 x 10"3
Embryo/fetal mortality
(Das et al., 2008)
0.93
3
10
1
1
3
100
9.5 x 10"3
1.0 x 10"2
Delayed eyes opening
(Das et al., 2008)
0.80
3
10
1
1
3
100
8.0 x 10"3
8.6 x 10"3
Delayed vaginal opening
(Das et al., 2008)
0.62
3
10
1
1
3
100
6.2 x 10"3
6.7 x 10"3
Delayed preputial separation
(Das et al., 2008)
29.37
3
10
1
1
3
100
2.9 x 10"1
3.2 x 10"1
a To calculate candidate values for salts of PFBA, multiply the candidate value of interest by the ratio of molecular
weights of the free acid and the salt. For example, for the ammonium salt of PFBA, the RfD would be calculated by
.... . . . . MW ammonium salt 231 . „ „ „ . . ....
multiplying the free acid RfD by 1.079: = — = 1.079. This same conversion can be applied to
MW free acid 214
other salts of PFBA, such as the potassium or sodium salts.
Selection of Lifetime Toxicity Value(s)
Selection of organ/system-specific oral reference doses fosRfDsl
From among the candidate values presented in Table 5-7, organ/system-specific RfDs
(osRfDs) are selected for the individual organ systems identified as hazards in Section 3. The osRfD
values selected were associated with increased liver hypertrophy for liver effects, decreased total
T4 for thyroid effects, and developmental delays (based on the candidate value for delayed time to
vaginal opening) for developmental effects. The confidence decisions about the study, evidence
base, quantification of the POD, and overall RfD for these organ/system-specific values are fully
described in Table 5-8, along with the rationales for selecting those confidence levels. In deciding
overall confidence, confidence in the evidence base is prioritized over the other confidence
decisions. The overall confidence in the osRfD for liver effects is medium, whereas the confidence in
the osRfDs for thyroid effects and developmental effects is medium-low. Selection of the overall RfD
is described in the following section.
This document is a draft for review purposes only and does not constitute Agency policy.
5-19 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Table 5-8. Confidence in the organ/system-specific oral reference doses
(osRfDs) for perfluorobutanoic acid (PFBA)
Confidence
categories
Designation
Discussion
Liver RfD = 1 x 10"3 mg/kg-d PFBA; 1 x 10"3 mg/kg-d NH4+ PFB
Confidence in
study3 used to
derive osRfD
High
Confidence in the study (Butenhoff et al., 2012; van Otterdiik, 2007b) is high given the
study evaluation results (i.e., rating of good or adequate in all evaluation categories)
and characteristics that make it suitable for deriving toxicity values, including the
relevance of the exposure paradigm (route, duration, and exposure levels), use of a
relevant species, and the study size and design.
Confidence in
evidence base
supporting this
hazard
Medium
Confidence in the evidence base for liver effects is medium because there are
consistent, dose-dependent, and biologically coherent effects on organ weight and
histopathology observed in multiple high and medium confidence studies. Although
the available mechanistic evidence also supports the human relevance of observed
effects, there is a sparsity of chemical-specific information. One in vivo PFBA study
(Foreman et al., 2009) is available that indicates non-PPARa modes-of-action are
active in the development of liver effects, but no PFBA-specific studies investigated
activation of other PPAR isoforms or additional pathways. Another limitation of the
database for PFBA-induced liver effects is the lack of a chronic duration study.
Confidence in
quantification
of the PODhed
Medium
Confidence in the quantification of the POD and osRfD is medium given the POD was
based on BMD modeling within the range of the observed data and dosimetric
adjustment was based on PFBA-specific toxicokinetic information, the latter of which
introduces some uncertainty. Another source of potential uncertainty is that
hypertrophy was observed only in the high dose group; however, modeling lesions of
"slight" severity only increased model uncertainty, and thus data for all lesions served
as the basis for BMD modeling.
Overall
confidence in
osRfD
Medium
The overall confidence in the osRfD is medium and is primarily driven by medium
confidence in both the evidence base supporting this hazard and the quantification of
the POD using BMD modeling of data from a high confidence study.
Thyroid RfD = 1 x 10"3 mg/kg-d PFBA; 1 x 10"3 mg/kg-d NH4+ PFB
Confidence in
study3 used to
derive osRfD
High
Confidence in the studv (Butenhoff et al., 2012; van Otterdiik, 2007b) is hiah given the
study evaluation results (i.e., rating of good or adequate in all evaluation categories)
and characteristics that make it suitable for deriving toxicity values, including the
relevance of the exposure paradigm (route, duration, and exposure levels), use of a
relevant species, and the study size and design.
Confidence in
evidence base
supporting this
hazard
Medium
Confidence in the evidence base for thyroid effects is medium because there were
consistent and coherent effects on hormone levels, organ weights, and
histopathology in a single high confidence study. Confidence is decreased by the lack
of coherence between histopathology and TSH, as well as the increased sensitivity of
rodents for developing thyroid hypertrophy compared to humans. Another limitation
of evidence base for thyroid effects is the lack of a chronic-duration or developmental
study.
This document is a draft for review purposes only and does not constitute Agency policy.
5-20 DRAFT-DO NOT CITE OR QUOTE
-------
Toxicological Review ofPFBA and Related Salts
Confidence
categories
Designation
Discussion
Confidence in
quantification
of the PODhed
Medium-low
Confidence in the quantification of the POD and osRfD is medium-low given the POD
was based on a NOAEL (BMD modeling did not provide an adequate fit to the data)
and dosimetric adjustment was based on PFBA-specific toxicokinetic information, the
latter of which introduces some uncertainty. Of note, however, is that a 15% decrease
in total T4 levels, upon which the NOAEL was based, is consistent with a 13%
decrease in total T4 that would correspond to a response level based on 1SD.
Therefore, this NOAEL might not be substantially more uncertain than a BMD-based
POD. This supports a determination that the confidence in the quantification of the
POD is medium-low.
Overall
confidence in
osRfD
Medium-low
The overall confidence in the osRfD is medium-low and is primarily driven by medium
confidence in the evidence base; however, the medium-to-low confidence in the
quantification of the POD does warrant decreasing the overall confidence in the
osRfD.
Developmental RfD = 6 x 10 3 mg/kg-d PFBA; 7 x 10 3 mg/kg-d NH4+ PFB
Confidence in
study3 used to
derive osRfD
High
Confidence in the studv (Das et al., 2008) is hiah given the studv evaluation results
(i.e., rating of good or adequate in all evaluation categories) and characteristics that
make it suitable for deriving toxicity values, including the relevance of the exposure
paradigm (route, duration, and exposure levels), use of a relevant species, and the
study size and design.
Confidence in
evidence base
supporting this
hazard
Medium
Confidence in the evidence base for developmental effects is medium. Although data
are only available in gestationally exposed animals in a single high confidence
developmental toxicity study, there were coherent delays in multiple developmental
milestones (general development, puberty).
Confidence in
quantification
of the PODhed
Medium-low
Confidence in the quantification of the POD and osRfD is medium-to-low given the
POD was based on BMD modeling and dosimetric adjustment was based on PFBA-
specific toxicokinetic information, the latter of which introduces some uncertainty.
Other sources of uncertainty are the use of dosimetric adjustments based on the ratio
of adult toxicokinetic parameters, and that the derived BMDL is approximately
ninefold below the observed range of the data.
Overall
confidence in
osRfD
Medium-low
The overall confidence in the osRfD is medium-low and is primarily driven by the
medium-to-low confidence in the quantification of the POD given the extrapolation
below the range of the observed data. Modeling data from a high confidence study in
a medium-confidence evidence base does not fully mitigate the medium-to-low
confidence in the actual modeling results in this case.
aAII study evaluation details can be found on HAWC.
Selection of overall oral reference dose (RfD) and confidence statement
1 Organ/system-specific RfD values for PFBA selected in the previous section are summarized
2 in Table 5-9.
This document is a draft for review purposes only and does not constitute Agency policy.
5-21 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Toxicological Review ofPFBA and Related Salts
Table 5-9. Organ/system-specific oral reference dose (osRfD) values for
perfluorobutanoic acid (PFBA)
System
Basis
POD
UFC
OSRfD
PFBA
(mg/kg-d)
OSRfD
nh4+ pfb
(mg/kg-d)b
Confidence
Hepatic
Increased
hepatocellular
hypertrophy in
adult male S-D
rats
BMDLhed from
Butenhoff et al.
(2012)
1,000
1 x 10"3
1 X 10"3
Medium
Thyroid
Decreased total
T4 in adult male
S-D rats
NOAELhed from
Butenhoff et al.
(2012)
1,000
1 x 10"3
1 x 10"3
Medium-low
Developmental
Developmental
delays after
gestational
exposure in CD1
mice3
BMDLhed from
Das et al. (2008)
100
6 x 10"3
7 x 10"3
Medium-low
aPOD based on delayed vaginal opening used to represent three developmental delays observed in the study.
b See Table 5-7 for details on how to calculate candidate values for salts of PFBA; the osRfDs presented in this table
have been rounded to 1 significant digit from the candidate values presented in Table 5-7.
From the identified human health hazards ofPFBA exposure and the derived osRfDs for
effects in the liver, thyroid, and developing organism, an overall RfD ofl x 10~3 mg/kg-day PFBA
based on increased liver hypertrophy and decreased total T4 is selected. The selected RfD for the
ammonium salt ofPFBA is also 1 x 10"3 mg/kg-day. These osRfDs are selected as the overall RfD as
they represent effects in two different organ systems with the same osRfD value, including the
osRfD with the highest confidence of all osRfDs derived (i.e., the hepatic osRfD, with medium
confidence). The other available osRfD was interpreted with medium-low confidence and had a
higher osRfD value; thus, it was not selected. Although the overall confidence in the individual liver
and thyroid osRfDs do differ slightly (medium for increased liver hypertrophy and medium-low for
decreased total T4), an overall confidence of medium is selected for the final RfD. This confidence
level of medium is supported given the two osRfDs come from the same high confidence study and
that the evidence bases for both organ systems were rated as medium. The difference in the overall
confidence for the two osRfDs was driven primarily by the confidence in the quantification of the
osRfDs: medium for increased liver hypertrophy and medium-low for decreased total T4. As noted
in Table 5-8, however, the use of the NOAEL approach for decreased total T4 is not substantially
more uncertain than using the BMD approach, given the relatively similar values in PODs that
would be derived using either approach. Thus, although the NOAEL approach is conceptually
associated with more uncertainty than the BMD approach, the confidence in the quantification of
the total T4 POD was downgraded only to medium-low, rather than to low in this specific case. This
This document is a draft for review purposes only and does not constitute Agency policy.
5-22 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Toxicological Review ofPFBA and Related Salts
supports the determination of medium confidence for the overall RfD on the basis of liver and
thyroid effects.
Another consideration in selecting the overall RfD is the difference in composite uncertainty
factors across the three candidate osRfDs. The composite UF for the liver and thyroid osRfDs was
greater than that for developmental effects (1,000 vs. 100), stemming from not applying a UFs for
the developmental effects. Application of the larger composite UF for liver and thyroid effects
results in osRfDs that are fivefold lower than the developmental osRfD and thus protective of PFBA-
induced effects on the developing organism. If the osRfD for developmental effects were chosen as
the overall RfD on the basis of the application of a smaller composite UF, this would raise concerns
that it would not be protective against potential liver and thyroid effects. Lastly, the selection of the
overall RfD based on liver and thyroid effects is further supported by the fact that the confidence in
that RfD is medium, compared with medium-low for developmental effects. Selection of the RfD
based on liver and thyroid effects is presumed to be protective of possible developmental effects in
humans.
Increased liver hypertrophy and decreased total T4 was observed only in male rats exposed
to PFBA, thus possibly identifying males as a susceptible population. As discussed in Section 3.3,
however, this observation in rats could be driven primarily by the observed sex-dependent
differences in toxicokinetics in rats. No compelling information is available that supports a
similarly strong sex dependence in toxicokinetics in humans. Therefore, this RfD is presumed
equally applicable to both male and female humans.
5.2.2. Subchronic Toxicity Values for Oral Exposure (Subchronic Oral Reference Dose [RfD])
Derivation
In addition to providing RfDs for lifetime exposures in multiple systems, this document also
provides an RfD for less-than-lifetime, subchronic-duration exposures. In the case ofPFBA, all
studies used to calculate the RfDs were subchronic or gestational in duration. Therefore, the
method to calculate the subchronic RfDs is identical to that used for calculating the RfDs, minus the
application of a 10-fold UFS for the subchronic studies (see Table 5-6). The individual organs and
systems for which specific candidate subchronic RfD values were derived were the liver, thyroid,
and the developing organism (see Table 5-10).
This document is a draft for review purposes only and does not constitute Agency policy.
5-23 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Toxicological Review ofPFBA and Related Salts
Table 5-10. Candidate subchronic oral reference dose (RfD) values for
perfluorobutanoic acid (PFBA)
Endpoint
PODhed
PFBA
(mg/kg-d)
UFa
UFh
UFs
UFl
UFd
UFC
Candidate
value PFBA
(mg/kg-d)
Candidate
value NH4+
PFB
(mg/kg-d)a
Increased relative liver weight
(Butenhoff et al., 2012)
2.04
3
10
1
1
3
100
2.0 x 10"2
2.2 x 10"2
Increased relative liver weight
(Das et al., 2008)
2.46
3
10
1
1
3
100
2.5 x 10"2
2.7 x 10"2
Increased liver hypertrophy
(Butenhoff et al., 2012)
1.15
3
10
1
1
3
100
1.1 x 10"2
1.2 x 10"2
Decreased total T4
(Butenhoff et al., 2012)
1.27
3
10
1
1
3
100
1.3 x 10"2
1.4 x 10"2
Embryo/fetal mortality
(Das et al., 2008)
0.93
3
10
1
1
3
100
9.3 x 10"3
1.0 x 10"2
Delayed eyes opening
(Das et al., 2008)
0.80
3
10
1
1
3
100
8.0 x 10"3
8.6 x 10"3
Delayed vaginal opening
(Das et al., 2008)
0.62
3
10
1
1
3
100
6.2 x 10"3
6.7 x 10"3
Delayed preputial separation
(Das et al., 2008)
29
3
10
1
1
3
100
2.9 x 10"1
3.1 x 10"1
a To calculate subchronic candidate values, osRfDs, or the subchronic RfD for salts of PFBA, multiply
the value of interest by the ratio of molecular weights of the free acid and the salt. For example, for
the ammonium salt of PFBA, the RfD would be calculated by multiplying the free acid RfD by 1.079:
MW ammonium salt 231 . „ „„ . . . . .
= — = 1.079. This same method of conversion can be applied to other salts of
MW free acid. 214
PFBA, such as the potassium or sodium salts, using the corresponding molecular weights.
From the identified human health hazards ofPFBA exposure and the derived candidate
RfDs, osRfDs of 1 x 10"2 mg/kg-day are selected for liver effects (increased liver hypertrophy) and
thyroid effects (decreased total T4) (corresponding osRfD of 1 x 10"2 mg/kg-day for the ammonium
salt), and an osRfD of 6 x 10"3 mg/kg-day PFBA is selected for developmental effects
(developmental delays based on the candidate value for delayed vaginal opening) (corresponding
osRfD of 7 x 10"3 mg/kg-day for the ammonium salt). The selection of these candidate values over
other candidates and the confidence in these subchronic osRfDs are identical to the confidence in
the osRfDs discussed in the previous section and presented in Table 5-8.
From these subchronic osRfDs, an overall subchronic RfD of 6 x 10~3 mg/kg-day PFBA
based on developmental delays is selected (the corresponding overall subchronic RfD is
7 x 10"3 mg/kg-day for the ammonium salt). This osRfD is selected as the overall subchronic RfD,
as it is the lowest osRfD among the derived subchronic osRfDs, even though it is not the osRfD
interpreted with the highest confidence. In the case of the subchronic RfD, selection need not
This document is a draft for review purposes only and does not constitute Agency policy.
5-24 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Toxicological Review ofPFBA and Related Salts
consider differences in the composite UF, as a value of 100 is applied to all PODs. This is because all
the studies considered for the subchronic RfD are subchronic or gestational duration studies. This
results in the osRfD for developmental delays being approximately 50% lower than the osRfD for
liver or thyroid effects. Although the overall confidence in the osRfD for developmental delays
[medium-low) is lower than for liver effects (medium confidence, see derivation of RfD section),
selection of the developmental osRfD as the overall subchronic RfD is presumed protective of
possible effects in other organ systems. Selection of the liver osRfD, although having a stronger
overall confidence determination, as the overall subchronic RfD would be considered inadequate to
protect against potential developmental effects. Also, although the subchronic RfD is intended to
protect health during a less-than-lifetime exposure to PFBA, developmental delays are appropriate
endpoints on which to base a subchronic RfD. First, as discussed above (Study Selection
subsection), given that the pubertal delays occur during critical periods of development, EPA's
Reproductive Toxicity Guidelines (U.S. EPA. 1996) state that "[significant effects on ... age at
puberty, either early or delayed, should be considered adverse...". Further, delays in reaching
developmental milestones are not phenomena that can be resolved (e.g., after PFBA exposure is
removed), and they can result from short (less-than-lifetime) exposures during discrete windows of
development More importantly, the consequences of these delays can have permanent impacts on
health (e.g., delays in eye opening leading to permanent decrements in visual acuity). So, although
the delay itself might occur only over a short portion of lifetime, the functional consequences are
permanent.
5.2.3. Inhalation Reference Concentration (RfC)
No published studies investigating the effects of subchronic, chronic, or gestational
exposure to PFBA in humans or animals have been identified. Therefore, an RfC is not derived.
5.3. CANCER
5.3.1. Cancer Weight-of-Evidence Descriptor and Derivation of Cancer Risk Values
No studies were identified that evaluated the carcinogenicity ofPFBA in humans or animals.
In accordance with the Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005). EPA concluded
that there is inadequate information to assess carcinogenic potential for PFBA (or salts of PFBA) for
any route of exposure. Therefore, the lack of data on the carcinogenicity of PFBA precludes the
derivation of quantitative estimates for either oral (oral slope factor [OSF]) or inhalation
(inhalation unit risk [IUR]) exposure.
This document is a draft for review purposes only and does not constitute Agency policy.
5-25 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Toxicological Review ofPFBA and Related Salts
REFERENCES
Abe. T: Takahashi. M: Kano. M: Amaike. Y: Ishii. C: Maeda. K: Kudoh. Y: Morishita. T: Hosaka. T:
Sasaki. T: Kodama. S: Matsuzawa. A: Koiima. H: Yoshinari. K. (2017). Activation of nuclear
receptor CAR by an environmental pollutant perfluorooctanoic acid. Arch Toxicol 91: 2365-
2374. http://dx.doi.Org/l0.1007/s00204-016-1888-3
Alexander. E: Pearce. E: Brent. G: Brown. R: Chen. H: Dosiou. C: Grobman. W: Laurberg. P: Lazarus. 1:
Mandel. S: Peeters. R: Sullivan. S. (2017). 2017 Guidelines of the American Thyroid
Association for the diagnosis and management of thyroid disease during pregnancy and the
postpartum. Thyroid 27: 315-389.
Angrish. M: Kaiser. 1: Mcqueen. C: Chorlev. B. (2016). Tipping the balance: Hepatotoxicity and the 4
apical key events of hepatic steatosis [Review], Toxicol Sci 150: 261-268.
http ://dx. doi. or g/10.10 9 3 /toxsci /kfwO 18
Antignac. 1: Vevrand. B: Kadar. H: Marchand. P: Oleko. A: Le Bizec. B: Vandentorren. S. (2013).
Occurrence of perfluorinated alkylated substances in breast milk of French women and
relation with socio-demographical and clinical parameters: Results of the ELFE pilot study.
Chemosphere 91: 802-808.
ATSDR (Agency for Toxic Substances and Disease Registry). (2017). Substance priority list (SPL)
resource page
https://www.atsdr.cdc.gov/spl/resources/ATSDR 2017 Full SPL Spreadsheetxlsx
ATSDR (Agency for Toxic Substances and Disease Registry). (2018a). An overview of perfluoroalkyl
and polyfluoroalkyl substances and interim guidance for clinicians responding to patient
exposure concerns. Interim guidance revised 5/7/2018.
https://www.atsdr.cdc.gov/pfas/docs/pfas clinician fact sheet 508.pdf
ATSDR (Agency for Toxic Substances and Disease Registry). (2018b). Toxicological profile for
perfluoroalkyls. Draft for public comment Atlanta, GA: U.S. Department of Health and
Human Services, Centers for Disease Control and Prevention.
https://www.atsdr.cdc.gOv/toxprofiles/tp200.pdf
Bao. W: Oian. Z: Geiger. S: Liu. E: Liu. Y: Wang. S: Lawrence. W: Yang. B: Hu. L: Zeng. X: Dong. G.
(2017). Gender-specific associations between serum isomers of perfluoroalkyl substances
and blood pressure among Chinese: Isomers of C8 Health Project in China. Sci Total Environ
607-608: 1304-1312. http: / /dx.doi.org/10.1016/i.scitotenv.2017.07.124
Barton. C: Kaiser. M: Russell. M. (2007). Partitioning and removal of perfluorooctanoate during rain
events: the importance of physical-chemical properties Journal of Environmental
Monitoring 9: 839-846.
Batt. A: Ferrari. L. (1995). Manifestations of chemically-induced liver-damage. Clinical Chemistry
41: 1882-1887.
Beggs. K: Mcgreal. S: McCarthy. A: Gunewardena. S: Lampe. I: Lau. C: Apte. U. (2016). The role of
hepatocyte nuclear factor 4-alpha in perfluorooctanoic acid- and perfluorooctanesulfonic
acid-induced hepatocellular dysfunction. Toxicol Appl Pharmacol 304: 18-29.
http://dx.doi.Org/10.1016/j.taap.2016.05.001
This document is a draft for review purposes only and does not constitute Agency policy.
R-l DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Toxicological Review ofPFBA and Related Salts
Bernal. I. (2015). Thyroid hormones in brain development and function. In LJ De Groot; G Chrousos;
K Dungan; KR Feingold; A Grossman; JM Hershman; C Koch; M Korbonits; R McLachlan; M
New; J Purnell; R Rebar; F Singer; A Vinik (Eds.). South Dartmouth, MA: MDTextcom, Inc.
Bjork. I: Wallace. K. (2009). Structure-activity relationships and human relevance for perfluoroalkyl
acid-induced transcriptional activation of peroxisome proliferation in liver cell cultures.
Toxicol Sci 111: 89-99. http://dx.doi.org/10.1093/toxsci/kfpQ93
Blaine. A: Rich. C: Hundal. L: Lau. C: Mills. M: Harris. K: Higgins. C. (2013). Uptake of perfluoroalkyl
acids into edible crops via land applied biosolids: field and greenhouse studies. Environ Sci
Technol 47: 14062-14069. http://dx.doi.org/10.1021/es403094q
Buhrke. T: Kriiger. E: Pevnv. S: RolSler. M: Bitter. K: Lampen. A. (2015). Perfluorooctanoic acid
(PFOA) affects distinct molecular signalling pathways in human primary hepatocytes.
Toxicology 333: 53-62. http://dx.doi.Org/10.1016/i.tox.2015.04.004
Buist. S: Cherrington. N: Choudhuri. S: Hartley. D: Klaassen. C. (2002). Gender-specific and
developmental influences on the expression of rat organic anion transporters. Journal of
Pharmacology and Experimental Therapeutics 301: 145-151.
Burkemper. I: Aweda. T: Rosenberg. A: Lunderberg. D: Peaslee. G: Lapi. S. (2017). Radiosynthesis
and biological distribution of F-18-labeled perfluorinated alkyl substances. Environ Sci
Technol Lett4: 211-215. http://dx.doi.org/10.1021/acs.estlett.7b00042
Butenhoff. 1: Bjork. I: Chang. S: Ehresman. D: Parker. G: Das. K: Lau. C: Lieder. P: van Otterdijk. F:
Wallace. K. (2012). Toxicological evaluation of ammonium perfluorobutyrate in rats:
Twenty-eight-day and ninety-day oral gavage studies. Reprod Toxicol 33: 513-530.
http://dx.doi.Org/10.1016/i.reprotox.2011.08.004
Butenhoff. 1: Kennedy. G: Frame. SR: O'Connor. I: York. R. (2004). The reproductive toxicology of
ammonium perfluorooctanoate (APFO) in the rat. Toxicology 196: 95-116.
http://dx.doi.Org/10.1016/i.tox.2003.ll.005
Calvo. R: Obregon. MI: Escobar del Rev. F: Morreale de Escobar. G. (1992). The rat placenta and the
transfer of thyroid hormones from the mother to fetus: Effects of maternal thyroid status.
Endocrinology 131: 357-365.
Calvo. R: Obregon. Ml: Ruiz de Ona. C: Ferreiro. B: Escobar Del Rev. E: Morreale de Escobar. G.
(1990). Thyroid hormone economy in pregnant rats near term: A "physiological" animal
model of nonthyroidal illness? Endocrinology 127: 10-16.
CDC (Centers for Disease Control and Prevention). (2019). Fourth National Report on Human
Exposure to Environmental Chemicals, Updated Tables, January 2019, Volume 1. Centers for
Disease Control and Prevention, National Health and Nutrition Examination Survey.
https://www.cdc.gov/exposurereport/pdf/FourthReport UpdatedTables Volumel Tan201
9-508.pdf
Cerrutti. I: Brandoni. A: Ouaglia. N: Torres. A. (2002). Sex differences in p-aminohippuric acid
transport in rat kidney: Role of membrane fluidity and expression of 0AT1. Molecular and
Cellular Biochemistry 233: 175-179.
Chang. S: Das. K: Ehresman. D: Ellefson. M: Gorman. G: Hart. 1: Noker. P: Tan. Y: Lieder. P: Lau. C:
Olsen. G: Butenhoff. 1. (2008). Comparative pharmacokinetics of perfluorobutyrate in rats,
mice, monkeys, and humans and relevance to human exposure via drinking water. Toxicol
Sci 104: 40-53. http://dx.d0i.0rg/l0.1093/toxsci/kfnO57
This document is a draft for review purposes only and does not constitute Agency policy.
R-2 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Toxicological Review ofPFBA and Related Salts
Chengelis. C: Kirkpatrick. 1: Radovskv. A: Shinohara. M. (2009). A 90-day repeated dose oral
(gavage) toxicity study of perfluorohexanoic acid (PFHxA) in rats (with functional
observational battery and motor activity determinations) Reprod Toxicol 27: 342-351.
Choksi. N: Tahnke. G: StHilaire. C: Shelby. M. (2003). Role of thyroid hormones in human and
laboratory animal reproductive health. Birth Defects Research, Part B: Developmental and
Reproductive Toxicology 68: 479-491.
Crebelli. R: Caiola. S: Conti. L: Cordelli. E: De Luca. G: Dellatte. E: Eleuteri. P: Iacovella. N: Leopardi.
P: Marcon. F: Sanchez. M: Sestili. P: Siniscalchi. E: Villani. P. (2019). Can sustained exposure
to PFAS trigger a genotoxic response? A comprehensive genotoxicity assessment in mice
after subacute oral administration of PFOA and PFBA Regulatory Toxicology and
Pharmacology 106: 169-177.
Crofton. K. (2004). Developmental disruption of thyroid hormone: Correlations with hearing
dysfunction in rats. Risk Anal 24: 1665-1671. http: //dx.doi.org/10.1111 /i.0272-
4332.2004.00557.x
Das. K: Grey. B: Zehr. R: Wood. C: Butenhoff. 1: Chang. S: Ehresman. D: Tan. Y: Lau. C. (2008). Effects
of perfluorobutyrate exposure during pregnancy in the mouse. Toxicol Sci 105: 173-181.
http: / /dx. doi. or g/10.109 3 /toxsci /kfn09 9
Das. K: Wood. C: Lin. M: Starkov. A: Lau. C: Wallace. K: Corton. 1: Abbott. B. (2017). Perfluoroalkyl
acids-induced liver steatosis: Effects on genes controlling lipid homeostasis. Toxicology
378: 37-52. http://dx.doi.Org/10.1016/i.tox.2016.12.007
di Masi. A: De Marinis. E: Ascenzi. P: Marino. M. (2009). Nuclear receptors CAR and PXR: Molecular,
functional, and biomedical aspects [Review], Mol Aspects Med 30: 297-343.
http://dx.doi.Org/10.1016/j.mam.2009.04.002
Espinosa. 1: Strvker. M. (2012). Development and plasticity of the primary visual cortex [Review],
Neuron 75: 230-249. http://dx.doi.Org/10.1016/i.neuron.2012.06.009
Feng. X: Cao. X: Zhao. S: Wang. X: Hua. X: Chen. L: Chen. L. (2017). Exposure of pregnant mice to
perfluorobutanesulfonate causes hypothyroxinemia and developmental abnormalities in
female offspring. Toxicol Sci 155: 409-419. http://dx.doi.org/10.1093/toxsci/kfw219
Finken. M: van Eiisden. M: Loomans. E: Vriikotte. T: Rotteveel. 1. (2013). Maternal
hypothyroxinemia in early pregnancy predicts reduced performance in reaction time tests
in 5- to 6-year-old offspring. J Clin Endocrinol Metab 98: 1417-1426.
http://dx.doi.org/10.1210/ic.2012-3389
Foreman. 1: Chang. S: Ehresman. D: Butenhoff. 1: Anderson. C: Palkar. P: Kang. B: Gonzalez. F: Peters.
L (2009). Differential hepatic effects of perfluorobutyrate mediated by mouse and human
PPAR-Alpha. Toxicol Sci 110: 204-211. http://dx.doi.org/10.1093/toxsci /kfp077
Forhead. A: Fowden. A. (2014). Thyroid hormones in fetal growth and prepartum maturation
[Review], J Endocrinol 221: R87-R103. http://dx.doi.org/10.1530/TOE-14-0Q25
Fu. Y: Wang. T: Fu. 0: Wang. P: Lu. Y. (2014). Associations between serum concentrations of
perfluoroalkyl acids and serum lipid levels in a Chinese population. Ecotoxicol Environ Saf
106: 246-252. http://dx.doi.Org/10.1016/i.ecoenv.2014.04.039
Fukuda. H: Ohshima. K: Mori. M: Kobavashi. I: Greer. MA. (1980). Sequential changes in the
pituitary-thyroid axis during pregnancy and lactation in the rat Endocrinology 107: 1711-
1716.
This document is a draft for review purposes only and does not constitute Agency policy.
R-3 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Toxicological Review ofPFBA and Related Salts
Gebbink. W: Berger. U: Cousins. I. (2015). Estimating human exposure to PFOS isomers and PFCA
homologues: the relative importance of direct and indirect (precursor) exposure. Environ
Int74: 160-169. http://dx.doi.Org/10.1016/i.envint.2014.10.013
Gilbert. M. (2011). Impact of low-level thyroid hormone disruption induced by propylthiouracil on
brain development and function. Toxicol Sci 124: 432-445.
http://dx.doi.org/10.1093/toxsci/kfr244
Gilbert. M: Sanchez-Huerta. K: Wood. C. (2016). Mild thyroid hormone insufficiency during
development compromises activity-dependent neuroplasticity in the hippocampus of adult
male rats : Supplemental data [Supplemental Data], Endocrinology 157: 774-787.
Gilbert. M: Zoeller. R. (2010). Thyroid hormones-impact on the developing brain: Possible
mechanisms of neurotoxicity. In GJ Harry; HA Tilson (Eds.), Neurotoxicology (3rd ed., pp.
79-111). New York, NY: Informa Healthcare.
Gomis. M: Vestergren. R: Borg. D: Cousins. I. f20181. Comparing the toxic potency in vivo of long-
chain perfluoroalkyl acids and fluorinated alternatives. Environ Int 113: 1-9.
http://dx.doi.Org/10.1016/j.envint2018.01.011
Guvatt. G: Oxman. A: Akl. E: Kunz. R: Vist. G: Brozek. I: Norris. S: Falck-Ytter. Y: Glasziou. P: DeBeer.
H: Taeschke. R: Rind. D: Meerpohl. I: Dahm. P: Schiinemann. H. (2011). GRADE guidelines: 1.
Introduction-GRADE evidence profiles and summary of findings tables. J Clin Epidemiol 64:
383-394. http://dx.doi.Org/10.1016/j.jclinepi.2010.04.026
Haddow. 1: Palomaki. G: Allan. W: Williams. 1: Knight. G: Gagnon. I: O'Heir. C: Mitchell. M: Hermos. R:
Waisbren. S: Faix. 1: Klein. R. (1999). Maternal thyroid deficiency during pregnancy and
subsequent neuropsychological development of the child. N Engl J Med 341: 549-555.
http: //dx.doi.org/l 0.1056/NETM199908193410801
Hall. A: Elcombe. C: Foster. 1: Harada. T: Kaufmann. W: Knippel. A: Kiittler. K: Malarkev. D:
Maronpot. R: Nishikawa. A: Nolte. T: Schulte. A: Strauss. V: York. M. (2012). Liver
hypertrophy: a review of adaptive (adverse and non-adverse) changes-conclusions from
the 3rd International ESTP Expert Workshop [Review], Toxicol Pathol 40: 971-994.
http://dx.doi.org/10.1177/0192623312448935
Hassan. I: El-Masri. H: Kosian. PA: Ford. I: Degitz. ST: Gilbert. ME. (2017). Neurodevelopment and
thyroid hormone synthesis inhibition in the rat: Quantitative understanding within the
adverse outcome pathway framework. Toxicol Sci 160: 57-73.
http://dx.doi.org/10.1093/toxsci/kfxl63
Henrichs. 1: Bongers-Schokking. I: Schenk. 1: Ghassabian. A: Schmidt. H: Visser. T: Hooiikaas. H: de
Muinck Keizer-Schrama. S: Hofman. A: Taddoe. V: Visser. W: Steegers. E: Verhulst. F: de
Riike. Y: Tiemeier. H. (2010). Maternal thyroid function during early pregnancy and
cognitive functioning in early childhood: The generation R study. J Clin Endocrinol Metab
95: 4227-4234. http://dx.doi.org/l0.1210/ic.2010-0415
Heo. I: Lee. 1: Kim. S: Oh. I. (2014). Foodstuff analyses show that seafood and water are major
perfluoroalkyl acids (PFAAs) sources to humans in Korea. J Hazard Mater 279: 402-409.
http://dx.doi.Org/10.1016/i.ihazmat2014.07.004
Hill. A. (1965). The environment and disease: Association or causation? Proc R Soc Med 58: 295-
300.
Hillman. N: Kallapur. S: lobe. A. (2012). Physiology of transition from intrauterine to extrauterine
life. Clin Perinatol 39: 769-783.
This document is a draft for review purposes only and does not constitute Agency policy.
R-4 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Toxicological Review ofPFBA and Related Salts
Hood. A: Hashmi. R: Klaassen. C. (1999). Effects of microsomal enzyme inducers on thyroid-
follicular cell proliferation, hyperplasia, and hypertrophy. Toxicol Appl Pharmacol 160: 163-
170. http://dx.doi.org/10.10Q6/taap.1999.8752
Hood. A: Klaassen. C. (2000). Differential effects of microsomal enzyme inducers on in vitro
thyroxine (T(4)) and triiodothyronine (T(3)) glucuronidation. Toxicol Sci 55: 78-84.
Howdeshell. K. (2002). A model of the development of the brain as a construct of the thyroid system
[Review], Environ Health Perspect 110 Suppl 3: 337-348.
http://dx.doi.org/10.1289/ehp.02110s3337
Hulbert. A. (2000). Thyroid hormones and their effects: A new perspective [Review], Biol Rev Camb
Philos Soc 75: 519-631.
Hurley. M: Andersen. M: Wallington. T: Ellis. D: Martin. 1: Maburv. S. (2004). Atmospheric chemistry
of perfluorinated carboxylic acids: Reaction with OH radicals and atmospheric lifetimes
Journal of Physical Chemistry A 108: 615-620.
Ikeda. T: Aiba. K: Fukuda. K: 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. http://dx.doi.org/10.1093/oxfordiournals.ibchem.al35302
Taeschke. H: Lemasters. 1. (2003). Apoptosis versus oncotic necrosis in hepatic
ischemia/reperfusion injury. Gastroenterology 125: 1246-1257.
Toshi-Barve. S: Kirpich. I: Cave. M: Marsano. L: McClain. C. (2015). Alcoholic, nonalcoholic, and
toxicant-associated steatohepatitis: Mechanistic similarities and differences [Review],
CMGH 1: 356-367. http://dx.doi.org/10.1016/i.icmgh.2015.05.006
Tulvez. 1: Alvarez-Pedrerol. M: Rebagliato. M: Murcia. M: Forns. 1: Garcia-Esteban. R: Lertxundi. N:
Espada. M: Tardon. A: Riano Galan. I: Sunver. 1. (2013). Thyroxine levels during pregnancy
in healthy women and early child neurodevelopment Epidemiology 24: 150-157.
lust. W: Gorgas. K: Hartl. F: Heinemann. P: Salzer. M: Schimassek. H. (1989). Biochemical effects and
zonal heterogeneity of peroxisome proliferation induced by perfluorocarboxylic acids in rat
liver. Hepatology 9: 570-581. http://dx.doi.org/10.1002/hep.1840090411
Kaiser. 1: Lipscomb. 1: Wesselkamper. S. (2012). Putative Mechanisms of Environmental Chemical-
Induced Steatosis. Int J Toxicol 31: 551-563.
http://dx.d0i.0rg/l 0.1177/1091581812466418
Kass. G. (2006). Mitochondrial involvement in drug-induced hepatic injury. Chemico-Biological
Interactions 163: 145-159.
Kato. Y: Kuge. K: Kusuhara. H: Meier. P: Sugivama. Y. (2002). Gender difference in the urinary
excretion of organic anions in rats. Journal of Pharmacology and Experimental Therapeutics
302: 483-489.
Kim. D: Kim. U. nl: Kim. H: Choi. SD: Oh. I. (2016). Perfluoroalkyl substances in serum from South
Korean infants with congenital hypothyroidism and healthy infants - Its relationship with
thyroid hormones. Environ Res 147: 399-404.
http://dx.doi.0rg/lO.lOl6/i.envres.2Ol6.O2.O37
Kim. S: Kannan. K. (2007). Perfluorinated acids in air, rain, snow, surface runoff, and lakes: relative
importance of pathways to contamination of urban lakes. Environ Sci Technol 41: 8328-
8334. http://dx.doi.org/10.1021/es072107t
This document is a draft for review purposes only and does not constitute Agency policy.
R-5 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Toxicological Review ofPFBA and Related Salts
Klaunig. I: Shinohara. M: Iwai. H: Chengelis. C: Kirkpatrick. 1: Wang. Z: Bruner. R. (2015). Evaluation
of the chronic toxicity and carcinogenicity of perfluorohexanoic acid (PFHxA) in Sprague-
Dawley rats. Toxicol Pathol 43: 209-220. http://dx.doi.org/10.n77/0192623314530532
Kleiner. D: Makhlouf. H. (2016). Histology of Nonalcoholic Fatty Liver Disease and Nonalcoholic
Steatohepatitis in Adults and Children. Clinics in Liver Disease 20: 293-+.
Kleiner. D: Weerasinghe. P: Buia. L. (2012). Oncosis: An important non-apoptotic mode of cell death.
Experimental and Molecular Pathology 93: 302-308.
Korevaar. T: Muetzel. R: Medici. M: Chaker. L: Taddoe. V: de Riike. Y: Steegers. E: Visser. T: White. T:
Tiemeier. H: Peeters. R. (2016). Association of maternal thyroid function during early
pregnancy with offspring IQ and brain morphology in childhood: A population-based
prospective cohort study. Lancet Diabetes Endocrinol 4: 35-43.
http://dx.doi.org/10.1016/S2213-8587ri5100327-7
Kotthoff. M: Miiller. 1: liirling. H: Schlummer. M: Fiedler. D. (2015). Perfluoroalkyl and
polyfluoroalkyl substances in consumer products. Environ Sci Pollut Res Int 22: 14546-
14559. http://dx.doi.org/10.1007/sl 1356-015-4202-7
Lau. C. (2015). Perfluorinated compounds: An overview. In JC DeWitt (Ed.), Toxicological effects of
perfluoroalkyl and polyfluoroalkyl substances (pp. 1-21). New York: Springer.
http://dx.doi. org/10.1007/978-3-319-15518-0 1
Lau. C: Butenhoff. 1: Rogers. 1. (2004). The developmental toxicity of perfluoroalkyl acids and their
derivatives [Review], Toxicol Appl Pharmacol 198: 231-241.
http://dx.doi.Org/10.1016/i.taap.2003.ll.031
Lau. C: Thibodeaux. 1: Hanson. R: Narotskv. M: Rogers. 1: Lindstrom. A: Strvnar. M. (2006). Effects of
perfluorooctanoic acid exposure during pregnancy in the mouse. Toxicol Sci 90: 510-518.
http: / /dx. doi. or g/10.109 3 /toxsci/kfj 105
Lau. C: Thibodeaux. 1: Hanson. R: Rogers. 1: Grey. B: Stanton. M: Butenhoff. 1: Stevenson. L. (2003).
Exposure to perfluorooctane sulfonate during pregnancy in rat and mouse. II: postnatal
evaluation. Toxicol Sci 74: 382-392. http://dx.doi.org/10.1093/toxsci/kfgl22
Lavado-Autric. R: Auso. E: Garcia-Velasco. I: Del Carmen Arufe. M: Escobar Del Rev. F: Berbel. P:
Morreale De Escobar. G. (2003). Early maternal hypothyroxinemia alters histogenesis and
cerebral cortex cytoarchitecture of the progeny. J Clin Invest 111: 1073-1082.
http://dx.doi.org/10.1172/TCI16262
Li. M: Zeng. X: Oian. Z: Vaughn. M: Sauve. S: Paul. G: Lin. S: Lu. L: Hu. L: Yang. B: Zhou. Y: Oin. X: Xu. S:
Bao. W: Zhang. Y: Yuan. P: Wang. I: Zhang. C: Tian. Y: Nian. M: Xiao. X: Fu. C: Dong. G.
(2017a). Isomers of perfluorooctanesulfonate (PFOS) in cord serum and birth outcomes in
China: Guangzhou Birth Cohort Study. Environ Int 102: 1-8.
http: / /dx. doi. or g/10.1016/i. envint 2017.03.006
Li. Y: Cheng. Y: Xie. Z: Zeng. F. (2017b). Perfluorinated alkyl substances in serum of the southern
Chinese general population and potential impact on thyroid hormones. Sci Rep 7: 43380.
http://dx.doi.org/10.1038/srep43380
Liu. X: Guo. Z: Krebs. K: Pope. R: Roache. N. (2014). Concentrations and trends of perfluorinated
chemicals in potential indoor sources from 2007 through 2011 in the US. Chemosphere 98:
51-57. http://dx.doi.Org/10.1016/i.chemosphere.2013.10.001
This document is a draft for review purposes only and does not constitute Agency policy.
R-6 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Toxicological Review ofPFBA and Related Salts
Ljubojevic. M: Balen. D: Breljak. D: Kusan. M: Anzai. N: Bahn. A: Burckhardt. G: Sabolic. I. (2007).
Renal expression of organic anion transporter 0AT2 in rats and mice is regulated by sex
hormones. American Journal of Physiology: Renal Physiology 292: F361-F372.
Ljubojevic. M: Herak-Kramberger. C: Hagos. Y: Bahn. A: Endou. H: Burckhardt. G: Sabolic. I. (2004).
Rat renal cortical 0AT1 and 0AT3 exhibit gender differences determined by both androgen
stimulation and estrogen inhibition. American Journal of Physiology: Renal Physiology 287:
F124-F138.
Loccisano. A: Campbell. 1. Tr: Andersen. M: Clewell. HI. (2011). Evaluation and prediction of
pharmacokinetics of PFOA and PFOS in the monkey and human using a PBPK model.
Regulatory Toxicology and Pharmacology 59: 157-175.
Loccisano. A: Campbell. 1. Tr: Butenhoff. 1: Andersen. M: Clewell. HI. (2012). Comparison and
evaluation of pharmacokinetics of PFOA and PFOS in the adult rat using a physiologically
based pharmacokinetic model. Reprod Toxicol 33: 452-467.
Lorenzo. M: Farre. M: Blasco. C: Onghena. M: Pico. Y: Barcelo. D. (2016). Perfluoroalkyl substances
in Breast milk, infant formula and baby food from Valencian Community (Spain).
Environmental Nanotechnology, Monitoring & Management 6: 108-115.
Luz. A: Anderson. 1: Goodrum. P: Durda. I. (2019). Perfluorohexanoic acid toxicity, part I:
Development of a chronic human health toxicity value for use in risk assessment.
Regulatory Toxicology and Pharmacology 103: 41-55.
Malhi. H: Gores. G: Lemasters. 1. (2006). Apoptosis and necrosis in the liver: A tale of two deaths?
Hepatology 43: S31-S44.
MDH (Minnesota Department of Health). (2017a). 2017 Health Based Guidance for Water:
Toxicological Summary for: Perfluorooctane Sulfonate (PFOS).
https://www.health.state.mn.us/communities/environment/risk/docs/guidance/gw/pfos.
pdf
MDH (Minnesota Department of Health). (2017b). PFBA and Drinking Water.
https://www.health.state.mn.us/communities/environment/risk/docs/guidance/gw/pfbai
nfo.pdf
MDH (Minnesota Department of Health). (2018). Health based guidance for water:
Perfluorooctanoic acid toxicological summary
https://www.health.state.mn.us/communities/environment/risk/docs/guidance/gw/pfoa.
pdf
MDH (Minnesota Department of Health). (2019). Health Based Guidance for Water: Toxicological
Summary for: Perfluorohexane sulfonate (PFHxS).
https://www.health.state.mn.us/communities/environment/risk/docs/guidance/gw/pfhx
s.pdf
Mellor. C: Steinmetz. F: Cronin. M. (2016). The identification of nuclear receptors associated with
hepatic steatosis to develop and extend adverse outcome pathways [Review], Crit Rev
Toxicol 46: 138-152. http: //dx.doi.org/10.3109/10408444.2015.1089471
Miller. M: Crofton. K: Rice. D: Zoeller. R. (2009). Thyroid-disrupting chemicals: Interpreting
upstream biomarkers of adverse outcomes [Review], Environ Health Perspect 117: 1033-
1041. http://dx.doi.org/10.1289/ehp.0800247
This document is a draft for review purposes only and does not constitute Agency policy.
R-7 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Toxicological Review ofPFBA and Related Salts
Mirabella. G: Feig. D: Astzalos. E: Perlman. K: Rovet. IF. (2000). The effect of abnormal intrauterine
thyroid hormone economies on infant cognitive abilities. J Pediatr Endocrinol Metab 13:
191-194.
Morgan. R: Thayer. K: Bero. L: Bruce. N: Falck-Ytter. Y: Ghersi. D: Guvatt. G: Hooijmans. C:
Langendam. M: Mandrioli. D: Mustafa. R: Rehfuess. E: Roonev. A: Shea. B: Silbergeld. E:
Sutton. P: Wolfe. M: Woodruff. T: Verbeek. 1: Hollowav. A: Santesso. N: Schiinemann. H.
(2016). GRADE: Assessing the quality of evidence in environmental and occupational health.
Environ Int 92-93: 611-616. http://dx.doi.Org/10.1016/j.envint.2016.01.004
Morreale De Escobar. G: Calvo. R: Obregon. M: Escobar Del Rev. F. (1990). Contribution of maternal
thyroxine to fetal thyroxine pools in normal rats near term. Endocrinology 126: 2765-2767.
Morreale de Escobar. G: Obregon. M: Escobar del Rev. F. (2004). Role of thyroid hormone during
early brain development [Review], Eur J Endocrinol 151 Suppl 3: U25-U37.
http://dx.doi.Org/10.1530/eie.0.151u025
Nian. M: Li. 0: Bloom. M: Oian. Z: Svberg. K: Vaughn. M: Wang. S: Wei. 0: Zeeshan. M: Gurram. N:
Chu. C: Wang. 1: Tian. Y: Hu. L: Liu. K: Yang. B: Liu. R: Feng. D: Zeng. X: Dong. G. (2019). Liver
function biomarkers disorder is associated with exposure to perfluoroalkyl acids in adults:
Isomers of C8 Health Project in China. Environ Res 172: 81-88.
http://dx.doi.Org/10.1016/i.envres.2019.02.013
Nilsson. H: Karrman. A: Westberg. H: Rotander. A: van Bavel. B: Lindstrom. G. (2010). A time trend
study of significantly elevated perfluorocarboxylate levels in humans after using fluorinated
ski wax. Environ Sci Technol 44: 2150-2155. http://dx.doi.org/10.1021/es9034733
NTDWOI (New Jersey Drinking Water Quality Institute). (2017). Public review draft: Health-based
maximum contaminant level support document: Perfluorooctane sulfonate (PFOS) (CAS #:
1763-23-1; Chemical Formula: C8HF1703S). NJDWQI Health Effects Subcommittee.
https://cswab.org/wp-content/uploads/2018/09/New-Tersey-Scientific-Documentation-
Supporting-PF0S-13-ppt.pdf
NLM (National Institutes of Health, National Library of Medicine). (2013). TOXNET Toxicology Data
Network, Hazardous Substances Database (HSDB): Perfluoro-n-nonanoic acid. Available
online at https://toxnet.nlm.nih.gov/cgi-
bin/sis/search/a?dbs+hsdb:@term+@DOCNO+8040
NLM (National Institutes of Health, National Library of Medicine). (2016). TOXNET Toxicology Data
Network, Hazardous Substances Database (HSDB): Perfluorohexanoic acid. Available online
athttps://toxnetnlm.nih.gov/cgi-bin/sis/search/a?dbs+hsdb:@term+@DOCNO+8299
NLM (National Institutes of Health, National Library of Medicine). (2017). TOXNET Toxicology Data
Network, Hazardous Substances Database (HSDB): Perfluorohexanesulfonic acid. Available
online at https://toxnet.nlm.nih.gov/cgi-
bin/sis /search/a?dbs+hsdb: @term+@DQCN0+82 74
NTP (National Toxicology Program). (2018). 28-day evaluation of the toxicity (C20613) of
perfluorohexanoic acid (PFHxA) (307-24-4) on Harlan Sprague-Dawley rats exposed via
gavage. Research Triangle Park, NC.
https://tools.niehs.nih.gov/cebs3/views/?action=main.dataReview&bin id=3879
NTP (National Toxicology Program). (2019). NTP technical report on the toxicity studies of
perfluoroalkyl carboxylates (perfluorohexanoic acid, perfluorooctanoic acid,
perfluorononanoic acid, and perfluorodecanoic acid) administered by gavage to Sprague
Dawley (Hsd:Sprague Dawley SD) rats. (Toxicity Report 97). T0X-97: Perfluorohexanoic
This document is a draft for review purposes only and does not constitute Agency policy.
R-8 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Toxicological Review ofPFBA and Related Salts
acid (307-24-41. Perfluorooctanoic acid (335-67-11. Perfluorononanoic acid (375-95-11.
Perfluorodecanoic acid (335-76-21. WY-14643 (50892-23-41 | Chemical Effects in Biological
Systems fnih.govl
O'Shaughnessv. KL: Kosian. PA: Ford. 1L: Oshiro. WM: Degitz. ST: Gilbert. ME. (2018). Developmental
thyroid hormone insufficiency induces a cortical brain malformation and learning
impairments: A cross-fostering study. Toxicol Sci 163: 101-115.
http ://dx. doi. or g/10.109 3 /toxsci/kfyO 16
Obregon. M: Calvo. R: Del Rev. F: de Escobar. G. (2007). Ontogenesis of thyroid function and
interactions with maternal function [Review], Endocr Dev 10: 86-98.
http: //dx.doi.org/10.1159/0000106821
OECD (Organisation for Economic Co-Operation and Development). (2010). TestNo. 417:
Toxicokinetics. Paris, France.
OECD (Organisation for Economic Co-operation and Development). (2018). Toward a new
comprehensive global database of per- and polyfluoroalkyl substances (PFASs): Summary
report on updating the OECD 2007 list of per- and polyfluoroalkyl substances (PFASs).
(ENV/JM/MONO(2018)7). http://www.oecd.org/chemicalsafety/portal-perfluorinated-
chemicals/
Patel. 1: Landers. K: Li. H: Mortimer. R: Richard. K. (2011). Thyroid hormones and fetal neurological
development J Endocrinol 209: 1-8.
Perez. F: Nadal. M: Navarro-Ortega. A: Fabrega. F: Domingo. 1: Barcelo. D: Farre. M. (2013).
Accumulation of perfluoroalkyl substances in human tissues. Environ Int 59: 354-362.
http://dx.doi.Org/10.1016/i.envint2013.06.004
Permadi. H: Lundgren. B: Andersson. K: Depierre. TW. (1992). Effects of perfluoro fatty acids on
xenobiotic-metabolizing enzymes, enzymes which detoxify reactive forms of oxygen and
lipid peroxidation in mouse liver. Biochem Pharmacol 44: 1183-1191.
http://dx.doi.Org/10.1016/0006-2952r92190383-T
Permadi. H: Lundgren. B: Andersson. K: Sundberg. C: Depierre. TW. (1993). Effects of perfluoro fatty
acids on peroxisome proliferation and mitochondrial size in mouse liver: dose and time
factors and effect of chain length. Xenobiotica 23: 761-770.
Post. G: Louis. 1: Lippincott. R: Procopio. N. (2013). Occurrence of perfluorinated compounds in raw
water from New Jersey public drinking water systems. Environ Sci Technol 47: 13266-
13275. http:/ /dx.doi.org/10.1021 /es402884x
Prevedouros. K: Cousins. I: Buck. R: Korzeniowski. S. (2006). Sources, fate and transport of
perfluorocarboxylates. Environ Sci Technol 40: 32-44.
Puig-Domingo. M: Vila. L. (2013). The implications of iodine and its supplementation during
pregnancy in fetal brain development [Review], Curr Clin Pharmacol 8: 97-109.
Rao. 1: Scott. A. (1992). A simple method for the analysis of clustered binary data. Biometrics 48:
577-585.
Roman. G: Ghassabian. A: Bongers-Schokking. 1: Taddoe. V: Hofman. A: de Riike. Y: Verhulst. F:
Tiemeier. H. (2013). Association of gestational maternal hypothyroxinemia and increased
autism risk. Annals of Neurology 74: 733-742.
This document is a draft for review purposes only and does not constitute Agency policy.
R-9 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Toxicological Review ofPFBA and Related Salts
Rosen. M: Das. K: Roonev. I: Abbott. B: Lau. C: Corton. 1. (2017). PPARa-independent transcriptional
targets of perfluoroalkyl acids revealed by transcript profiling. Toxicology 387: 95-107.
http://dx.doi.Org/10.1016/i.tox.2017.05.013
Rosen. M: Das. K: Wood. C: Wolf. C: Abbott. B: Lau. C. (2013). Evaluation of perfluoroalkyl acid
activity using primary mouse and human hepatocytes. Toxicology 308: 129-137.
http://dx.doi.Org/10.1016/i.tox.2013.03.011
Rosenmai. A: Ahrens. L: le Godec. T: Lundqvist. 1: Oskarsson. A. (2018). Relationship between
peroxisome proliferator-activated receptor alpha activity and cellular concentration of 14
perfluoroalkyl substances in HepG2 cells. J Appl Toxicol 38: 219-226.
http: / /dx. doi. or g/10.10 0 2 /i at 3 515
Ruan. T: Sulecki. L: Wolstenholme. B: Tiang. G: Wang. N: Buck. R. (2014). 6:2 Fluorotelomer iodide in
vitro metabolism by rat liver microsomes: comparison with [1,2-(14)C] 6:2 fluorotelomer
alcohol. Chemosphere 112: 34-41. http://dx.doi.Org/10.1016/i.chemosphere.2014.02.068
Russell. M: Himmelstein. M: Buck. R. (2015). Inhalation and oral toxicokinetics of 6:2 FTOH and its
metabolites in mammals. Chemosphere 120: 328-335.
http://dx.doi.Org/10.1016/i.chemosphere.2014.07.092
Schiinemann. H: Hill. S: Guvatt. G: Akl. E: Ahmed. F. (2011). The GRADE approach and Bradford
Hill's criteria for causation. J Epidemiol Community Health 65: 392-395.
http://dx.doi.org/10.1136/jech.2010.119933
Sferruzzi-Perri. A: Vaughan. 0: Forhead. A: Fowden. A. (2013). Hormonal and nutritional drivers of
intrauterine growth [Review], Curr Opin Clin Nutr Metab Care 16: 298-309.
http: //dx.doi.org/10.1097/MCO.QbOl 3e32835e3643
Song. X: Tang. S: Zhu. H: Chen. Z: Zang. Z: Zhang. Y: Niu. X: Wang. X: Yin. H: Zeng. F: He. C. (2018).
Biomonitoring PFAAs in blood and semen samples: Investigation of a potential link between
PFAAs exposure and semen mobility in China. Environ Int 113: 50-54.
http://dx.doi.Org/10.1016/j.envint2018.01.010
Stahl. L: Snyder. B: Olsen. A: Kincaid. T: Wathen. I: Mccarty. H. (2014). Perfluorinated compounds in
fish from U.S. urban rivers and the Great Lakes. Sci Total Environ 499: 185-195.
http: / /dx. doi. or g/10.1016/i. scitotenv. 2014.07.126
Stenzel. D: Huttner. W. (2013). Role of maternal thyroid hormones in the developing neocortex and
during human evolution. Frontiers in Neuroanatomy 7: 9.
http://dx.doi.org/10.3389/fnana.2013.00Q19
Sun. M: Arevalo. E: Strvnar. M: Lindstrom. A: Richardson. M: Kearns. B: Pickett. A: Smith. C: Knappe.
EL (2016). Legacy and emerging perfluoroalkyl substances are important drinking water
contaminants in the Cape Fear River Watershed of North Carolina. Environ Sci Technol Lett
3: 415-419. http://dx.doi.org/10.1021/acs.estlett6b00398
Sundstrom. M: Bogdanska. I: Pham. H: Athanasios. V: Nobel. S: Mcalees. A: Eriksson. 1: Depierre. I:
Bergman. A. (2012a). Radiosynthesis of perfluorooctanesulfonate (PFOS) and
perfluorobutanesulfonate (PFBS), including solubility, partition and adhesion studies.
Chemosphere 87: 865-871. http://dx.doi.Org/10.1016/i.chemosphere.2012.01.027
Sundstrom. M: Chang. SC: Noker. PE: Gorman. GS: Hart. TA: Ehresman. DT: Bergman. A: Butenhoff. TL.
(2012b). Comparative pharmacokinetics of perfluorohexanesulfonate (PFHxS) in rats, mice,
and monkeys. Reprod Toxicol 33: 441-451.
http://dx.doi.Org/10.1016/i.reprotox.2011.07.004
This document is a draft for review purposes only and does not constitute Agency policy.
R-10 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Toxicological Review ofPFBA and Related Salts
U.S. EPA (U.S. Environmental Protection Agency). (1991). Guidelines for developmental toxicity risk
assessment. Fed Reg 56: 63798-63826.
U.S. EPA (U.S. Environmental Protection Agency). (1994). Methods for derivation of inhalation
reference concentrations and application of inhalation dosimetry [EPA Report],
(EPA/600/8-90/066F). Research Triangle Park, NC: U.S. Environmental Protection Agency,
Office of Research and Development, Office of Health and Environmental Assessment,
Environmental Criteria and Assessment Office.
https://cfpub.epa. gov/ncea/risk/recordisplay.cfm?deid=71993&CFID=51174829&CFTOKE
N=25006317
U.S. EPA (U.S. Environmental Protection Agency). (1996). Guidelines for reproductive toxicity risk
assessment. Fed Reg 61: 56274-56322.
U.S. EPA (U.S. Environmental Protection Agency). (1998). Assessment of thyroid follicular cell
tumors [EPA Report] (pp. 1-51). (EPA/630/R-97/002). Washington, DC: U.S. Environmental
Protection Agency, Risk Assessment Forum.
https://www.epa.gov/sites/production/files/2014-ll/documents/thyroid.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2002). A review of the reference dose and
reference concentration processes. (EPA/630/P-02/002F). Washington, DC: U.S.
Environmental Protection Agency, Risk Assessment Forum.
https://www.epa.gov/sites/production/files/2014-12/documents/rfd-final.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2005). Guidelines for carcinogen risk assessment
[EPA Report], (EPA/630/P-03/001B). Washington, DC: U.S. Environmental Protection
Agency, Risk Assessment Forum, https: //www.epa.gov/sites/production /files/2013-
09/documents/cancer guidelines final 3-25-05.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2011). Recommended use of body weight 3/4 as
the default method in derivation of the oral reference dose (pp. 1-50). (EPA/100/R-
11/0001). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum,
Office of the Science Advisor, https://www.epa.gov/sites/production/files/2013-
09/documents/recommended-use-of-bw34.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2012). Benchmark dose technical guidance.
(EPA/100/R-12/001). Washington, DC: U.S. Environmental Protection Agency, Risk
Assessment Forum, https: //www.epa.gov/risk/benchmark-dose-technical-guidance
U.S. EPA (U.S. Environmental Protection Agency). (2013). U.S. Environmental Protection Agency
endocrine disruptor screening program SAP review of EDSP tier 1 screening: Assay and
battery performance.
U.S. EPA (U.S. Environmental Protection Agency). (2016a). Health effects support document for
perfluorooctane sulfonate (PFOS). (EPA 822-R-16-002).
https://www.epa.gOv/sites/production/files/2016-05/documents/pfos hesd final 508.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2016b). Health effects support document for
perfluorooctanoic acid (PFOA). (EPA 822-R-16-003).
https://www.epa.gOv/sites/production/files/2016-05/documents/pfoa hesd final-
plain.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2018a). Chemistry Dashboard. Washington, DC.
Retrieved from https: //comptox.epa.gov/dashboard
This document is a draft for review purposes only and does not constitute Agency policy.
R-ll DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Toxicological Review ofPFBA and Related Salts
U.S. EPA (U.S. Environmental Protection Agency). (2018b). Human health toxicity values for
perfluorobutane sulfonic acid and related compound potassium perfluorobutane sulfonate
(Public comment draft). (EPA-823-R-18-307).
U.S. EPA (U.S. Environmental Protection Agency). (2019a). ChemView [Database], Retrieved from
https://chemview.epa.gov/chemview
U.S. EPA (U.S. Environmental Protection Agency). (2019b). EPA Actions to Address PFAS. Available
online at https://www.epa.gov/pfas/epa-actions-address-pfas
Van Cruchten. S: Van Den Broeck. W. (2002). Morphological and biochemical aspects of apoptosis,
oncosis and necrosis. Anatomia, Histologia, Embryologia 31: 214-223.
van Otterdiik. F. (2007a). Repeated dose 28-day oral toxicity study with MTDID-8391 by daily
gavage in the rat, followed by a 21-day recovery period. (Study Number 06-226).
Maple wood, MN: 3M.
van Otterdiik. F. (2007b). Repeated dose 90-day oral toxicity study with MTDID 8391 by daily
gavage in the rat followed by a 3-week recovery period. (Study Number 06-398).
Maple wood, MN: 3M.
Venkatesan. A: Halden. R. (2014). Loss and in situ production of perfluoroalkyl chemicals in
outdoor biosolids-soil mesocosms. Environ Res 132: 321-327.
http://dx.doi.Org/10.1016/i.envres.2014.04.024
Vulsma. T: Gons. M: de Vijlder. 1. (1989). Maternal-fetal transfer of thyroxine in congenital
hypothyroidism due to a total organification defect or thyroid agenesis. N Engl J Med 321:
13-16. http://dx.doi.org/10.1056/NETM198907063210103
Wahlang. B: Beier. 1: Clair. H: Bellis-Tones. H: Falkner. K: Mcclain. C: Cave. M. (2013). Toxicant-
associated steatohepatitis [Review], Toxicol Pathol 41: 343-360.
http://dx.doi. org/10.1177/0192623312468517
Wang. 1: Zeng. X: Bloom. M: Oian. Z: Hinvard. L: Belue. R: Lin. S: Wang. S: Tian. Y: Yang. M: Chu. C:
Gurram. N: Hu. L: Liu. K: Yang. B: Feng. D: Liu. R: Dong. G. (2019). Renal function and
isomers of perfluorooctanoate (PFOA) and perfluorooctanesulfonate (PFOS): Isomers of C8
Health Project in China. Chemosphere 218: 1042-1049.
http://dx.doi.Org/10.1016/i.chemosphere.2018.ll.191
Wang. Z: Cousins. I: Scheringer. M: Buck. R: Hungerbiihler. K. (2014). Global emission inventories
for C4-C14 perfluoroalkyl carboxylic acid (PFCA) homologues from 1951 to 2030, Part I:
production and emissions from quantifiable sources [Review], Environ Int 70: 62-75.
http://dx.doi.Org/10.1016/i.envint2014.04.013
Watkins. D: Tosson. 1: Elston. B: Bartell. S: Shin. H: Vieira. V: Savitz. D: Fletcher. T: Wellenius. G.
(2013). Exposure to perfluoroalkyl acids and markers of kidney function among children
and adolescents living near a chemical plant. Environ Health Perspect 121: 625-630.
http://dx.doi.org/10.1289/ehp.1205838
Weaver. Y: Ehresman. D: Butenhoff. 1: Hagenbuch. B. (2010). Roles of rat renal organic anion
transporters in transporting perfluorinated carboxylates with different chain lengths.
Toxicol Sci 113: 305-314.
Wiesel. T. (1982). The postnatal development of the visual cortex and the influence of environment
[Review], Biosci Rep 2: 351-377. http://dx.doi.org/10.1007/bfO 1119299
This document is a draft for review purposes only and does not constitute Agency policy.
R-12 DRAFT-DO NOT CITE OR QUOTE
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Toxicological Review ofPFBA and Related Salts
Williams. G. (2008). Neurodevelopmental and neurophysiological actions of thyroid hormone
[Review], J Neuroendocrinol 20: 784-794. http://dx.doi.org/10.ll 11/i. 1365-
2826.2008.01733.x
Williams. G: Iatropoulos. M. (2002). Alteration of liver cell function and proliferation:
Differentiation between adaptation and toxicity. Toxicol Pathol 30: 41-53.
Wolf. C: Schmid. 1: Lau. C: Abbott. B. (2012). Activation of mouse and human peroxisome
proliferator-activated receptor-alpha (PPARa) by perfluoroalkyl acids (PFAAs): further
investigation of C4-C12 compounds. Reprod Toxicol 33: 546-551.
http://dx.doi.Org/10.1016/i.reprotox.2011.09.009
Wolf. C: Takacs. M: Schmid. 1: Lau. C: Abbott. B. (2008). Activation of mouse and human peroxisome
proliferator-activated receptor alpha by perfluoroalkyl acids of different functional groups
and chain lengths. Toxicol Sci 106: 162-171. http: //dx.doi.org/10.1093 /toxsci/kfnl66
Yang. C: Glover. K: Han. X. (2009). Organic anion transporting polypeptide (Oatp) lal-mediated
perfluorooctanoate transport and evidence for a renal reabsorption mechanism of Oatplal
in renal elimination of perfluorocarboxylates in rats Toxicology Letters 190: 163-171.
https://search.proquest.com/docview/20796308?accountid=171501
Zhang. Y: Dong. X: Fan. L: Zhang. Z: Wang. 0: Tiang. N: Yang. X. (2017). Poly- and perfluorinated
compounds activate human pregnane X receptor. Toxicology 380: 23-29.
http://dx.doi.Org/10.1016/j.tox.2017.01.012
Zhao. H: Ou. B: Guan. Y: Tiang. 1: Chen. X. (2016). Influence of salinity and temperature on uptake of
perfluorinated carboxylic acids (PFCAs) by hydroponically grown wheat (Triticum aestivum
L.). SpringerPlus 5: 541. http://dx.doi.org/10.1186/s40064-016-2016-9
Zoeller. R: Rovet. 1. (2004). Timing of thyroid hormone action in the developing brain: Clinical
observations and experimental findings [Review], J Neuroendocrinol 16: 809-818.
http://dx.doi. org/10.1111 /i.l 365-2826.2004.01243.x
Zoeller. R: Tan. S: Tvl. R. (2007). General background on the hypothalamic-pituitary-thyroid (HPT)
axis [Review], Crit Rev Toxicol 37: 11-53. http: //dx.doi.org/10.1080/10408440601123446
This document is a draft for review purposes only and does not constitute Agency policy.
R-13 DRAFT-DO NOT CITE OR QUOTE
------- |