Human Health Toxicity Values for Perfluorobutane Sulfonic Acid (CASRN 375-73-5) and Related Compound Potassium Perfluorobutane Sulfonate (CASRN 29420-49-3)


c EDA
WtrM
EPA/600/R-20/345F
January 2021
Human Health Toxicity Values for
Perfluorobutane Sulfonic Acid
(CASRN 375-73-5)
and Related Compound
Potassium Perfluorobutane Sulfonate
(CASRN 29420-49-3)

-------
Human Health Toxicity Values for
Perfluorobutane Sulfonic Acid (CASRN 375-73-5) and Related Compound
Potassium Perfluorobutane Sulfonate (CASRN 29420-49-3)
Prepared by:
U.S. Environmental Protection Agency
Office of Research and Development (8101R)
Washington, DC 20460
EPA Document Number: EPA/600/R-20/345F
JANUARY 2021

-------
Disclaimer
This document is a Clearance draft for review purposes only. This information is distributed
solely for the purpose of Clearance review. 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.
i

-------
PAGE LEFT INTENTIONALLY BLANK
ii

-------
PREFACE
This assessment titled Human Health Toxicity Values for Perfluorobutane Sulfonic Acid and
Related Compound Potassium Perfluorobutane Sulfonate (PFBS) is an EPA toxicity assessment
developed in support of the Agency's PFAS Action Plan.)
The PFBS toxicity assessment is one of the key goals of the Agency's PFAS Action Plan and
provides qualitative and quantitative toxicity information that can be used along with exposure
information and other important considerations to assess potential health risks to determine if,
and when, it is appropriate to take action to address this chemical. This assessment is an update
that replaces the existing 2014 Provisional Peer Reviewed Toxicity Value (PPRTV) for PFBS
assessment used by the EPA's Superfund Program. In addition, this assessment is available for
use across multiple EPA program and regional offices, other federal agencies, states, tribes,
external stakeholders, and other entities as needed. .
The PFBS human health toxicity values presented in this assessment were developed based on
the best available science. The assessment provides high quality evaluations and conclusions
drawn from publicly available information on the toxicity of PFBS. This assessment is not a
regulation; rather, it provides a critical part of the scientific foundation for risk assessment
decision-making. Risk assessors and risk managers should carefully consider how their specific
circumstances (e.g., exposure pathways, concentrations, presence of sensitive subpopulations)
compare with the assessment's evaluation of potential hazard, the synthesis of the information,
and the uncertainties in the assessment when determining how to incorporate these toxicity
values into their specific risk characterizations.
The PFBS toxicity assessment underwent a rigorous and thorough development and review
process, as described below.
Overview of Major Steps in the PFBS Assessment Development and Review Process
•	Draft assessment development by EPA's Office of Research and Development (ORD) Center
for Public Health and Environmental Assessment (CPHEA)
•	Review by EPA Program and Regional offices (i.e., Agency review)
•	Review by other Federal Agencies (i.e., Interagency review)
•	External peer review
•	Public comment period
•	2nd External peer review
•	Agency and Interagency Review
This assessment was provided for review to scientists in EPA's program and regional offices in
the early and late stages of the assessment process. Comments were submitted by:
Office of the Administrator/Office of Children's Health Protection
Office of the Administrator/Office of Policy
Office of Chemical Safety and Pollution Prevention
Office of Land and Emergency Management
Office of Research and Development
Office of Water
i

-------
Region 2, New York, NY
Region 3, Boston, MA
Region 4, Atlanta, GA
Region 5, Chicago, IL
Region 8, Denver, CO
This assessment was provided for review to other federal agencies in the early and late stages of
the assessment process. Representatives from Federal Agencies and from the Environmental
Council of the States (ECOS) were briefed during the assessment scoping and draft development
process on March 9, 2018, May 2, 2018, and August 27, 2018. In the latter stages, this
interagency review was conducted through the oversight of the Office of Management and
Budget's PFAS Technical Working Group (TWG). Comments were submitted by:
Department of Defense
Department of Health and Human Services/Agency for Toxic Substances and Disease Registry
Food and Drug Administration
National Institute of Environmental Health Sciences/National Toxicology Program
National Institute of Occupational Safety and Health,
National Aeronautics and Space Administration
Executive Office of the President/Office of Management and Budget
This assessment was peer reviewed by independent, expert scientists external to EPA prior to
public comment, and following public comment. The reports of the two external peer reviews of
the EPA's draft Human Health Toxicity Values for PFBS, dated November 2018 and August
2020, are available at https://www.epa.gov/pfas/genx-and-pfbs-draft-toxicitv-assessments.
Comments from external peer review were submitted by:
Karen Chou, PhD
Dale Hattis, PhD
Lisa M. Kamendulis, PhD
Angela M. Leung, MD
Angela L. Slitt, PhD
David Alan Warren, MPH, PhD
R. Thomas Zoeller, PhD
This assessment was released for public comment from November 21, 2018 to January 22, 2019.
The public comments are available on Regulations.gov in the Docket ID No. EPA-HQ-OW-
2018-0614.

-------
Contents
1.0 Background	1
1.1	Physical and Chemical Properties	1
1.2	Occurrence	2
1.3	Toxicokinetics	3
1.3.1	Overview	3
1.3.2	Absorption	7
1.3.3	Distribution	7
1.3.4	Metabolism	9
1.3.5	Elimination	9
1.3.6	Physiologically Based Pharmacokinetic Models	12
1.3.7	Summary	13
2.0 Problem Formulation	15
2.1	Conceptual Model	15
2.2	Objective	17
2.3	Methods	17
2.3.1	Literature Search	17
2.3.2	Screening Process	17
2.3.3	Study Evaluati on	18
2.3.4	Data Extraction	20
2.3.5	Evidence Synthesis	21
2.3.6	Evidence Integration and Hazard Characterization	21
2.3.7	Derivation of Values	23
3.0 Overview of Evidence Identification for Synthesis and Dose-Response Analysis	26
3.1	Literature Search and Screening Results	26
3.2	Study Evaluation Results	28
4.0 Evidence Synthesis: Overview of Included Studies	31
4.1	Thyroid Effects	31
4.1.1	Human Studies	31
4.1.2	Animal Studies	32
4.2	Reproductive Effects	33
4.2.1	Human Studies	33
4.2.2	Animal Studies	33
4.3	Offspring Growth and Early Development	37
4.3.1	Human Studies	37
4.3.2	Animal Studies	37
4.4	Renal Effects	38
4.4.1	Human Studies	38
4.4.2	Animal Studies	38
iii

-------
4.5	Hepatic Effects	39
4.5.1	Human Studies	39
4.5.2	Animal Studies	39
4.6	Lipids and Lipoproteins	40
4.6.1	Human Studies	40
4.6.2	Animal Studies	41
4.7	Other Effects	42
4.7.1	Human Studies	42
4.7.2	Animal Studies	42
4.8	Other Data	43
4.8.1	Tests Evaluating Genotoxicity and Mutagenicity	46
4.8.2	Acute Duration and Other Routes of Exposure	46
5.0 Evidence Integration and Hazard Characterization	47
5.1	Thyroid Effects	50
5.2	Developmental Effects	51
5.3	Reproductive Effects	51
5.4	Renal Effects	53
5.5	Hepatic Effects	54
5.6	Effects on Lipid or Lipoprotein Homeostasis	54
5.7	Immune Effects	54
5.8	Cardiovascular Effects	55
5.9	Evidence Integration and Hazard Characterization Summary	55
6.0 Derivation of Values	65
6.1	Derivation of Oral Reference Doses	65
6.1.1	Derivation of Subchronic RfD	65
6.1.2	Derivation of the Chronic RfD	79
6.2	Derivation of Inhalation Reference Concentrations	82
6.3	Cancer Weight-of-Evidence Descriptor and Derivation of Cancer Risk Values	82
6.4	Susceptible Populations and Life Stages	82
Appendix A: Literature Search Strategy	A-l
Appendix B: Detailed PECO Criteria	A-l
Appendix C: Study Evaluation Methods	A-l
Appendix D: HAWC User Guide and Frequently Asked Questions	A-l
Appendix E. Additional Data Figures	A-l
Appendix F. Benchmark Dose Modeling Results	A-l
Appendix G. Quality Assurance	A-16
Appendix H. References	A-18
iv

-------
Figures
Figure 1. Chemical structures ofPFBS andK+PFBS 1
Figure 2. Conceptual model forPFBS and/or potassium salt 16
Figure 3. Approach for evaluating epidemiological and animal toxicology studies 20
Figure 4. Literature search and screening flow diagram for PFBS (CASRN 375-73-5) 27
Figure 5. Evaluation results for epidemiological studies assessing effects ofPFBS (click to
see interactive data graphic for rating rationales) 29
Figure 6. Evaluation results for animal studies assessing effects ofPFBS exposure (click to
see interactive data graphic for rating rationales) 30
Figure 7. Thyroid effects from K+PFBS exposure (click to see interactive data graphic and
rationale for study evaluations for effects on the thyroid in HAWC) 32
Figure 8. Reproductive hormone response to K+PFBS exposure (click to see interactive
data graphic and rationale for study evaluations for reproductive hormone levels
in HAWC) 35
Figure 9. Effects to reproductive development and estrous cycling following PFBS
exposure (click to see interactive data graphic) 36
Figure D-l. HAWC homepage for the public PFBS assessment A-l
Figure D-2. Representative study list A-2
Figure D-3. Representative study evaluation pie chart with the reporting domain selected
and text populating to the right of pie chart A-3
Figure D-4A. Visualization example for PFBS. (Note that the records listed under each
column (study, experiment endpoint, units, study design, observation time, dose)
and data within the plot are interactive.) A-4
Figure D-4B. Example pop-up window after clicking on interactive visualization links. (In
Figure D-4A the red circle for study NTP (2019); male at a dose of 500 mg/kg-
day was clicked leading to the pop-up shown above. Clicking on blue text will
open a new window with descriptive data.) A-4
Figure D-5. Representative data download page A-5
Figure D-6A. Example BMD modeling navigation A-6
Figure D-6B. Example BMD session A-6
Figure E-l. Serum free and total thyroxine (T4) response in animals following K+PFBS
exposure (click to see interactive data graphic) A-l
Figure E-2. Serum total triiodothyronine (T3) response in animals following K+PFBS
exposure (click to see interactive data graphic) A-2
Figure E-3. Serum thyroid-stimulating hormone (TSH) response in animals following
K+PFBS exposure (click to see interactive data graphic) A-3
v

-------
Figure E-4. Developmental effects (eye opening) following K+PFBS in rats (click to see
interactive data graphic)	A-3
Figure E-5. Developmental effects (first estrus) following K+PFBS in rats (click to see
interactive data graphic)	A-4
Figure E-6. Developmental effects (vaginal patency) following K+PFBS in rats (click to
see interactive data graphic)	A-4
Figure E-7. Kidney histopathological effects following K+PFBS in rats (click to see
interactive data graphic)	A-5
Figure E-8. Renal effects following K+PFBS in rats (click to see interactive data graphic)	A-6
Figure E-9. Kidney weight effects following K+PFBS in rats (click to see interactive data
graphic)	A-7
Figure E-10. Liver effects following K+PFBS in rats (click to see interactive data graphic)	A-8
Figure E-l 1. Effects on lipids and lipoproteins following K+PFBS in rats and mice (click to
see interactive data graphic)	A-9
Figure F-l. Candidate PODs for the derivation of the subchronic and chronic RfDs for
PFBS (click to see interactive data graphic)	A-3
Figure F-2. Exponential (Model 4) for total T4 in PND 1 female offspring (litter n)
exposed GDs 1-20 (Feng et al. (2017)	A-3
Tables
Table 1. Physicochemical properties of PFBS (CASRN 375-73-5) and related compound
K+PFBS (CASRN 29420-49-3)	2
Table 2. Summary of toxicokinetics of serum PFBS (mean ± standard error)	5
Table 3. Criteria for overall evidence integration judgments	22
Table 4. Epidemiological studies excluded based on study evaluation	28
Table 5. Other studies	44
Table 6. Summary of noncancer data for oral exposure to PFBS (CASRN 375-73-5) and
the related compound K+PFBS (CASRN 29420-49-3)	48
Table 7. Summary of hazard characterization and evidence integration judgments	56
Table 8. Mouse, Rat, and Human half-lives and data-informed dosimetric adjustment
factors	70
Table 9. PODs considered for the derivation of the subchronic RfD for K+PFBS
(CASRN 29420-49-3)	72
Table 10. UFs for the subchronic RfD for thyroid effects for K+PFBS
(CASRN 29420-49-3)	77
vi

-------
Table 11. Confidence descriptors for the subchronic RfD for PFBS (CASRN 375-73-5)
and the related compound K PFBS (CASRN 29420-49-3)	78
Table 12. UFs for the chronic RfD for thyroid for K+PFBS (CASRN 29420-49-3)	80
Table 13. Confidence descriptors for chronic RfD for PFBS (CASRN 375-73-5) and the
related compound K PFBS (CASRN 29420-49-3)	81
Table A-l. Synonyms and MESH terms	A-l
Table B-l. Population, exposure, comparator, and outcome criteria	A-l
Table C-l. Questions used to guide the development of criteria for each domain in
epidemiology studies	A-2
Table C-2. Criteria for evaluation of exposure measurement in epidemiology studies	A-5
Table C-3. Questions used to guide the development of criteria for each domain in
experimental animal toxicology studies	A-7
Table F-l. Candidate PODs for the derivation of the subchronic and chronic RfDs for
PFBS (CASRN 375-73-5) and the related compound K PFBS
(CASRN 29420-49-3)	A-l
Table F-2. Modeling results for total T4 in PND 1 female offspring (litter n) exposed
GDs 1-20 a	A-2
vii

-------
Commonly Used Abbreviations
AEC	absolute eosinophil count
AFFF	Aqueous Film-Forming Foam
AIC	Akaike's information criterion
ALT	alanine aminotransferase
AST	aspartate aminotransferase
AUC	area under the curve
BMD	benchmark dose
BMDL	benchmark dose lower confidence
limit
BMDS	Benchmark Dose Software
BMR	benchmark response
BUN	blood urea nitrogen
BW	body weight
CA	chromosomal aberration
CASRN	Chemical Abstracts Service
Registry Number
CHO	Chinese hamster ovary (cell line
cells)
CI	confidence interval
CPHEA	Center for Public Health and
Environmental Assessment
CPN	chronic progressive nephropathy
D3	deiodinase 3
DAF	dosimetric adjustment factor
DNA	deoxyribonucleic acid
ECP	eosinophilic cationic protein
EPA	U.S. Environmental Protection
Agency
GD	gestation day
GLP	good laboratory practices
HAWC	Health Assessment Workspace
Collaborative
HED	human equivalent dose
HPT	hypothalamic-pituitary-thyroid
i.v.	intravenous
ICR	Institute of Cancer Research
K+PFBS potassium perfluorobutane sulfonate
keiim	serum elimination rate constant
LD	lactation day
LD50	median lethal dose
LOAEL	lowest observed adverse effect level
Acronyms
NCEA	National Center for Environmental
Assessment
NHANES National Health and Nutrition
Examination Survey
NOAEL	no observed adverse effect level
NTP	National Toxicology Program
NZW	New Zealand White (rabbit breed)
OR	odds ratio
PECO	population, exposure, comparator,
outcome
PFAA	perfluoroalkyl acid
PFAS	per- and polyfluoroalkyl substances
PFOA	perfluorooctanoic acid
PFOS	perfluorooctane sulfonic acid
PFBS	perfluorobutane sulfonic acid
PFHxA	perfluorohexanoic acid
PND	postnatal day
POD	point of departure
RfC	inhalation reference concentration
RfD	oral reference dose
ROS	reactive oxygen species
rT3	reverse triiodothyronine
S-D	Sprague Dawley
SD	standard deviation
T2	3,5-diiodo-L-thyronine
T3	triiodothyronine
T4	thyroxine
TBG	thyroid binding globulin
TSH	thyroid-stimulating hormone
TTR	transthyretin
UF	uncertainty factor
UFa	interspecies uncertainty factor
UFc	composite uncertainty factor
UFd	database uncertainty factor
UFh	intraspecies uncertainty factor
UFl	LOAEL-to-NOAEL uncertainty factor
UFs	subchronic-to-chronic uncertainty
factor
VLDL	very low density lipoprotein
viii

-------
molecular weight

-------
X

-------
Executive Summary
Summary of Occurrence and Health Effects
The U.S. Environmental Protection Agency (EPA) is issuing draft subchronic and chronic oral
toxicity values for perfluorobutane sulfonic acid (PFBS) (Chemical Abstracts Service Registry
Number [CASRN] 375-73-5) and its related salt, potassium perfluorobutane sulfonate (K+PFBS)
(CASRN 29420-49-3). The ionic state of PFAS such as PFBS influence physicochemical
properties such as water or lipid solubility and bioaccumulative potential, which in turn impact
fate and transport in the environment and potential human health and ecological effects in
exposed populations. K+PFBS fully dissociates in aqueous solutions of pH ranging from 4-9, as
such, the oral toxicity values derived in this document are also applicable to the deprotonated
anionic form of PFBS (i.e., PFBS"; CASRN 45187-15-3).
The toxicity assessment for PFBS is a scientific and technical report that includes toxicity values
associated with potential noncancer health effects following oral exposure (in this case, oral
reference doses [RfDs]). This assessment evaluates human health hazards. The toxicity
assessment and the values contained within is not a risk assessment as it does not include an
exposure assessment nor an overall risk characterization. Further, the toxicity assessment does
not address the legal, political, social, economic, or technical considerations involved in risk
management. When final, the PFBS toxicity assessment can be used by EPA, states, tribes, and
local communities, along with specific exposure and other relevant information, to determine,
under various appropriate regulations and statutes, if, and when, it is necessary to take action to
address potential risk associated with human exposures to PFBS.
PFBS and K+PFBS are both four-carbon, fully fluorinated alkane members of a large and diverse
class of linear and branched compounds known as "per- and polyfluoroalkyl substances," or
PFAS. In the early 2000s, concerns grew over the environmental persistence, and long half-lives
in humans and bioaccumulation potential of longer chain PFAS, in particular, perfluorooctanoic
acid (PFOA) and perfluorooctane sulfonic acid (PFOS). As a result, shorter chain PFAS such as
PFBS were developed and integrated into various consumer products and applications, as this
compound has the desired properties and characteristics associated with this class of compounds
with faster elimination from the body than PFOA and PFOS. PFBS has been found in food
contact materials, dust, and source and finished drinking water. It is also associated with
Aqueous Film-Forming Foams and used during chrome electroplating as a mist suppressant. As
such, oral intake of water and food, inhalation, and dermal contact are plausible modes of PFBS
exposure, with the oral route being the primary route of exposure. PFBS has been detected in
human urine, confirming exposure to this PFAS; however, the magnitude of human exposure
likely depends on factors such as occupation (e.g., processing and/or manufacture of PFBS or
PFBS-containing products and chrome electroplating) and living conditions (e.g., proximity to
locations that make or use PFBS-containing products and well-water use).
Human studies have examined possible associations between PFBS exposure and potential
health outcomes such as alteration of menstruation, reproductive hormones or semen parameters,
kidney function (uric acid production), lung function (induction of asthma), and lipid profile. The
ability to draw conclusions about associations was limited due to the small number of human
studies per outcome. Of the examined health outcomes, only asthma and serum cholesterol levels
in humans were found to exhibit a statistically significant positive association with PFBS
xi

-------
exposure. No studies have been identified that evaluate the association between PFBS exposure
and potential cancer outcomes. While the epidemiology studies were not influential to drawing
evidence integration judgments or the derivation of toxicity values, the general findings identify
potential areas of future research
Animal studies of repeat-dose PFBS exposure have been exclusively via the oral route, used the
potassium salt of PFBS (K+PFBS) as the source exposure material, and have examined
noncancer effects only. The available rat and mouse studies support identification of thyroid,
developmental, and kidney endpoints as potential health effects following repeated exposures in
utero and/or during adulthood. Animal studies also evaluated other health outcomes such as
liver, reproductive parameters, lipid/lipoprotein homeostasis, spleen, and hematology; however,
the available evidence does not support a clear association with PFBS exposure.
Noncancer Effects Observed Following Oral Exposure
Oral exposures to PFBS or its K+ salt in adult and developing rats and mice have been shown to
result in thyroid, developmental, and kidney effects. Thyroid effects in adult exposed rats and
mice and in developing mice were primarily expressed through significant decreases in
circulating levels of hormones such as thyroxine (T4) and triiodothyronine (T3). In early
developmental life stages in mice (e.g., newborn), decreases in thyroid hormone were
accompanied by other effects indicative of delayed maturation or reproductive development
(e.g., vaginal patency and eyes opening). Kidney weight and/or histopathological alterations
(e.g., renal tubular and ductal epithelial hyperplasia) were observed in rats following short-term
and subchronic oral exposures. Many of the kidney effects, however, occurred at higher doses
than did the thyroid and developmental effects. The limited number of human studies examining
oral PFBS exposure does not inform the potential for effects in thyroid, developing offspring, or
the renal system.
Oral Reference Doses for Noncancer Effects
Subchronic1 and chronic2 oral RfDs were derived for PFBS. The hazards of potential concern
include thyroid, developmental, and kidney effects. From these identified targets of PFBS
toxicity, perturbation of thyroid hormone levels (e.g., thyroxine [T4]) was used as the critical
effect for derivation of a subchronic and chronic RfD. Based on recommendations in the EPA's
Recommended Use of Body Weight4 as the Default Method in Derivation of the Oral Reference
Dose (U.S. EPA. 2011b). chemical specific toxicokinetic data (e.g., serum half-lives) were used
to scale a toxicologically equivalent dose of orally administered PFBS from animals to humans.
Following the EPA's Benchmark Dose Technical Guidance Document (U.S. EPA. 2012).
benchmark dose (BMD) modeling of thyroid effects in a developmental life stage following
exposure to K+PFBS in utero resulted in a BMDLo.ssd human equivalent dose (FLED) of 0.16
1 Subchronic Exposure: Repeated exposure by the oral, dermal, or inhalation route for more than 30 days, up to
approximately 10% of the life span in humans (more than 30 days up to approximately 90 days in typically used
laboratory animal species).
2 Chronic Exposure: Repeated exposure by the oral, dermal, or inhalation route for more than approximately 10% of
the life span in humans (more than approximately 90 days to 2 years in typically used laboratory animal species).
(https://ofmpub.epa.gov/sor internet/registrv/termreg/searchandretrieve/glossariesandkevwordlists/search.do?details
=&glossarvName=IRIS%20Glossarv#formTop)
xii

-------
milligrams per kilogram per day (mg/kg-day). This HED associated with thyroid effects served
as the point of departure (POD) for derivation of the subchronic and chronic RfDs.
In the process of developing the subchronic and chronic RfDs, scientific rationales were
provided for assigning a value for the database uncertainty factor (UFD)of 1 and of 3. Each
argument was considered by EPA to have merit. Therefore, EPA has presented RfDs for K+
PFBS and for PFBS (free acid) derived using both an UFd of 1 or an UFd of 3. Risk assessors
may evaluate the justifications for application of either UFd and decide whether the risk scenario
under consideration warrants use of the higher or lower RfD considering the purpose and scope
of their risk assessment and the decision-making it supports, i.e., which is fit-for-purpose of the
specific risk assessment.
The lower subchronic RfD for K+PFBS was calculated by dividing the PODhed for decreased
serum total T4 observed in newborn (PND 1) mice, conducted by Feng et al. (2017). by a
composite uncertainty factor (UFc) of 100 to account for extrapolation from mice to humans (an
interspecies UF, or UFa, of 3), for interindividual differences in human susceptibility
(intraspecies UF, or UFh, of 10), and for deficiencies in the toxicity database (database UF, or
UFd, of 3) (a value of 1 was applied for sub chronic-to-chronic UF, or UFs, and LOAEL-to-
NOAEL UF, or UFl) (see Table 10), yielding a subchronic RfD of 0.0016 mg/kg-day rounded to
2 x 10 3 mg/kg-day. As K+PFBS is fully dissociated in water at the environmental pH range of
4-9 to the PFBS anion (PFBS") and the K+ cation, data for K+PFBS were used to derive a
subchronic RfD for the free acid (PFBS) by adjusting for differences in molecular weight (MW)
between K+PFBS (338.19) and PFBS (300.10), yielding the value of 0.0014 mg/kg-day rounded
to 1 x 10 3 mg/kg-day for a subchronic RfD for PFBS (free acid). The higher subchronic RfD
for K+PFBS and PFBS (free acid) was calculated in the same way with the exception of using an
UFd of 1.
The lower chronic RfD for K+PFBS associated with thyroid effects was calculated by dividing
the PODhed for decreased serum total T4 observed in newborn (PND 1) mice, conducted by
Feng et al. (2017). by a UFc of 300 to account for extrapolation from mice to humans (UFa of 3),
for interindividual differences in human susceptibility (UFh of 10), and deficiencies in the
toxicity database (UFd of 10) (a value of 1 was applied for UFs and UFl) (see Table 12),
yielding a chronic RfD of 0.00053 mg/kg-day rounded to 5 x 10 4 mg/kg-day. Like the
subchronic RfD for thyroid effect, based on the data for K+PFBS, a chronic RfD for PFBS (free
acid) of 0.00047 mg/kg-day rounded to 5 x 10 4 mg/kg-day was derived. The higher chronic RfD
for K+PFBS and PFBS (free acid) was calculated in the same way with the exception of using an
UFd of 1.
Confidence in the Oral RfDs
The overall confidence in the subchronic RfD for thyroid effects is medium. The gestational
exposure study conducted by Feng et al. (2017) reports administration of K+PFBS by gavage in
pregnant Institute of Cancer Research (ICR) mice (10/dose) from gestation days (GDs) 1 to 20.
This study was of good quality (i.e., high confidence) with adequate reporting and consideration
of appropriate study design, methods, and conduct (click to see risk of bias analysis in HAWC3).
3 HAWC: A Modular Web-Based Interface to Facilitate Development of Human Health Assessments of Chemicals;
see Appendix D for details.
xiii

-------
Confidence in the oral toxicity database for derivation of the subchronic RfD is medium because,
although there are multiple short-term studies and a subchronic-duration toxicity study in
laboratory animals, a two-generation reproductive toxicity study in rats (Lieder et al.. 2009b).
and multiple developmental toxicity studies in mice and rats, there are no PFBS studies available
that have specifically evaluated health effect domains of emerging concern across the PFAS
class such as immunotoxicity and mammary gland development (Dewitt et al.. 2012; White et
al., 2007). Further, neurodevelopmental effects are of particular concern when perturbations in
thyroid hormone occur during a sensitive early life stage, and the absence of a study evaluating
neurodevelopmental effects following PFBS exposure is a source of uncertainty in the
assessment.
The overall confidence in the chronic RfD for thyroid effects is low. While the RfD was derived
using the same high-confidence principal study conducted by Feng et al. (2017) that was used for
the subchronic RfD, there is increased concern pertaining to the potential for identification of
hazards following longer (i.e., chronic) duration PFBS exposures. Thus, due to the lack of
studies that specifically evaluated health effect domains of emerging concern across the PFAS
class such as immunotoxicity, mammary gland development, or neurodevelopmental at any
exposure duration but particularly for chronic duration, confidence in the database specifically
for a chronic RfD is low.
Effects other than Cancer Observed Following Inhalation Exposure
There are no studies available that examine toxicity in humans or experimental animals
following inhalation exposure, precluding the derivation of an inhalation reference concentration
(RfC).
Evidence for Carcinogenicity
Under the EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005), the Agency
concluded that there is "inadequate evidence to assess carcinogenic potential" for PFBS and
K+PFBS by either oral or inhalation routes of exposure. Therefore, the lack of data on the
carcinogenicity of PFBS and the related compound K+PFBS precludes the derivation of
quantitative estimates for either oral (oral slope factor) or inhalation (inhalation unit risk)
exposure.
xiv

-------
1.0 Background
1.1 Physical and Chemical Properties
Perfluorobutane sulfonic acid (PFBS) (Chemical Abstracts Service Registry Number
[CASRN] 375-73-5)4 and its related salt, potassium perfluorobutane sulfonate (K+PFBS)
(CASRN 29420-49-3), are members of the group of per- and polyfluoroalkyl substances (PFAS),
more specifically the short-chain perfluoroalkane sulfonates. For purposes of this assessment,
"PFBS" will signify the ion, acid, or any salt of PFBS. Concerns about PFBS and other PFAS
stem from the resistance of these compounds to hydrolysis, photolysis, and biodegradation,
which leads to their persistence in the environment (Sundstrom et al., 2012). The chemical
formula of PFBS is C4HF9O3S and the chemical formula of K+PFBS is C4F9KO3S. Their
respective chemical structures are presented in Figure 1. K+PFBS differs from PFBS by being
associated with a potassium ion. The reported water solubility of each species suggests that in
aqueous environments, the sulfonate would be the predominant form. The preferential use of
K+PFBS in laboratory studies is related to the optimal dissociation of the salt to the sulfonate
(i.e., PFBS") at pH ranging from 4 to 9 (see Table 1). Table 1 provides a list of physicochemical
properties for PFBS and K+PFBS.
K*
PFBS
K+PFBS
Figure 1. Chemical structures of PFBS and K+PFBS.
4 The CASRN given is for linear PFBS; the source PFBS used in toxicity studies was assayed at >98% linear, suggesting some
minor proportion of other chemicals, such as branched PFBS isomers, are present. Thus, observed health effects may apply to the
total linear and branched isomers in a given exposure source.
1

-------
Table 1. Physicochemical properties of PFBS (CASRN 375-73-5) and related compound
K+PFBS (CASRN 29420-49-3)
Property (unit)
Value*
PFBS (free acid)3
K+PFBS (potassium salt)b
Boiling point (°C)
152
447
Density (g/cm3)
1.83 (predicted)
1.83 (predicted)
Vapor pressure (mm Hg)
0.104 (predicted)
1.12 x l(T8
pH
ND
ND
Solubility in water (mol/L)
0.0017
0.08
Molecular weight (g/mol)
300.09
338.18
Dissociation constant
NA
Fully dissociated in water over the pH range of 4-9
Sources'.
'Values are experimentally determined unless otherwise indicated
aU.S. EPA Chemistry Dashboard for CASRN 375-73-5.
bU.S. EPA Chemistry Dashboard for CASRN 29420-49-3.
Notes'. °C = degrees Celsius; g/cm3 = grams per cubic centimeter; g/mol = grams per mole; mm HG = millimeters of mercury;
mol/L = moles per liter; NA = not applicable; ND = no data.
1.2 Occurrence
PFBS-based compounds are surfactants used primarily in the manufacture of paints, cleaning
agents, and water- and stain-repellent products and coatings. They serve as replacements for
perfluorooctane sulfonic acid (PFOS) (3M. 2002b). Various sources report detection or
occurrence in environmental media and consumer products, including drinking water, ambient
water, dust, carpeting and carpet cleaners, floor wax, and food packaging.
Oral exposure via drinking water might be expected in areas where contamination has been
reported. EPA Unregulated Contaminant Monitoring Rule data for public drinking water utilities
in 2013-2015 showed levels of PFBS above the Minimum Reporting Level (> 0.09 micrograms
per liter [|ig/L]) in water systems serving Alabama, Colorado, Georgia, the Northern Mariana
Islands, and Pennsylvania (U.S. EPA. 2017; Hu et al.. 2016). These utilities included both
ground and surface drinking water sources, with concentrations ranging from 0.09 to 0.37 |ig/L.
The estimated combined number of people served by these water systems is more than 340,000
(U.S. EPA. 2018).
Measurements from 37 surface water bodies in the northeastern United States (metropolitan New
York area and Rhode Island) collected in 2014 showed an 85% site detection rate (Zhang et al„
2016). PFBS has also been identified in surface waters in Georgia, New Jersey, North Carolina,
and the Upper Mississippi River Basin (Post et al.. 2013; Lasier et al.. 2011; Nakavama et al..
2010; Nakavama et al.. 2007). It has been detected in wastewater treatment plant effluent,
seawater, soil, and biosolids (Houtz et al.. 2016; Zhao et al.. 2012; Sepulvado et al.. 2011).
PFBS contamination, which has been associated with the use of Aqueous Film-Forming Foams
(AFFFs) (ESTCP. 2017; Anderson et al.. 2016). was reported at Superfund sites and areas under
assessment for Superfund designation. Contaminated sites include the former Wurtsmith Air
Force Base, Ellsworth Air Force Base, and Dover Air Force Base (Aerostar SES LLC, 2017;
2

-------
Anonymous, 2017; ASTSWMO, 2015). At the Wurtsmith site, PFBS was detected at a
concentration of 6.4 |ig/L in ground water contaminated by a PFAS plume originating from the fire
training area (ASTSWMO. 2015). It is also present in some drinking water samples from nearby
residential wells at low nanograms per liter concentrations, which were below the screening value
cited by the Michigan Department of Community Health (MDCH. 2015). Other sources of PFAS
and/or PFBS contamination include chrome plating operations, PFAS manufacture, and sites that
use PFAS in product formulations such as textile and electronic industries.
PFBS has also been detected in household dust and consumer products. There was a 92%
detection frequency for PFBS among 39 household dust samples (10 from the United States)
analyzed with levels ranging from 86 nanograms per gram (ng/g) for the 25th percentile to
782 ng/g for the 75th percentile (Kato et al.. 2009). In a separate study, PFBS dust levels were
measured in Boston area offices (n = 31), homes (// = 30), and vehicles (// = 13) with detection
frequencies being relatively low—10%, 3%, and 0%, respectively—and ranging in the low parts
per billion (Fraser et al.. 2013). Consumer products could also be an exposure source. Limited
quantitative testing showed the presence of PFBS in carpet and upholstery protectors (45.8 and
89.6 ng/g), carpet shampoo (25.7 and 911 ng/g), textiles (2 ng/g), and floor wax (143 ng/g)
purchased in the United States (Liu et al.. 2014).
PFBS was detected in fast food packaging (7/20 samples) in one U.S. study (Schaider et al..
2017) although the magnitude of the detection was not reported.
The European Food Safety Authority reported the presence of PFBS in various food and drink
items, including fruits, vegetables, cheese, and bottled water. For average adult consumers, the
estimated exposure ranges for PFBS were 0.03-1.89 nanograms per kilogram per day
(ng/kg-day) (minimum) to 0.10-3.72 ng/kg-day (maximum) (EFSA. 2012).
PFBS has been reported in serum of humans in the general population. In American Red Cross
samples collected in 2015, 8.4% had a quantifiable serum PFBS concentration; the majority of
samples were below the lower limit of quantitation (4.2 nanograms per milliliter [ng/mL]) (01 sen
et al„ 2017). The National Health and Nutrition Examination Survey (NHANES) included PFBS
in consecutive biomonitoring cycles, including 2013-2014 where the 95th percentile reported for
PFBS was at or below the level of detection (0.1 ng/mL). Considering the relatively rapid rate of
elimination of PFBS (days to weeks), compared to longer chain PFAS (years), the lack of
biomonitoring detects (e.g., NHANES 2013-2014 cycle) should not be interpreted as a lack of
occurrence or exposure potential. Another study with a lower limit of detection (0.013 ng/g)
reported increasing levels of PFBS in serum from primiparous nursing women in Sweden from
1996 to 2010 (Glynn et al.. 2012).
1.3 Toxicokinetics
1.3.1 Overview
Animal evidence has shown that PFBS, like other PFAS, is well absorbed following oral
administration. PFBS distributes to all tissues of the body (Bogdanska et al.. 2014). but a study
evaluating the volume of distribution concluded that distribution is predominantly extracellular
(Olsen et al.. 2009). Because of its resistance to metabolic degradation, PFBS is primarily
eliminated unchanged in urine and feces.
3

-------
Three sets of investigators have conducted toxicokinetic studies in rats and monkeys (Huang et
al.. 2019a; Chengelis et al.. 2009; Olsen et al.. 2009). 01sen et al. (2009) and Xu et al. (2020)
have measured the half-life of PFBS in humans. Bogdanska et al. (2014) and Lau et al. (2020)
have reported limited toxicokinetic information in mice. One study developed a physiologically
based pharmacokinetic (PBPK) model that includes parameterization for PFBS (Fabrega et al..
2015).
Results of all studies discussed in this section are summarized in Table 2.
4

-------
Table 2. Summary of toxicokinetics of serum PFBS (mean ± standard error)
Species/Sex
Study design
Elimination
half-life (hr)
AUC
(jig-hr/mL)
Clearance
Volume of
distribution (L/kg)
Reference
Mice
Mice/male
Single oral dose (30 mg/kg)
3.7
1515
0.019 (L/hr-kg)
0.129
Lau et al. (2020)
Single oral dose (300
mg/kg)
6.0
7178
0.039 (L/hr-kg)
0.291
Lau et al. (2020)
Single oral dose (combined
30/300 mg/kg)
5.8

0.038 (L/hr-kg)
0.275
Lau et al. (2020)
Mice/female
Single oral dose (30 mg/kg)
4.4
520
0.056 (L/hr-kg)
0.145
Lau et al. (2020)
Single oral dose (300
mg/kg)
4.6
4587
0.064 (L/hr-kg)
0.308
Lau et al. (2020)
Single oral dose (combined
30/300 mg/kg)
4.5

0.063 (L/hr-kg)
0.278
Lau et al. (2020)
Rats
Rats/male
Single i.v. dose (10 mg/kg)
2.1
254
0.0394 (L/hr-kg)
0.118
Cheneelis et al. (2009)
Single i.v. dose (30 mg/kg)
4.51 ± 2.22°
294 ± 77
119 ± 34 (L/hr)a
0.330 ±0.032
Olsen et al. (2009)
Single oral dose (30 mg/kg)
4.68 ± 0.43°
163 ± 10
NA
0.676 ± 0.055
Olsen et al. (2009)
Single i.v. dose (4 mg/kg)
4.22 ± 0.28d
116 ± 7
0.0345 ± 0.002 (L/hr-kg)
0.188 ±0.017d
Huane et al. (2019a)
Single oral dose (4 mg/kg)
4.89 ± 1.67d
154 ± 15
0.0265 ± 0.003 (L/hr-kg)
0.174 ±0.614d
Huane et al. (2019a)
Single oral dose (20 mg/kg)
5.36 ± 1.24d
533 ±45
0.0376 ± 0.003 (L/hr-kg)
0.167 ±0.039d
Huane et al. (2019a)
Single oral dose
(100 mg/kg)
5.25 ± 1.19d
1320 ± 100
0.0755 ± 0.006 (L/hr-kg)
0.335 ±0.041d
Huane et al. (2019a)

Rats/female
Single i.v. dose (10 mg/kg)
0.64
32
0.311 (L/hr-kg)
0.288
Cheneelis et al. (2009)
Single i.v. dose (30 mg/kg)
3.96 ± 0.21°
65 ±5
469 ± 40 (L/hr)b
0.351± 0.034
Olsen et al. (2009)
Single oral dose (30 mg/kg)
7.42 ± 0.79°
85 ± 12
NA
0.391± 0.105
Olsen et al. (2009)
Single i.v. dose (4 mg/kg)
0.95 ± 0.10d
16 ± 1
0.252 ±0.018 (L/hr-kg)
0.165 ±0.015d
Huane et al. (2019a)
Single oral dose (4 mg/kg)
1.50 ± 0.10d
29 ±3
0.152 ±0.020 (L/hr-kg)
0.328 ± 0.042d
Huane et al. (2019a)
Single oral dose (20 mg/kg)
1.23 ± 0.12d
109 ±23
0.183 ±0.039 (L/hr-kg)
0.326 ± 0.073d
Huane et al. (2019a)
5

-------
Species/Sex
Study design
Elimination
half-life (hr)
AUC
(jig-hr/mL)
Clearance
Volume of
distribution (L/kg)
Reference

Single oral dose
(100 mg/kg)
l.ll±0.10d
387 ±50
0.259 ± 0.033 (L/hr-kg)
0.415 ± 0.063d
Huang et al. (2019a)
Monkeysb
Cynomolgus
macaque/male
Single i.v. dose (10 mg/kg)
15 (9.65)e
1,115 ±859
0.016 (L/hr-kg)
0.209 ± 0.028
Chengelis et al. (2009)
Single i.v. dose (10 mg/kg)
95.2 ±27.1
24.3 ± 8.6
511 ± 141 (mL/hr)
0.254 ±0.031
Olsen et al. (2009)
Cynomolgus
macaque/female
Single i.v. dose (10 mg/kg)
8.1
489 ±180
0.0229 ± 0.0099 (L/hr-kg)
0.248 ± 0.045
Chengelis et al. (2009)
Single i.v. dose (10 mg/kg)
83.2 ±41.9
35.4 ± 13.3
368 ± 120 (mL/hr)
0.255 ±0.017
Olsen et al. (2009)
Humans
Males and female
Occupational (n=6)
619.2f
NA
NA
NA
Olsen et al. (2009)
Males
Occupational (n=5)
552f
NA
NA
NA
Olsen et al. (2009)
Female
Occupational (n=l)
1,096.8
NA
NA
NA
Olsen et al. (2009)
Males and females
Occupational (n=26)
1,056
NA
NA
NA
Xu et al. (2020)
Notes'. AUC = area under the curve; hr = hour; i.v. = intravenous; L/hr-kg = liters per hour per kilogram; L/kg = liter per kilogram; mL/hr = milliliters per hour; (ig-hr/mL =
micrograms per hour per milliliter; NA = not available.
aBody weights were reported to be 0.200-0.250 kg (approximately 476 L/kg-hour).
bThe data were monitored 48 hours and 31 days postdosing for Chengelis et al. (2009) and Olsen et al. (20091 respectively.
cQlsen et al. (2009) reported T0.501 and T0.5P in rats, presenting data for T0.5P
dHuang et al. (2019a) reported T0.501, T0.5P, and To.skio in male rats (both oral and i.v.) and female rats (i.v. only); only To.skio was reported in female rats (oral). Presenting data
for T0.5P for male rats (both oral and i.v.) and female rats (i.v.) and To.skio for female rats (oral). The volume of distribution (Vd) was calculated as the sum of volume terms of the
central compartment and that of the peripheral compartment except for orally-exposed female rats. The volume of peripheral compartment was not reported for orally-exposed
female rats, representing the volume of central compartment only.
eOne male monkey had a serum concentration more than tenfold higher than the others at 48 hours postdosing with an estimated half-life of 26 hours.
fQlsen et al. (2009) reported mean and geometric mean values for males only and all subjects, presenting data for geometric mean values.
6

-------
1.3.2 Absorption
Olsen et al. (2009) conducted intravenous (i.v.) and oral uptake studies in rats (n=3/sex) that
were given a single oral dose (30 milligrams per kilogram [mg/kg]) of potassium PFBS
(K+PFBS). The serum area under the concentration curve (AUC) after i.v. administration was
294 ± 77 and 65 ± 5 (|ig-h/mL) in male and female rats, respectively, and 163 ± 10 and 85 ± 12
in males and females, respectively, after oral dosing. The large variance in AUC for male rats
after i.v. dosing and greater AUC after oral dosing compared to i.v. dosing in females makes it
difficult to interpret these results with certainty, but it appears that PFBS is 100% bioavailable in
female rats, while the nominal bioavailability in male rats is only 55% based on AUC. Peak
concentrations (Cmax) occurred at 0.3-0.4 hours after oral dosing, showing that absorption was
fairly rapid. Bioavailability based on Cmax is 60% in male rats and 85% in female rats, suggesting
a similar sex difference as estimated from AUC.
The findings are generally confirmed in a recent paper by Huang et al. (2019a). It was found that
absorption of PFBS usually occurred within 24 h, along with the time reaching the maximal
plasma concentration (Tmax) under 2.4 h in male rats and under 1.4 h in female rats, following
single dose of gavage administration in Hsd:Sprague Dawley SD rats (4, 20, 100 mg/kg of
K+PFBS). However, bioavailability calculated based on the AUC after 4 mg/kg i.v. and oral
doses reported by Huang et al. (2019a) is 75% in males and 60% in females, and based on Cmax
respective values of 45% and 27% in males and females are obtained, qualitatively the opposite
of results from Olsen et al. (2009).
Given the range of estimated bioavailability from the results of Olsen et al. (2009) and Huang et
al. (2019a). a sex difference in this parameter for rats cannot be determined. Averaging the AUC-
based values for both males and females from the two studies yields an overall average of 73%.
Notably, Huang et al. (2019a) also observed that, the dose-adjusted AUC decreased with
increasing doses for both males and females. However, this result could occur because of
saturation of renal resorption at higher doses, rather than a reduction in absorption.
Similar observations indicating rapid absorption of PFBS have been reported for CD-I mice
orally exposed to PFBS at 30 or 300 mg/kg, where Tmax was estimated between 1 to 2 hours after
oral gavage (Lau et al.. 2020).
1.3.3 Distribution
PFBS has been shown to distribute to tissues within 24 hours of exposure with liver and kidney
being the organs with highest distribution.
Lau et al. (2020) evaluated the pharmacokinetic properties of PFBS in CD-I mice at 8 weeks of
age. Male and female mice were given a single dose of 0, 30, or 300 mg/kg body weight PFBS
via gavage. Liver and kidney were harvested 24 hours postdosing. PFBS distributed to both
organs readily in a dose-dependent manner but did not accumulate in either liver or kidney. Lau
et al. (2020) reported similar volume of distribution (Vd) of 0.28 liter per kilogram [L/kg] in both
male and female mice from both dose groups.
Olsen et al. (2009) estimated volumes of distribution for K+PFBS as 0.7 and 0.4 L/kg in male
and female rats, respectively, and 0.25 L/kg in male and female cynomolgus macaques and
7

-------
concluded that K+PFBS is primarily distributed in the extracellular space. Consistent with the
observations by 01 sen et al. (2009). Huang et al. (2019a) found that the overall Vd was generally
comparable between male rats (0.167-0.335 L/kg) and female rats (0.165-0.415 L/kg).
Chengelis et al. (2009) calculated a Vd of 0.25 L/kg in female cynomolgus macaques, consistent
with females from Olsen et al. (2009). The male monkey Vd from Chengelis et al. (2009) was
slightly lower (0.21 L/kg) than corresponding females and males from Olsen et al. (2009). These
results indicate Vd is generally comparable between male and female primates. Huang et al.
(2019a) also evaluated tissue concentrations in the liver, kidney, and brain and reported higher
PFBS concentrations in the liver compared to the kidney in male and female rats and lowest
concentrations in the brain.
Bogdanska et al. (2014) characterized the tissue distribution of 35S-labeled PFBS in male
C57BL/6 mice. Animals (3/group) were exposed for either 1, 3, or 5 days to an average of 16 mg
of PFBS/kg/day in the diet. Following 1, 3, and 5 days of exposure, total estimated recovery of
PFBS from all tissues evaluated was 10%, 5%, and 3.4% of the ingested dose, respectively. The
declining recovery with time reflects the lack of accumulation in tissues after the first few days,
with continued elimination in the urine. The study authors suggest that these low recovery rates
most likely reflect rapid excretion of PFBS and/or potentially limited uptake of the compound,
but the results of Lau et al. (2020) and Olsen et al. (2009) suggest that limited tissue distribution
is also a factor.
Bogdanska et al. (2014) found that blood levels of PFBS did not change when comparing values
observed after 1 and 5 days of exposure. As with PFOS, PFBS was found to distribute to most of
the 20 tissues examined at all exposure durations, but the levels of PFBS were significantly
lower (five-fold to forty-fold lower) than those of PFOS in tissues after similar exposure to
PFOS, especially in liver and lungs (Bogdanska et al.. 2014). These differences might be
attributed to chain length-dependent active transport of perfluorinated chemicals (Weaver et al..
2010). Excluding stomach and fat tissue, PFBS tissue levels increased between 1 and 3 days of
exposure, but there were no significant changes in tissue levels between 3 and 5 days of exposure
in any tissue examined. Similar to PFOS, whole bone, liver, blood, skin, and muscle accounted
for approximately 90% of the recovered PFBS at all time points. The highest tissue
concentrations outside of blood, however, were found in liver, GI tissues, kidney, and cartilage.
The significant total PFBS mass found in muscle and skin was due to the large total volume of
these tissues as much as the concentration in them. The liver contained the highest tissue
concentration of PFBS at all time points, while the brain contained the lowest.
Human studies were not available on lactational transfer of PFBS. Studies are sparse pertaining
to the transplacental transfer of PFBS in humans; in a Spanish mother-child paired cohort, PFBS
was not found in maternal blood samples or in corresponding cord blood during the first
trimester of pregnancy (Manzano-Salgado et al„ 2015). However, developmental studies in
animals indicate the potential for effects in offspring following gestational exposure suggesting
direct (i.e., fetus) and/or indirect (maternal/pregnant dam) effects of PFBS on offspring (Feng et
al.. 2017; York. 2003a. 2002).
Volume of distribution (Vd) is expected to be similar across mammalian species. For PFBS, the
average value for male and female monkeys (0.23 L/kg) is in the range estimated for male and
8

-------
female rats by Huang et al. (2019a) (0.17-0.42 L/kg), although estimates by 01 sen et al. (2009)
were a bit higher.
1.3.4	Metabolism
There is no evidence of biotransformation of PFBS. It is expected that PFBS, a short-chain (C4)
of perfluoroalkyl acids (PFAAs), is metabolically inert because of the chemical stability that also
exists in the longer chain PFAA chemicals, including perfluorohexane sulfonic acid (PFHxS)
(C6), PFOS (C8), and perfluorooctanoic acid (PFOA) (C8).
1.3.5	Elimination
To facilitate comparison of differing studies for a given species, results for elimination are
organized by species.
1.3.5.1	Mice
Lau et al. (2020) dosed male and female CD-I mice with 0, 30, or 300 mg/kg body weight PFBS
via a single gavage dose. Trunk blood was collected at 0.5, 1, 2, 4, 8, 16, 24, and 48 hours and
urine at 24 hours after dosing. Within 24 hours of gavage dosing, more than 95% of the PFBS
measured in serum was excreted into urine. Although the rate of PFBS clearance was linear with
administered doses, urine accounted for only 30-43% of the original gavage doses. The half-life
of PFBS was estimated to be 4.5 hours in the female mice and 5.8 hours in the males. Sex
difference in PFBS elimination is also noted that the elimination rate of absorbed PFBS is about
28% faster in female mice than male mice. Similarly, AUC estimates for the serum, kidney, and
liver compartments were higher in males than in females. The findings are generally comparable
to previous studies on rats (Huang et al.. 2019a; 01 sen et al.. 2009).
1.3.5.2	Rats
Chengelis et al. (2009) conducted a single-dose pharmacokinetic study in Sprague-Dawley (S-D)
rats, designed to compare the toxicokinetic behavior of PFBS to that of perfluorohexanoic acid
(PFHxA), another PFAA. In this study, 12 male and 12 female rats were each administered a
bolus dose of PFBS (10 mg/kg) via i.v. injection. Blood samples were collected from
three animals per sex at 0.5, 1, 1.5, 2, 4, 8, and 24 hours after dose administration. Additionally,
to determine urinary excretion, three animals per sex were housed in metabolic cages following
dose administration and urine was collected over the following time intervals: 0-6, 6-12, and
12-24 hours postdosing. Chengelis et al. (2009) fit the data to a non-compartmental model to
calculate pharmacokinetic parameters. Female rats had an approximately three-fold shorter mean
elimination half-life of PFBS in serum (0.64 h) than male rats (2.1 h). This could be in part due
to the difference in clearance and volume of distribution; the mean apparent clearance of PFBS
from the serum was approximately eightfold higher for female rats (0.311 L/h/kg) than for male
rats (0.0394 L/h/kg) and the mean apparent volume of distribution for PFBS in the serum was
approximately 2.4-fold higher for female rats (0.288 L/kg) than for male rats (0.118 L/kg).
Approximately 70% of the administered dose of PFBS was recovered in the urine over 24 hours
postdosing regardless of sex. Using the urine data, the mean half-life values for male rats and
female rats were determined to be 3.1 and 2.4 hours, respectively; the finding of longer urinary
half-lives in males is consistent with those observed for serum half-lives.
9

-------
Olsen et al. (2009) evaluated the elimination of PFBS in S-D rats after i.v. and oral exposure to
K+PFBS. The terminal serum elimination half-lives following i.v. administration of 30 mg/kg
K+PFBS were 4.51 ± 2.22 hours for males and 3.96 ± 0.21 hours for females (mean ± s.d.).
Although there was not a statistically significant difference between the terminal serum half-lives
in male and female rats, there was a statistically significant difference in the urinary clearance
rates (p < 0.01), with female rats (469 ± 40 mL/h) having faster clearance rates than male rats
(119 ± 34 mL/h). (Since clearance [CL] is calculated from the ratio of the volume of distribution
[Vd] to the half-life [ti/2], CL = 0.693 *Vd/ti/2, differences in Vd can lead to differences in CL,
even when ti/2 is similar between comparison groups.) For rats receiving an oral dose, terminal
serum K+PFBS elimination half-lives were significantly different (p < 0.05) for males
(ti/2= 4.68 ± 0.43 h) versus females (ti/2 = 7.42 ± 0.79 h).
Huang et al. (2019a) also evaluated elimination of PFBS following a single intravenous or
gavage dose in male or female Hsd:Sprague Dawley SD rats (4, 20, 100 mg/kg of K+PFBS).
Huang et al. (2019a) report elimination half-lives (ti/2,p) following i.v. administration of PFBS in
male and female rats of 4.22 and 0.95 h, respectively. The data for male rats after both oral and
i.v. dosing and female rats administered PFBS by i.v. fit a two-compartment model, whereas data
in female rats dosed via gavage fit a one-compartment model. Thus, elimination half-lives were
only reported for male rats following oral exposure and ranged from 4.89 - 5.36 hours. Overall
plasma elimination half-lives (kio ti/2) reported in female rats after oral administration were
between 1.11 - 1.50 hours, approximately 2 to 3-fold faster than in males that ranged from 2.7 -
4.4 hours. Similarly, clearance was 3 to 6-fold higher in females than males given the same dose
(26.0-75.5 mL/h/kg in males, 152-259 mL/h/kg in females).
The serum K+PFBS elimination half-lives reported by Huang et al. (2019a) are consistent with
the findings of Olsen et al. (2009) in male rats but not in female rats. In general, the elimination
half-life of serum PFBS observed by Huang et al. (2019a) in female rats was 2-to 4-fold shorter
than seen by Olsen et al. (2009). Similarly, Chengelis et al. (2009) calculated half-lives using a
one compartment model for each group, while Olsen et al. (2009) determined separate alpha and
beta phases via a two-compartment model. Thus, the half-life estimates of Olsen et al. (2009)
following i.v. administration (4.5-3.96 h) are higher than those estimated by Chengelis et al.
(2009) based on urine data (2.4 and 3.1 h).
1.3.5.3 Monkeys
Similar to their study in rats, Chengelis et al. (2009) investigated the toxicokinetic profile of
PFBS through a series of experiments in the cynomolgus macaque (Macaca fascicularis).
Monkeys (three males and three females) were each administered a bolus i.v. dose of 10 mg/kg
PFBS. The controlled exposure to PFBS occurred 7 days after the same animals were each
administered a bolus dose of PFHxA (10 mg/kg). Blood samples were collected at 0 hours
(immediately prior to dosing) and at 1, 2, 4, 8, 24, and 48 hours after dose administration and
were analyzed to determine PFBS concentration in serum. Only a single clearance half-life was
estimated. The estimated half-life of PFBS in serum ranged from 5.8 to 26.0 hours in this
experiment, and the median half-life was 9.55 hours for the six animals.
10

-------
Olsen et al. (2009) also evaluated the elimination of PFBS (specifically, K+PFBS) in cynomolgus
macaques after i.v. dosing. A significant difference in design from the study of Chengelis et al.
(2009) is that Olsen et al. (2009) followed PFBS elimination for 31 days in monkeys (versus
48 hours), allowing Olsen and colleagues to identify both an initial clearance half-life and a
terminal phase-half-life. Olsen et al. (2009) did not observe statistically significant sex-related
differences in half-life or clearance between male and female monkeys, unlike those observed in
rats. In monkeys, the mean terminal serum elimination half-lives, after i.v. administration of
10	mg/kg K+PFBS, were 95 ± 27 hours in males and 83 ± 42 hours in females.
The serum half-life data in Olsen et al. (2009) clearly show a slow elimination phase in monkeys
that does not begin until 4-10 days after dosing. Chengelis et al. (2009) followed elimination for
only 48 hours, hence could not have observed this terminal clearance phase. The initial
elimination half-life (ti/2,p) estimated by Olsen et al. (2009) in monkeys—13 hours for males,
11	hours for females—is essentially identical to the values estimated by Chengelis et al.
(2009)—10 or 15 hours for males (without/with outlier) and 8 hours in females. Hence the two
studies appear consistent in identifying an initial elimination half-life, but the difference in
design precluded Chengelis and colleagues from identifying the longer (terminal) half-life of
PFBS.
1.3.5.4 Humans
In addition to their experimental studies in rats and monkeys, Olsen et al. (2009) evaluated the
elimination of human serum K+PFBS in a group of workers with occupational exposure, with
serum concentrations measured up to 180 days after cessation of further K+PFBS work-related
activity. Given that the workers had been occupationally exposed, distribution into the tissues is
expected to have been complete before the observations began. The reported mean serum half-
life was 24.1 days in males (n=5) and 45.7 days in females (n=l). Among the six subjects
(five males, one female), the reported geometric mean serum elimination half-life for K+PFBS
was 25.8 days (95% confidence interval = 16.6-40.2 days). Since there was only one female
subject, these data cannot be used to establish a significant sex difference in elimination. Urine
appeared to be a major route of elimination based on observed levels of PFBS in urine in the
human study.
Xu et al. (2020) also measured PFBS elimination in a study population with previous
occupational exposure, in this case airport employees who were exposed to firefighting foam that
contained PFBS. Eleven male and six female employees provided repeated blood samples during
a period of observation with minimal exposure and the data were analyzed with a linear mixed-
effects pharmacokinetic model. The average half-life (95% CI) was 44 d (37-55 d). While Xu et
al. (2020) evaluated age and sex as covariates of their statistical model, they do not report either
as being a significant factor for PFBS. The average half-life (44 d) is larger than that reported by
Olsen et al. (2009) (25.8 d), but there is significant overlap: the range of Xu et al. (2020) is 21.6-
87.2 d while the range of Olsen et al. (2009) is 13.1-45.7 d.
For the sake of comparison, the linear mixed model used by Xu et al. (2020) was also applied to
the estimated serum PFBS elimination half-life for the population and each individual worker
(five male, one female) who manufactured K+PFBS, described in Olsen et al. (2009). In brief, a
linear mixed effect model is an extension of simple linear models that can be used to estimate
toxicokinetic parameters such as serum elimination rate constant (keiim) and half-life by assuming
11

-------
one-compartment first-order elimination kinetics. The details of the linear mixed-effect model
have been reported previously Li et al. (2018). Because of the limited sample size (only one
female worker) and the participant age was not available for each worker in the study, age and
sex were not included in the linear mixed model for reanalysis of the 01 sen et al. (2009) data,
whereas both were included in Xu et al. (2020). In general, the estimated half-life using the
linear mixed effect model were similar to originally reported values in Olsen et al. (2009). For
instance, as compared to the reported average of 25.8 d ranging from 13.1-45.7 d (Olsen et al.,
2009). the estimated population elimination half-life for serum PFBS was 25.0 d with individual
estimates of 14.6-42.9 d using the linear mixed effect model.
While the estimated serum half-life of PFBS in Olsen et al. (2009) overlapped with those of Xu
et al. (2020) (mean=43.8 d, range = 21.9-87.6 d), there is a statistically significant difference
between these two studies as suggested by both parametric (One-Way Analysis of Variance,
ANOVA) and non-parametric analyses (Kruskal-Wallis test). Overall, the estimated serum half-
life of PFBS by Xu et al. (2020) is about two folds higher than Olsen et al. (2009).
Some of the difference between Xu et al. (2020) and Olsen et al. (2009) may result from the
difference in initial concentration, where the Olsen et al. (2009) subjects had initial
concentrations ranging from 100-1000 ng/mL PFBS, while the highest initial concentrations in
Xu et al. (2020) was 1.3 ng/mL. It is possible that the higher serum levels in the Olsen et al.
(2009) subjects resulted in saturation of renal resorption, hence more rapid excretion/shorter
half-lives. However, to the extent that some ongoing low-level exposure occurred during the
period of observation, such exposure would cause a greater bias towards over-estimation of the
elimination half-life for the Xu et al. (2020) subjects than those of Olsen et al. (2009). The data
of Olsen et al. (2009) might also have a greater signal/noise ratio than the data of Xu et al.
(2020). Despite this uncertainty, the fact that the blood concentrations of the Xu et al. (2020) are
more representative of environmental exposure, that their population was larger, and a significant
statistical difference was observed, the two data sets will not be combined and the half-life
estimated by Xu et al. (2020) is presumed to better predict human dosimetry at environmental
levels.
The possibility that menstrual blood loss could contribute to overall clearance was evaluated,
assuming that the concentration of PFBS in menstrual blood is the same as in the general
circulation and that the volume of distribution in humans is equal to the average value estimated
for monkeys (0.23 L/kg). The results indicate that this avenue of loss is more than 2 orders of
magnitude slower than that indicated by the measured PFBS half-life in humans. Thus, menstrual
blood loss is unlikely to contribute significantly to overall PFBS elimination.
1.3.6 Physiologically Based Pharmacokinetic Models
Fabrega et al. (2015) developed a physiologically based pharmacokinetic model to estimate the
concentration of PFAS, including PFBS, in human tissues, based on an existing model and
experimental data on concentrations of perfluoroalkyl substances in human tissues from
individuals in Catalonia, Spain. Several uncertainties in the model limit the use for this
assessment of PFBS.
12

-------
There are three chemical-specific parameters which determine the rate of elimination: the free
fraction in blood, the maximum rate of resorption in the kidney (Tm), and the saturation constant
for that resorption (Kt). No details beyond a rough description are provided on how these
parameter values were identified. The data used for calibration are population samples in adults,
who would essentially be at steady state, and only a single average level of exposure and
corresponding blood concentration are reported, precluding the possibility of evaluating
exposure- or concentration-dependence. In this situation it is not possible to uniquely identify the
three parameters. This lack of identifiability is likely to be an underlying cause of the extreme
variability in the individual parameter values (among the 11 PFAS evaluated) reported by
Fabrega et al. (2015).
In addition, the rate constant for elimination from the glomerular filtrate compartment to the
urine "storage" compartment (i.e., the bladder) is the total glomerular filtration rate (GFR),
which is approximately 10 L/h in a 70 kg adult. But most of the glomerular flow is resorbed in
the nephrons and human urinary output is less than 2 L/d. Hence, use of GFR for elimination is
not realistic. Finally, while the model structure and the equations listed by Fabrega et al. (2015)
appear to be appropriate for most humans, it should be noted that excretion via lactation is not
included.
Of considerable concern is the way in which partition coefficients (PCs) were identified. In
particular, PCs were obtained by taking tissue concentration data from cadavers and comparing
those to average blood concentrations from volunteer subjects, albeit from the same geographical
area (county in Spain). The livenblood PC for PFDA was thereby estimated to be 0.001 while
the value for PFNA was 1.65. By contrast Kim et al. (2019) obtained values of- 0.6-0.7 for
PFDA in male and female rats, -1.2 for PFNA in male rats, and - 0.5 for PFNA in female rats.
Thus, there appears to be extreme inconsistency and hence uncertainty in these parameters as
estimated by Fabrega et al. (2015). Generally, human PCs should be similar in value to those in
rats.
The authors do not compare model predictions for Tarragona County, Spain, to measured values
for county residents; i.e., the data used for model calibration. Also, the authors state that 20-30
years of simulated time are required to reach steady state. These steady state estimates are
inconsistent with the elimination data from 01 sen et al. (2009). where the half-life in males was
24 days, and in a female subject of 46 days; these empirical half-lives are consistent with a time
to steady state of less than a year, indicating that the predicted clearance from Fabrega et al.
(2015) may be an order of magnitude or more too low. At the same time, the simulated levels of
5 PFAS (average levels) were consistently lower than the averages in a validation data, 4 of these
being low by an order of magnitude or more.
Thus, predictions of the Fabrega et al. (2015) model are considered highly uncertain and data
other than those used by the authors will be needed to accurately estimate key PK parameters for
PFBS and these other PFAS, a task that would require significant additional research.
1.3.7 Summary
Collectively, elimination half-lives appear to be similar for mice and rats, with potential
sex-specific toxicokinetic differences being reported (i.e., females appearing to have a faster
13

-------
elimination rate). Humans have a longer serum elimination half-life (-weeks) than both rodents
(-hours) and monkeys (-days). Further, although volume of distribution information is not
available for humans, observations in male and female mice, rats and monkeys exposed to
comparable doses indicate comparability across species. Results of all studies discussed in this
section are summarized in Table 2.
14

-------
2.0 Problem Formulation
2.1 Conceptual Model
A conceptual model was developed to summarize the availability of data to understand potential
health hazards related to exposure to PFBS and/or K+PFBS. The potential sources of these
chemicals, the routes of exposure for biological receptors of concern (e.g., various human
activities related to ingested drinking water, and food preparation and consumption), the
potential organs and systems affected by exposure (e.g., effects such as developmental toxicity),
and potential populations at risk due to exposure to PFBS and/or potassium salt are depicted in
the conceptual diagram in Figure 2. Arrows indicate linkage between one or more boxes between
levels of organization.
15

-------
STRESSOR
POTENTIAL
SOURCES OF
EXPOSURE
PFBS and Potassium Salt
Drinking
Water
Ambient
Water
Industrial
Uses
Consumer
Products
Food
Dust
Air
Soil
Fire-
fighting
Foams
EXPOSURE ROUTES
POTENTIAL
ORGANS/
SYSTEMS
AFFECTED
Oral
Dermal
Thyroid Effects
POTENTIAL
RECEPTORS IN
GENERAL
POPULATION
Reproductive
Effects
Developmental
Effects
Renal Effects
Hepatic Effects
Adults
Children
Pregnant Women
and Fetuses

Lipid and
Lipoprotein
Effects
Lactating Women
Inhalation
LEGEND
Data Selected for
Assessment
Limited Data
Unknown
Quantitative Data
Figure 2. Conceptual model for PFBS and/or potassium salt.
16

-------
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
2.2 Objective
The overall objective of this assessment is to provide the health effects basis for the development
of oral reference doses (RfDs) for PFBS (CASRN 375-73-5) and a related compound, K+PFBS
(CASRN 29420-49-3), including the science-based decisions providing the basis for
identification of potential human health effects and estimating PODs. Based on the needs of the
EPA partner Program Offices, Regions, States, and/or Tribes as they pertain to diverse exposure
scenarios and human populations, subchronic and chronic RfDs have been derived. The
assessment includes studies and information previously provided in the 2014 Provisional
Peer-Reviewed Toxicity Value assessment (U.S. EPA. 2014f) and builds upon the amount of
literature containing studies published since that review.
2.3 Methods
2.3.1 Literature Search
Four online scientific databases (PubMed, Web of Science, Toxline, and TSCATS via Toxline)
were searched by the EPA's Health and Environmental Research Online (HERO) staff and
stored in the HERO database.5 The literature search focused on chemical name and synonyms
with no limitations on publication type, evidence stream (i.e., human, animal, in vitro, and in
silico), or health outcomes. Full details of the search strategy for each database are presented in
appendix A. The initial database searches were conducted on July 18, 2017, and updated on
February 28, 2018, May 1, 2019, and May 15, 2020. Additional studies (e.g., Lau et al. (2020);
Xu et al. (2020)) were identified during subsequent review periods and integrated into the
assessment as appropriate. Studies were also identified from other sources relevant to PFBS,
including studies submitted to the EPA by the manufacturer of PFBS (i.e., 3M) as part of Toxic
Substances Control Act (TSCA) premanufacture notices for other PFAS chemicals or as required
under TSCA reporting requirements and studies referenced in prior evaluations of PFBS toxicity
(MDH. 2017; ATSDR. 2015). In addition, on March 29, 2018, the National Toxicology Program
(NTP) published study tables and individual animal data from a 28-day toxicity study of PFBS
(http://doi.org/10.22427/NTP-DATA-002-01134-0003-0000-4). with a protocol outlining the
NTP study methods available in HERO
(https://hero.epa.gov/hero/index.cfm/reference/details/reference id/4309741) (NTP. 2011). The
final NTP Technical Report on the Toxicity Studies ofPerfluoroalkyl Sulfonates Administered by
Gavage to Sprague Daw ley Rats was published in August 2019 (NTP. 2019).
2.3.2 Screening Process
Two screeners independently conducted a title and abstract screening of the search results using
DistillerSR6 to identify study records that met the Population, Exposure, Comparator, Outcome
(PECO) eligibility criteria (see appendix B for a more detailed summary):
• Population: Human and nonhuman mammalian animal species (whole organism) of any
life stage and in vitro models of genotoxicity.
5The EPA's Health and Environmental Research Online (HERO) database provides access to the scientific literature
behind EPA science assessments. The database includes more than 2,500,000 scientific references and data from the
peer-reviewed literature used by the EPA to develop its regulations.
6DistillerSR is a web-based systematic review software used to screen studies available at
https://www.evidencepartners.com/products/distillersr-svstematic-review-software.
17

-------
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
•	Exposure: Any qualitative or quantitative estimates of exposure of PFBS or K+PFBS, via
oral or inhalation routes of exposure. (Note: Non-oral and non-inhalation studies are
tracked as potential supplemental material and are presented in Section 4.8.2.)
•	Comparator: A comparison or reference population exposed to lower levels or for shorter
periods of time for humans. Exposure to vehicle-only or untreated control in animals.
•	Outcome: Any examination of cancer or noncancer health outcomes.
In addition to the PECO criteria, the following additional exclusion criteria were applied,
although these study types were tracked as supplemental material as described following the
exclusion criteria:
•	Records that do not contain original data such as other agency assessments, scientific
literature reviews, editorials, and commentaries;
•	Abstract only (e.g., conference abstracts); and
•	Retracted studies.
Records that were not excluded based on title and abstract screening advanced to full-text review
using the same PECO eligibility criteria. Studies that have not undergone peer review were
included if the information could be made public and sufficient details of study methods and
findings were included in the reports. Full-text copies of potentially relevant records identified
from title and abstract screening were retrieved, stored in the HERO database, and independently
assessed by the screeners using DistillerSR to confirm eligibility. At both title/abstract and
full-text review levels, screening conflicts were resolved by discussion between the primary
screeners in consultation with a third reviewer to resolve any remaining disagreements. During
title/abstract or full-text level screening, studies that were not directly relevant to the PECO, but
could provide supplemental information, were categorized (or "tagged") by the type of
supplemental information they provided (e.g., review, commentary, or letter with no original
data; conference abstract; toxicokinetics; mechanistic information aside from in vitro
genotoxicity studies; other routes of exposure; exposure only). Conflict resolution was not
required during the screening process to identify supplemental information (i.e., tagging by a
single screener was sufficient to identify the study as potential supplemental information).
2.3.3 Study Evaluation
Study evaluation was conducted by one reviewer for epidemiological studies and by two
independent reviewers for animal studies using the EPA's version of Health Assessment
Workspace Collaborative (HAWC), a free and open source web-based software application
designed to manage and facilitate the process of conducting literature assessments.7 For
pragmatic purposes, only one reviewer was considered necessary for epidemiological studies
because it was apparent during literature screening that the animal evidence would be most
informative for deriving toxicity values. The available outcomes in the epidemiological studies
were heterogeneous and unrelated to each other, and only a single study was available for each
outcome. This approach is consistent with recommendations from the National Academies of
Science encouraging the EPA to explore ways to make systematic review more feasible,
including a "rapid review in which components of the systematic review process are simplified
7HAWC: A Modular Web-Based Interface to Facilitate Development of Human Health Assessments of Chemicals.
https ://hawcproi ect. org/.
18

-------
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
or omitted (e.g., the need for two independent reviewers)" (NASEM, 2017). Study evaluation
was not conducted for studies tagged as supplemental information during screening.
The general approach for evaluating epidemiology and animal toxicology was the same
(see Figure 3), but the specifics of applying the approach differed. These evaluations were
focused on the methodological approaches and completeness of reporting in the individual
studies, rather than on the direction or magnitude of the study results. Evaluation of
epidemiology studies was conducted for the following domains: exposure measures, outcome
measures, participant selection, confounding, analysis, sensitivity, and selective reporting. For
animal studies, the evaluation process focused on assessing aspects of the study design and
conduct through three broad types of evaluations: reporting quality, risk of bias, and study
sensitivity. A set of domains with accompanying core questions fall under each evaluation type
and directed individual reviewers to evaluate specific study characteristics. For each domain
evaluated for experimental animal studies—reporting quality, selection or performance bias,
confounding/variable control, reporting or attrition bias, exposure methods sensitivity, and
outcome measures and results display—basic considerations provided additional guidance on
how a reviewer might evaluate and judge a study for that domain. Core and prompting questions
used to guide the criteria and judgment for each domain are presented in appendix C. Key
concerns for the review of epidemiology and animal toxicology studies are potential sources of
bias (factors that could systematically 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).
For each study in each evaluation domain, reviewers reached a consensus rating regarding the
utility of the study for hazard identification, with categories of good, adequate, deficient, not
reported, or critically deficient. These ratings were then combined across domains to reach an
overall classification of high, medium, or low confidence or uninformative (definitions of these
classifications are available in appendix C). The rationale for the classification, including a brief
description of any identified strengths and/or limitations from the domains and their potential
impact on the overall confidence determination, is documented and retrievable in HAWC.
Uninformative studies were not used in evidence synthesis or dose-response analysis. Studies
were evaluated for their suitability for each health outcome investigated and could receive
different ratings for each outcome.
For epidemiological studies, exposure-specific criteria were developed prior to evaluation and
are described in detail in appendix C. In brief, standard analytical methods of measurement of
PFBS in serum or whole-blood using quantitative techniques such as liquid chromatograph-triple
quadrupole mass spectrometry and high-pressure liquid chromatography with tandem mass
spectrometry were preferred. In addition, exposure must have been assessed in a relevant
time-window for development of the outcome.
19

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Individual evaluation domains
V7
Domain judgments
Animal
Epidemiological
Reporting Quality
Exposure measurement
Selection or Performance Bias
Outcome ascertainment
Confounding/Variable Control
Population Selection
Reporting or Attrition Bias
Confounding
Exposure Methods Sensitivity
Outcome Measures and Results
Display
Analysis
Sensitivity
Selective reporting

Judgement
Interpretation
0
Good
Appropriate study conduct relating to the domain & minor
deficiencies not expected to influence results.
o
Adequate
A study that may have some limitations relating to the domain, but
they are not likely to be severe or to have a notable impact on
results.
•
Deficient
Identified biases or deficiencies interpreted as likely to have had a
notable impact on the results or prevent reliable interpretation of
study findings.
o
Critically
Deficient
A serious flaw identified that is interpreted to be the primary driver
of any observed effect or makes the study uninterpretable. Study is
not used without exceptional justification.
Overall study rating
for an outcome
Rating
Interpretation
High
Medium
Low
Uninformative
No notable deficiencies or concerns identified: potential for bias unlikely
or minimal; sensitive methodology.
Possible deficiencies or concerns noted, but resulting bias or lack of
sensitivity would be unlikely to be of a notable degree.
Deficiencies or concerns were noted, and the potential for substantive
bias or inadequate sensitivity could have a significant impact on the study
results or their interpretation.
Serious flaw(s) makes study results unusable for hazard identification.
Figure 3. Approach for evaluating epidemiological and animal toxicology studies.
2.3.4 Data Extraction
Information on study design, methods, results, and data from animal toxicology studies were
extracted into the HAWC and are available at https://hawcprd.epa.gov/assessment/100000037/.
Visual graphics prepared from HAWC are embedded as hyperlinks and are fully interactive
when viewed online by way of a "click to see more" capability. Clicking on content allows
access to study evaluation ratings, methodological details, and underlying study data. The action
of clicking on content contained in those visual graphics (e.g., data points, endpoint, and study
design) will yield the underlying data supporting the visual content. NOTE. The following
browsers are fully supported for accessing HAWC: Google Chrome (preferred), Mozilla Firefox,
and Apple Safari. There are errors in functionality when viewed with Internet Explorer. Study
methods and findings from epidemiological studies were described in narratives given the small
size and heterogeneity of the evidence base. Data extraction was performed by one member of
the evaluation team and checked by one to two other members. Any discrepancies in data
extraction were resolved by discussion or consultation with a third member of the evaluation
team. Digital rulers such as WebPlotDigitizer and Grab It (https://automeris.io/WebPlotDigi tizer/
and https://grab-it.softl 12.com/, respectively) were used to extract numerical information from
20

-------
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
figures. Use of digital rulers was documented during extraction. Dose levels were extracted as
reported in the study and converted to milligrams per kilogram per day(mg/kg-day) human
equivalent dose (HED) for endpoints that were considered for use in the dose-response and
derivation of toxicity values.
2.3.5	Evidence Synthesis
For the purposes of this assessment, after study evaluation, the informative evidence for each
outcome was summarized from the available human studies and, separately, the available animal
studies. This synthesis provides a short synopsis of the breadth of data available to inform each
outcome and summarizes information on the general study design, doses tested, outcomes
evaluated, and results for the endpoints of interest within each study. While the evidence
synthesis describes inferences about the methodological rigor and sensitivity of the individual
studies (i.e., study confidence) and discusses the pattern and magnitude of the experimental
findings within studies, it does not include conclusions drawn across the sets of studies (see
"Evidence Integration and Hazard Characterization," next).
2.3.6	Evidence Integration and Hazard Characterization
In this assessment, the evaluation of the available evidence from informative human and animal
studies was described in an evidence integration narrative for each outcome, including overall
evidence integration judgments as to whether the data provide evidence sufficient to support a
hazard. These integrated judgments serve to characterize the extent of the available evidence for
each outcome, including information on potential susceptible populations and life stages, as well
as important uncertainties in the interpretation of the data.
The evidence integration for each health effect considered aspects of an association that might
suggest causation first introduced by Austin Bradford Hill (Hill, 1965), including the
consistency, exposure-response relationship, strength of association, biological plausibility, and
coherence of the evidence. This involved weighing the PFBS-specific human and animal
evidence relating to each of these considerations within or across studies, including both
evidence that supported causation as well as evidence that indicated lack of support. For
example, the evaluation of consistency examined the similarity of results across studies (e.g.,
direction and magnitude). When inconsistencies across studies were identified, the evaluation
considered whether results were "conflicting" (i.e., unexplained positive and negative results in
similarly exposed human populations or in similar animal models) or "differing" (i.e., mixed
results explained by differences between human populations, animal models, exposure
conditions, or study methods), based on analyses of potentially important explanatory factors
such as confidence in studies' results (the results of higher confidence studies were emphasized),
exposure levels or duration, or differences in populations or species (including potential
susceptible groups) across studies (U.S. EPA, 2005). While consistent evidence across studies
increases support for hazard, unexplained inconsistency or conflicting evidence decreases
support for hazard. The evaluations of these considerations were informed by EPA guidelines,
including Guidelines for Developmental Toxicity Risk Assessment (U.S. EPA, 1991a) and
Guidelines for Reproductive Toxicity Risk Assessment (U.S. EPA, 1996b).
The overall evidence integration judgments were developed using a structured framework based
on evaluation of the considerations above (see Table 3). Using this framework, the human and
21

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
animal evidence for each health effect was judged separately as supports a hazard, equivocal, or
supports no hazard. Evidence integration judgments of supports a hazard span a range of
supportive evidence bases that can be further differentiated by the quantity and quality of
information available to rule out alternative explanations for the results. Equivocal evidence is
limited in terms of the quantity, consistency, or confidence level of the available studies and
serves to encourage additional research. Supports no hazard requires several high-confidence
studies across potentially susceptible populations with consistent null results; this judgment was
not reached in this assessment. Overall evidence integration judgments were drawn across the
human and animal conclusions, considering the available information on the human relevance of
findings in animals. Thus, for example, evidence in animals that supports a hazard alongside
equivocal human evidence in the absence of information indicating that the responses in animals
are unlikely to be relevant to humans would result in an overall judgment of supports a hazard
for that outcome.
Table 3. Criteria for overall evidence integration judgments

Animal
Human
Supports a
hazard
The evidence for effects is consistent or largely
consistent in at least one high- or medium-
confidence experiment.3 Although notable
uncertainties across studies might remain, any
inconsistent evidence or remaining uncertainties
are insufficient to discount the cause for concern
from the positive experiments. In the strongest
scenarios, the set of experiments provide evidence
supporting a causal association across
independent laboratories or species. In other
scenarios, including evidence for an effect in a
single study, the experiment(s) demonstrate
additional support for causality such as coherent
effects across multiple related endpoints; an
unusual magnitude of effect, rarity, age at onset,
or severity; a strong dose-response relationship;
and/or consistent observations across exposure
scenarios (e.g., route, timing, or duration), sexes,
or animal strains.
One or more high- or medium-confidence
independent studies reporting an association
between the exposure and the health outcome. In
general, the study results are largely consistent or
any inconsistent results are not sufficient to
discount the cause for concern from the higher
confidence study or studies, and there is
reasonable confidence that alternative
explanations, including chance, bias, and
confounding, have been ruled out. In situations in
which only a single study is available, the results
of multiple studies are heterogeneous, or
alternative explanations, including chance, bias
and confounding, have not been ruled out, there is
additional supporting evidence such as
associations with biologically related endpoints in
other human studies (coherence), large estimates
of risk, or strong evidence of an
exposure-response within or across studies.
Equivocal
The evidence is generally inadequate to determine
hazard. This includes a lack of relevant studies
available or a set of low-confidence experiments.
It also includes scenarios with a set of high- or
medium-confidence experiments that are not
reasonably consistent or not considered
informative to the hazard question under
evaluation. This category would also include a
single high- or medium-confidence experiment
with weak evidence of an effect (e.g., changes in
one endpoint among several related endpoints,
and without additional evidence supporting
causality).
The evidence is considered inadequate to describe
an association between exposure and the health
outcome with confidence. This includes a lack of
studies available in humans, only low-confidence
studies, or considerable heterogeneity across
medium- or high-confidence studies. This also
includes scenarios in which there are serious
residual uncertainties across studies (these
uncertainties typically relate to exposure
characterization or outcome ascertainment,
including temporality) in a set of largely
consistent medium- or high-confidence studies.
22

-------
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

Animal
Human
Supports no
hazard
A set of high-confidence experiments examining
the full spectrum of related endpoints within a
type of toxicity, with multiple species, and testing
a reasonable range of exposure levels and
adequate sample size in both sexes, with none
showing any indication of effects. The data are
compelling in that the experiments have examined
the range of scenarios across which health effects
in animals could be observed, and an alternative
explanation (e.g., inadequately controlled features
of the studies' experimental designs) for the
observed lack of effects is not available. The
experiments were designed to specifically test for
effects of interest, including suitable exposure
timing and duration, post-exposure latency, and
endpoint evaluation procedures, and to address
potentially susceptible populations and life stages.
Several high-confidence studies, showing
consistently null results (e.g., an odds ratio of 1.0)
ruling out alternative explanations including
chance, bias, and confounding with reasonable
confidence. Each of the studies should have used
an optimal outcome and exposure assessment and
adequate sample size (specifically for higher
exposure groups and for sensitive populations).
The set as a whole should include the full range of
levels of exposures that human beings are known
to encounter, an evaluation of an exposure
response gradient, and at-risk populations and life
stages and should be mutually consistent in not
showing any indication of effect at any level of
exposure.
Note:
a "Experiment" refers to measurements in a single population of exposed animals (e.g., a study that included separate evaluations
of rats and of mice, or separate cohorts exposed at different life stages, would be considered as multiple experiments).
Conversely, two papers or studies that report on the same cohort of exposed animals (e.g., examining different endpoints) would
not be considered separate experiments.
The primary evidence and rationale supporting these decisions were summarized in a single
evidence profile table to transparently convey the aspects of the evidence that were considered to
increase or decrease the hazard support for each health effect. For the purposes of this
assessment, only the integrated evidence that supports a hazard was considered for use in the
dose-response and derivation of toxicity values.
2.3.7 Derivation of Values
Development of the dose-response assessment for PFBS and/or the potassium salt has followed
the general guidelines for risk assessment put forth by the National Research Council (NRC.
1983) and the EPA's l'ramework for Human Health Risk Assessment to Inform Decision Making
(U.S. EPA. 2014c). Other EPA guidelines and reviews considered in the development of this
assessment include the following:
• A Review of the Reference Dose and Reference Concentration Processes (U.S. EPA.
2002).
• A Framework for Assessing Health Risks of Environmental Exposures to Children (U.S.
EPA. 2006).
• Exposure Factors Handbook (U.S. EPA. 201 la)8.
• Recommended Use of Body Weight4 as the Default Method in Derivation of the Oral
Reference Dose (U.S. EPA. 201 lb).
• Guidance for Applying Quantitative Data to Develop Data-Derived Extrapolation
Factors for Interspecies and Intraspecies Extrapolation (U.S. EPA. 2014d).
• Benchmark Dose Technical Guidance Document (U.S. EPA. 2012).
8 please note that specific updates to this Handbook are available at
https ://cfpub. epa. gov/ncea/risk/recordisplav. cfm?deid=236252
23

-------
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
• Child-Specific Exposure Scenarios Examples (U.S. EPA. 2014a).
The EPA's A Review of the Reference Dose and Reference Concentration Processes document
describes a multistep approach to dose-response assessment, including analysis in the range of
observation followed by extrapolation to lower levels (U.S. EPA. 2002). As described above,
prior to deriving toxicity values, the EPA conducted a comprehensive evaluation of available
human epidemiological and animal toxicity studies to identify potential health hazards and
associated dose-response information through the literature search and screening, study
evaluation, evidence synthesis, and evidence integration steps. This evaluation informed the
selection of candidate key studies and critical effects for dose-response analysis, from which the
EPA identified a critical effect and point of departure (POD) for subchronic and chronic
reference value derivation and extrapolated a selected POD to a corresponding RfD (e.g.,
subchronic RfD). For dose-response analysis of PFBS and/or the potassium salt, the EPA used
the BMD approach to identify a POD. The steps for deriving an RfD using the BMD approach
are summarized below.
• Step 1: Evaluate the data to identify and characterize endpoints related to exposure
to PFBS chemicals. This step involved determining the relevant studies and adverse
effects to be considered for BMD modeling. Once the appropriate data were collected,
evaluated for study quality, and characterized for adverse outcomes, endpoints were
selected that were judged to be relevant (i.e., for the purposes of this assessment, effects
that were sufficient to support a hazard) and sensitive as a function of dose (typically
defined by the no observed adverse effect level [NOAEL] value). In this assessment,
these decisions were directly informed by the evidence integration judgments arrived at
for each assessed health outcome. Some of the most important considerations that
influenced selection of endpoints for BMD modeling include data with dose-response,
percent change from controls, adversity of effect, and consistency across studies. For
PFBS, thyroid, developmental, and kidney endpoints were considered for toxicity value
derivations.
• Step 2: Convert the adjusted daily doses to an HED. The adjusted daily doses were
converted to HEDs by considering EPA's Recommended Use of Body Weight4 as the
Default Method in Derivation of the Oral Reference Dose (U.S. EPA. 201 lb).
• Step 3: Select the benchmark response (BMR) level. Using the EPA's Benchmark
Dose Technical Guidance Document (U.S. EPA. 2012). the endpoints selected were
modeled. The BMR is a predetermined change in the response rate of an adverse effect. It
serves as the basis for obtaining the benchmark dose lower confidence limit (BMDL),
which is the 95% lower bound of the BMD. BMRs were identified and applied consistent
with quantal and continuous data and, when possible, informed by understanding of
biological significance.
• Step 4: BMD Model the data. This step involved fitting a statistical model to the
dose-response data that describes the data set of the identified adverse effect. Typically,
this involved selecting a family or families of models (e.g., polynomial continuous, hill
continuous, or exponential continuous) for further consideration based on the data and
experimental design. In this step, a BMDL was derived by placing confidence limits
(one- or two-sided) and a confidence level (typically 95%) on a BMD to obtain the dose
that ensures with high confidence that the BMR is not exceeded.
24

-------
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
•	Step 5: Determine a PODhed. If modeling was feasible, the estimated BMDL(HED)s
were used as PODs (i.e., PODhed). If dose-response modeling was not feasible, NOAEL
(HED)s or lowest observed adverse effect level (LOAEL) (HED)s were identified.
•	Step 6: Provide rationale for selecting Uncertainty Factors (UFs). UFs were selected
in accordance with EPA guidelines considering variations in sensitivity among humans,
differences between animals and humans, the duration of exposure in the key study
compared to a lifetime of the species studied, and the potential limitations of the
toxicology database.
•	Step 7: Calculate the subchronic and chronic RfDs. The RfDs were calculated by
dividing a PODhed by the selected UFs.
RfD = PODhfd
UFc
where:
PODhed = The PODhed is calculated from the BMDL or NOAEL using a BW3 4
allometric scaling approach consistent with EPA guidance (U.S. EPA. 2011b)
UFc = Composite UF established in accordance with EPA guidelines considering
variations in sensitivity among humans, differences between animals and humans, the
duration of exposure in the key study compared to a lifetime of the species studied, and
the potential limitations of the toxicology database.
•	Step 8: Assignment of Confidence Levels. In assessments in which an RfD or RfC is
derived, characterization of the level of confidence in the principal study(ies), the
database associated with that reference value, and the overall confidence in the reference
value(s) are provided. Details on characterizing confidence are provided in Ch.4
(specifically section 4.3.9.2) of the U.S. EPA's Methods for Derivation of Inhalation
Reference Concentrations and Application of Inhalation Dosimetry (U.S. EPA,
1994). For example, the confidence ranking in database (low, medium, or high) reflects
the degree of belief that the reference value (e.g., RfD) will change (in either direction)
with the acquisition of new data.
25

-------
1
2
3
4
5
6
7
8
9
10
11
12
3.0	Overview of Evidence Identification for Synthesis and
Dose-Response Analysis
3.1	Literature Search and Screening Results
The database searches yielded 434 unique records, with 50 records identified from additional
sources such as TSCA submissions, posted NTP study tables, peer-review recommendations, and
review of reference lists from other authoritative sources. Of the 434 studies identified, 317 were
excluded during title and abstract screening, 117 were reviewed at the full-text level, and 42
were considered relevant to the PECO eligibility criteria (see Figure 4). This included 19
epidemiologic studies (described in 22 publications), 10 in vivo animal studies (described in 15
peer-reviewed and nonpeer-reviewed publications), and five in vitro genotoxicity studies. The
detailed search approach, including the query strings and PECO criteria, is provided in
appendix A and appendix B, respectively.
26

-------
Figure 4. Literature search and screening flow diagram for PFBS (CASRN 375-73-5).
27

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
3.2 Study Evaluation Results
Based on the study evaluations, seven human epidemiology studies were considered
uninformative and are not discussed any further in this assessment (see Table 4). No animal
studies were considered uninformative and, thus, all animal studies identified as relevant during
literature screening were included in the evidence synthesis and dose-response analysis. Overall,
12 epidemiologic studies (described in 15 publications) and 10 in vivo animal studies (described
in 15 peer-reviewed and nonpeer-reviewed publications) were included in the evidence synthesis
and further evaluated for use in the development of toxicity values for PFBS. As shown in
Figures 5 and 6, while the database of studies on PFBS is not large, a number of high- and
medium-confidence oral exposure studies in animals were identified, as were several medium-
confidence studies in humans. Multiple publications of the same study are not listed as
independent studies in HAWC, they are reviewed together in one entry. In addition, Shiue (2016)
was not evaluated because the outcome (i.e., sleep disturbances) was considered a nonspecific
effect, and thus was not entered into HAWC. No studies were identified evaluating the toxicity
of PFBS or K+PFBS following inhalation exposure or on the carcinogenicity of PFBS or
K+PFBS in humans or animals.
Table 4. Epidemiological studies excluded based on study evaluation
Reference
Outcome
Reason for exclusion
Bao et al. (2017)
Blood pressure
Extremely poor sensitivity (96% of participants below the
LOD for PFBS measurement) with no observed
association.
Berketal. (2014)
Depression
Serious concerns with temporality between exposure and
outcome, confounding, and analysis.
Gvllenhammar et al. (2018)
Birth size, weight gain
Extremely poor sensitivity (median exposure = 0.01 ng/g,
IQR LOD-0.04, 43% below the LOD for PFBS
measurement) with no observed association.
Kim et al. (2016)
Congenital
hypothyroidism
Excluded from full statistical analysis by study authors
because of high percent below the LOD (72%) for PFBS
measurement.
Seo et al. (2018)
Cholesterol, uric acid,
diabetes, BMI, thyroid
hormones
No consideration of potential confounding.
Shiue (2016)
Sleep disturbances
Not evaluated due to nonspecific effect.
Wane et al. (2017)
Endometriosis-related
infertility
Exposure measured concurrent with outcome for chronic
outcome; serious concerns for exposure and outcome
misclassification.
Note: LOD = limit of detection.
Shiue ('2016') was not evaluated because the outcome was sleep disturbances, which was considered a nonspecific effect, and thus
was not entered in HAWC.
28

-------

.e?
,0°
~\1	-lO^'oO^ ^av-}0"^ '>o0'

Participant selection -
Exposure measurement -
Outcome ascertainment -
Confounding -
Analysis -
Sensitivity
Selective Reporting
Overall confidence

Good (metric)

Adequate (metric)

Deficient

++
or High
+
or Medium
-
(metric) or
N/A

(overall)

(overall)

Low (overall)

Not assessed due
to critical
deficiency in
other domain
I
Critically deficient
(metric) or
Uninformative
(overall)
Figure 5. Evaluation results for epidemiological studies assessing effects of PFBS
(click to see interactive data graphic for rating rationales).
29

-------
0

1
2
Reporting
Allocation
Blinding
Variable Control
Selective Reporting & Attrition
Exposure Characterization
Utility of Study Design
Outcome Assessment -
Results Presentation
Overall confidence
+ +
+ +

+ +
+ +
+ +
+ +
+ +
+ +
+ +
+ +
+ +


+ +
+ +
+ +
+ +
+ +
+ +
NR
NR
NR
NR
NR
+ +
+ +
+ +
NR
+

Good (metric)

Adequate (metric)

Deficient

Not
H
j
Critically deficient
++
or High
+
or Medium
-
(metric) or
NR
reported

(metric) or

(overall)

(overall)

Low (overall)

for metric
J
Uninfonnative (overall)
Figure 6. Evaluation results for animal studies assessing effects of PFBS exposure
(click to see interactive data graphic for rating rationales).
30

-------
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
4.0 Evidence Synthesis: Overview of Included Studies
The database of all repeated-dose oral toxicity studies for PFBS and the related compound
K+PFBS that are potentially relevant to the derivation of RfD values includes a short-term range
finding study in rats (3M, 2000d), two 28-day studies in rats (NTP, 2019; 3M, 2001). one
subchronic-duration study in rats (Lieder et al.. 2009a; York. 2003b). one sub chronic-duration
lipoprotein metabolism study in mice (Bijland et al.. 2011; 3M. 2010). three gestational exposure
studies in mice and rats (Feng et al., 2017; York, 2003a, 2002), and one two-generation
reproductive toxicity study in rats (Lieder et al„ 2009b; York, 2003c, d, e). In addition, 19
epidemiologic studies (described in 22 publications) were identified that report on the association
between PFBS and human health effects. Specific study limitations identified during evaluation
(see HAWC) are discussed only for studies interpreted as low confidence or if a limitation
impacted a specific inference for drawing conclusions.
Human and animal studies have evaluated potential effects on the thyroid, reproductive systems,
development, kidneys, liver, and lipid and lipoprotein homeostasis following exposure to PFBS.
The evidence base for these outcomes is presented in this section. For each potential health
effect, the synthesis describes the database of human and animal studies, as well as an array of
the animal results across studies. NOAELs and LOAELs presented in figures and text are based
on statistical significance and/or biological significance (e.g., directionality of effect [statistically
significantly decreased cholesterol/triglycerides is of unclear toxicological relevance], abnormal
or irregular dose-response [nonmonotonicity], tissue-specific considerations for magnitude of
effect [nonstatistically significant increase of >10% in liver weight interpreted as biologically
significant]). A summary of the available database is presented in Table 6 of Section 5. For
information in this section, evidence to inform organ-/sy stem-specific effects of PFBS in animals
following developmental exposure is discussed in the individual organ-/system-specific sections
(e.g., reproductive cycling endpoints after developmental exposure are discussed in
"Reproductive Effects"). Other effects informing potential developmental effects (e.g., pup BW)
are discussed in the "Offspring Growth and Early Development" section.
Evidence integration analyses and overall judgments on the hazard support for each outcome
domain provided by the available human and animal studies are discussed in "Evidence
Integration and Hazard Characterization." Notably, in that section, the evidence informing organ-
dy stem-specific endpoints after developmental exposure was considered potentially informative
to both the developmental effects outcome domain and the organ-/system-specific outcome
domain.
4.1 Thyroid Effects
4.1.1 Human Studies
One low confidence study examined cross-sectional associations between PFBS exposure and
thyroid hormones in women with premature ovarian insufficiency (Zhang et al„ 2018) and
reported no association with free T3, free T4, or thyroid stimulating hormone. However, this
study had poor sensitivity and methodological limitations that make interpretation of these null
results difficult and further, the results in this highly selected population may not be
generalizable.
31

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
4.1.2 Animal Studies
Two high-confidence studies evaluated the effects of PFBS exposure on thyroid, specifically
thyroid hormone levels, thyroid histopathology, and thyroid weight (NTP, 2019; Feng et al.,
2017) (see Figure 7). Dams exposed to K+PFBS through gestation (gestation days [GDs] 1-20)
exhibited a statistically significant decrease in total triiodothyronine (T3). total thyroxine (T4).
and free T4 (reduced 17%, 21%, and 12%, respectively, relative to control at 200 mg/kg-day and
reduced 16%, 20%, and 11%, respectively, relative to control at 500 mg/kg-day) on GD 20 at
doses of 200 and 500 mg/kg-day, but not at 50 mg/kg-day (Feng et al.. 2017). Decreased total T3
and total T4 were also reported at postnatal day (PND) 1, PND 30, and PND 60 in offspring
gestationally exposed to K+PFBS at the same doses (up to 37% reduction in T3 and 52%
reduction in T4). Increased thyroid-stimulating hormone (TSH) was reported in dams and
pubertal (PND 30) offspring (21% and 14% relative to control at 200 mg/kg-day, respectively)
exposed gestationally to K+PFBS. Statistically significant dose-dependent decreases in total T3.
total T4. and free T4 were also reported after exposure in male and female rats to K+PFBS for
28 days at all doses tested (> 62.6 mg/kg-day) (NTP, 2019). The reported reductions in rat total
T3 were up to -57% (male) and -43% (female), -86% (male) and -77% (female) in free T4, and
-97% (male) and -71% (female) in total T4, respectively. Dose-response graphics for T4, T3,
and TSH, including effect size and variability, are included in appendix E, Figures E-l, E-2, and
E-3, respectively. Thyroid gland weight, thyroid histopathology, and TSH levels were not
changed after 28 days of PFBS exposure in male or female rats at up to 1,000 mg/kg-day (NTP.
2019).
Kndpoinl Name
Study Name
Study Type
Animal Description
Observation Time


PFBS Thyroid Effects

Tetraiodothyronine (T4), Free
NTP 2018, 4309741
short-term (28 days)
Rat. Harlan Sprague-Dawley (cT)
Day 28
~-W-
—V-
	~	

• Doses



Rat. Harlan Sprague-Dawley (9)
Day 28

—5VL
	~	

I—| Dose Range

Feng 2017. 3856465
developmental (GDI to 20)
P0 Mouse. ICR (9)
GD20
	
^	
	~

A Significant Increase
Tetraiodothyronine (T4). Total
NTP 2018, 4309741
short-term (28 days)
Ral, Harlan Sprague-Dawley (d1)
Day 28

—V-
	~	

V Significant Decrease



Rat. Harlan Sprague-Dawley (9)
Day 28

—57-
	~	



Feng 2017, 3856465
developmental (GDI to 20)
P0 Mouse, ICR (9)
GD20
M	
	
	V





F1 Mouse. ICR (9)
PND1
	
^	
	~






PND30
	
^	
	~






PND60
M	
^	
	w


Triiodothyronine (T3)
NTP 2018,4309741
short-term (28 days)
Rat, Harlan Sprague-Dawley (Cf)
Oay 28
»vv
—S7-
	V	





Rat. Harlan Sprague-Dawley (9)
Day 28
~ vv
—5V-
	V	
	~


Feng 2017, 3856465
developmental (GDI to 20)
P0 Mouse, ICR (9)
GD20
M	
-~—
	w





F1 Mouse, ICR (9)
PND1
	
-V—
	V






PND30
*-•	
-V—
	V






PND60
	
-V—
	w


Thyroid Stimulating Hormone (TSH)
NTP 2018, 4309741
short-term (28 days)
Rat, Harlan Sprague-Dawley {(f)
Day 28
~ » »







Rat. Harlan Sprague-Dawley (9)
Day 28
~ • •	•	•	~


Feng 2017, 3856465
developmental (GDI to 20)
P0 Mouse, ICR (9)
GD20
4-»	
-A—
	A





F1 Mouse, ICR (9)
PND1
	








PND30
M	
—
	~






PND60
	




Thyroid Weight. Absolute
NTP 2018. 4309741
short-term (28 days)
Rat. Harlan Sprague-Dawley (cT)
Day 28
~ » «


	~




Rat. Harlan Sprague-Dawley (9)
Day 28
4 • •


~

Thyroid Weight. Relative
NTP 2018,4309741
short-term (28 days)
Rat, Harlan Sprague-Dawley (Cf)
Day 28
4 • •


~




Rat, Harlan Sprague-Dawley (9)
Day 28
4 • •


	~

Thyroid Histopathology
NTP 2018, 4309741
short-term (28 days)
Rat. Harlan Sprague-Dawley (d1)
Day 28
4 • •


	4




Rat. Harlan Sprague-Dawley (9)
Day 28
4 • •


	~


3M, 2000, 4289992
short-term (10 days)
Rat, Crl: Cd (Sd) lbs Br(cT)
Day 11
4	•—


	4




Rat, Crl: Cd (Sd) lbs Br (9)
Day 11
~	•—


	~





-1
DO 0 100
200 30
0 400 500 600 700
800 900 1.000 1.
00







Dose (mg/kg/day)

Figure 7. Thyroid effects from K+PFBS exposure (click to see interactive data graphic and
rationale for study evaluations for effects on the thyroid in HAWC).
32

-------
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
4.2 Reproductive Effects
4.2.1	Human Studies
Five studies of populations in China and Taiwan examined different reproductive outcomes in
women and men (Yao et al.. 2019; Song et al.. 2018; Zhang et al.. 2018; Zhou et al.. 2017a;
Zhou et al.. 2016).
Three low-confidence studies examined reproductive hormones in newborn boys and girls in
China (Yao et al.. 2019). adolescent boys and girls in Taiwan (Zhou et al.. 2016). and adult
women in China (Zhang et al.. 2018). The study in newborns reported lower testosterone (P: -
0.23, 95% CI -0.46,0.01) and estradiol (P: -0.09, 95% CI -0.2,0.01) in cord blood in male babies,
but these differences were not statistically significant (Yao et al.. 2019). The other two studies
reported no clear associations between PFBS levels and reproductive hormones in women with
premature ovarian insufficiency (Zhang et al.. 2018) or adolescents, among the entire study
population or stratified by sex (Zhou et al.. 2016).
One low-confidence cross-sectional study (Song et al.. 2018) examined the association between
PFBS exposure and semen parameters. There was no indication of decreased semen quality in
this study (correlation coefficients of-0.022 for semen concentration and 0.195 \p < 0.05] for
progressive motility), although issues were noted regarding the ability of this study to detect an
effect and important methodological details were missing.
Two studies examined other female reproductive effects - a cross-sectional study of menstrual
cycle characteristics in a general population sample of women planning to become pregnant,
enrolled at preconception care clinics in China (Zhou et al.. 2017a) and a case-control study in
China of premature ovarian insufficiency (Zhang et al.. 2018). defined by FSH level and
oligo/amenorrhea. For any outcome related to menstruation, there is significant potential for
reverse causation because menstruation is a potential mechanisms by which PFAS are removed
from the body (Wong et al„ 2014; Zhang et al., 2013), and thus both of these studies are
considered low confidence. Although not statistically significant, (Zhou et al„ 2017a) reported
adjusted odds ratios (OR) of 1.30 (95% CI: 0.54-3.12) for menorrhagia and 1.48 (95% CI:
0.54-4.03) for hypomenorrhea in preconception women in China for each one unit increase in
PFBS. However, they also reported inverse non-statistically significant associations for these two
outcomes based on exposure quartiles (OR range: 0.61-0.84 for the highest quartiles relative to
the referent) with no evidence of an exposure-response relationship, indicating that the
associations are not robust. All of the analyses in this study examined continuous outcome
measures. Zhang et al. (2018) reported no increase in odds of premature ovarian insufficiency
with higher PFBS exposure (OR (95% CI) for second tertile vs. first: 0.84 (0.44,1.60), third
tertile: 0.92 (0.48,1.76)).
4.2.2	Animal Studies
Reproductive outcomes were evaluated in a high-confidence study of prenatal exposure to PFBS
in mice (Feng et al„ 2017), in two high-confidence gestational exposure studies in rats (York,
2003c, 2002), in high-confidence short-term and subchronic-duration studies in rats (NTP, 2019;
Lieder et al., 2009a), and in a high-confidence two-generation reproductive study in rats (Lieder
et al„ 2009b). Endpoints evaluated in these studies include fertility and pregnancy outcomes,
hormone levels, markers of reproductive development, and reproductive organ weights.
33

-------
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
4.2.2.1 Female fertility and pregnancy outcomes
Female fertility parameters were evaluated in both Feng et al. (2017) and Lieder et al. (2009b),
which reported generally no effects in exposed parents, but some effects after gestational
exposure in the F1 offspring (click to see interactive graphic for Female Fertility Effects in
HAWC). Female fertility (e.g., fertility index and days in cohabitation) and delivery parameters
(e.g., length of gestation, % deliveries, stillborn pups, and implantation sites) evaluated in Lieder
et al. (2009b) were generally unaffected by K+PFBS treatment for P0- and F1-generation dams
up to 1,000 mg/kg-day. The mean number of live born F1 pups was statistically significantly
decreased in the 30-mg/kg-day group, but this change was not dose-dependent. The viability
index in F1 pups and the lactation index in F1 and F2 pups showed statistically significant
changes at various doses but were not dose-dependent (Lieder et al.. 2009b). Similarly, no effects
were observed in delivery and litter parameters (e.g., implantations, litter sizes, live fetuses,
corpora lutea, and early resorptions) following prenatal exposure from GDs 6 to 20 (York.
2003c, 2002). Adult (PND 60) F1 females gestationally exposed to PFBS at doses greater than
200 mg/kg-day, however, exhibited fewer primordial follicles, primary follicles, secondary
follicles, early antral follicles, antral follicles, and preovulatory follicles, as well as fewer corpora
lutea compared to control (Feng et al„ 2017). Importantly, no effects on the health (e.g., weight
gain) of the exposed dams were observed at any dose (Feng et al., 2017). Lieder et al. (2009b)
evaluated ovarian follicles in F1 females after they were mated and their pups had been weaned
(i.e., lactation day [LD] 22), and observed no effects compared to controls at 1,000 mg/kg-day;
however, the data were not reported. These parameters were not evaluated in York (2002).
4.2.2.2 Male fertility
Two studies using S-D rats evaluated several potential responses in the male reproductive system
(NTP, 2019; Lieder et al„ 2009b). Male fertility parameters and reproductive effects (e.g., sperm
parameters) were generally unaffected by K+PFBS treatment in P0- and F1-generation males
observed by Lieder et al. (2009b). At the highest dose, there were statistically significant
increases in the percentage of abnormal sperm in F1 animals and decreases in testicular sperm
count in PO-generation males. In addition, the study authors report the number of spermatids per
gram testis was within the historical control of the testing facility. These effects were not
statistically changed at lower doses. Alterations in parameters such as sperm count/number and
morphology are considered indicative of adverse responses in the male reproductive system
(Foster and Gray, 2013; Mangelsdorf et al„ 2003; U.S. EPA, 1996a). A 28-day exposure study
reported a decreased trend in testicular spermatid count per mg testis evaluated at the time of
necropsy; however, no significant effects on other sperm measures were reported, including
caudal epididymal sperm count and sperm motility (NTP, 2019). It should be noted that a
complete spermatogenesis cycle in male rats is typically 7 weeks in length, thus study designs of
shorter duration could potentially miss effects of chemical exposure on some sperm parameters.
As such, the differences in responses observed in the two available studies might have been due
to experimental design differences as Lieder et al. (2009b) exposed P0 animals for 70 days and
F1 animals during the entire period of gestation plus lactation, whereas NTP (2019) exposed
animals for 28 days. Future studies should be developed to ascertain whether long-term and/or
gestational exposure to PFBS significantly affects sperm measures in sexually mature and
developing animals.
34

-------
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
4.2.2.3 Reproductive hormones (female and male)
Reproductive hormones were evaluated in mice (Feng et al., 2017) and, to a limited extent, in
rats (NTP. 2019) (see Figure 8). Exposure to K+PFBS for 28 days resulted in a significant trend
for increased testosterone levels in females, but not in males (NTP. 2019). The increase in
testosterone was not statistically significant when compared to control at any dose by pairwise
analysis. Prenatal exposure to PFBS at and above 200 mg/kg-day resulted in statistically
significant reduced serum estradiol levels and increased serum luteinizing hormone levels in
pubertal offspring (i.e., PND 30) (Feng et al., 2017). The change in serum estradiol levels, but
not luteinizing hormone, continued into adulthood in the K+PFBS-exposed offspring
(i.e., PND 60). Adult PFBS-exposed offspring also exhibited decreased serum progesterone
levels at doses of 200 mg/kg-day and greater. PFBS exposure did not alter maternal estradiol-,
progesterone-, or gonadotropin-releasing hormone. Reproductive hormone levels in males and
females were not evaluated by Lieder et al. (2009b). The changes in follicle and corpora lutea
development reported in the same study, however, may be associated with alterations in hormone
production/levels, as ovarian follicles and corpora lutea produce estrogen and progesterone,
respectively (Foster and Gray. 2013; U.S. EPA. 1996a).
The hormonal effects observed in the NTP (2019) and Feng et al. (2017) studies might be
associated with adverse reproductive effects reported in these studies. Androgens, luteinizing
hormone, estradiol, and progesterone play an important role in normal development and
functions of the female reproductive system (Woldemeskel. 2017; Foster and Gray. 2013).
Alterations in the levels and production of these reproductive hormones can disrupt endocrine
signals at the hypothalamic-pituitary level and lead to delayed reproductive development and
changes in functions (Rudmann and Foley. 2018; Woldemeskel. 2017; Foster and Gray. 2013).
Endpoint
Study
Exposure
Animal Group
Observation time
PFBS Reproductive Hormone Effects
Testosterone (T)
NTP 2018,4309741
28 Day Oral
Rat, Harlan Sprague-Dawley (9)
Day 28







Rat. Harlan Sprague-Dawley (cf)
Day 28
_ _ _ _ _ ^







Estrogen
Feng 2017, 3856465
20 Day Oral Gestation
PO Mouse. ICR (9)
GD20
	•	~
• Doses




Fl Mouse. ICR (9)
PNDI

A Significant Increase

PND30
~ Y	~	~
^7 Significant Descrease

PND60
	~	~
|—| Dose Range

Progesterone (P4)
Feng 2017, 3856465
20 Day Oral Gestation
P0 Mouse, ICR (9)
GD20
	•	»




Fl Mouse. ICR (9)
PNDI
	•	•

PND30


PND60
+~	~	~

Luteinizing Hormone (LH)
Feng 2017,3856465
20 Day Oral Gestation
Fl Mouse, ICR (9)
PNDI


PND30
•A	A	~

PND60
	~	~

Gonadotropin Releasing Hormone (GnRH)
Feng 2017, 3856465
20 Day Oral Gestation
PO Mouse, ICR (9)
GD20
• •	•	~




Fl Mouse, ICR (9)
PNDI


PND30
	•	~


PND60




»0 0 100 200 300 400 500 600
Dose (mg/kg/day)
700 800 900 1,000 1,
00
Figure 8. Reproductive hormone response to K+PFBS exposure (click to see interactive
data graphic and rationale for study evaluations for reproductive hormone levels in
HAWC).
35

-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
4.2.2.4 Reproductive system development, including markers of sexual differentiation and
maturation (female and male)
Several measures of female reproductive development were affected by gestational K+PFBS
exposure in mice (see Figure 9). Feng et al. (2017) reported a delayed first estrous in female
PFBS-exposed offspring (> 200 mg/kg-day) compared to control. Estrous cyclicity was also
affected in K+PFBS-exposed PNDs 40-60 offspring as exhibited by a prolongation of the
diestrus stage compared to control. Estrous cycling was generally not statistically significantly
altered in P0- or F1-generation females treated with K+PFBS in the two-generation study by
Lieder et al. (2009b). An increase in the number of rats with > 6 consecutive days of diestrus was
observed in the F1 females exposed to 100 mg/kg/day; however, the increase was not present at
higher doses (Lieder et al.. 2009b). Estrous cyclicity was affected after adult exposure to
K+PFBS for 28 days exhibited by a dose-dependent prolongation of diestrus at doses of
250 mg/kg-day and greater with marginal significance at the lowest dose tested (125 mg/kg-day)
(p = 0.063) (NTP. 2019). Lieder et al. (2009b) reported a delay in the days to preputial separation
in F1 males of the 30- and 1,000-mg/kg-day groups;9 however, the measure was no longer
statistically significant when adjusted for BW. There was similarly no change in the days to
vaginal patency in F1 female rats (Lieder et al.. 2009b). Unlike Lieder et al. (2009b). Feng et al.
(2017) reported a delay in vaginal patency in F1 females after gestational exposure of
200 mg/kg-day and greater.
Kndpoinl Name
Study Name
Experiment Name
Animal Description
Observation Time
PFBS Reproductive Development and Estrous Cycling
>6 Days of Diestrus
Lieder. 2009, 1578545
2 Generation Oral
F1 Rat. Sprague-Dawley (9)

m-A	•	~
• Doses



P0 Rat, Sprague-Dawley (9)

M—•	1	1
|—| Dose Range
>6 Days of Estrus
Lieder. 2009, 1578545
2 Generation Oral
F1 Rat, Sprague-Dawley (9)

M—¦	¦	•
A Significant Increase



P0 Rat, Sprague-Dawley (9)
6 days
H—•	•	~
V Significant decrease
hstrous Cycle, Diestrus
Feng 2017, 3856465
20 Day Oral Gestation
F1 Mouse, ICR (9)
PND40-60
~-	A	A


NTP 2018, 4309741
28 Day Oral
Rat, Harlan Sprague-Dawley (9)
Day 28
~ » *	A	A	~

Estrous Cycle. Estrus
Feng 2017. 3856465
20 Day Oral Gestation
F1 Mouse. ICR (9)
PND40-60
	•	~


NTP 2018, 4309741
28 Day Oral
Rat. Harlan Sprague-Dawley (9)
Day 28
» • •	•			~

Estrous Cycle. Mctcstrus
NTP 2018. 4309741
28 Day Oral
Rat. Harlan Sprague-Dawley (9)
Day 28
~ » »	•	•	*

Estrous Cyelc. Proesirus
Feng 2017. 3856465
20 Day Oral Gestation
F1 Mouse. ICR (9)
PND40-60
	•	~


NTP 2018. 4309741
28 Day Oral
Rat. Harlan Sprague-Dawley (9)
Day 28
~ • •	1	•	~

Estrous Stage, Diestrus (at sacrifice)
l.ieder, 2009, 1578545
2 Generation Oral
F1 Rat, Sprague-Dawley (9)
Sacrifice
M—¦	•	•

HO Rat. Sprague-Dawley (9)
M » « •

Estrous Stage, Estrus (at sacrifice)
Lieder, 2009. 1578545
2 Generation Oral
F1 Rat. Sprague-Dawley (9)
Sacrifice
M—•	•	•

P0 Rat, Sprague-Dawley (?)
M—•	•	~

Estrous Stage. Mclcslrus (at sacrifice)
Lieder. 2009. 1578545
2 Generation Oral
F1 Rat. Sprague-Dawley (9)
Sacrifice
M—•	•	•

P0 Rat, Sprague-Dawley (9)
•	•	•

Estrous Stage. Proesirus (al sacrifice)
Lieder. 2009. 1578545
2 Generation Oral
F1 Rat. Sprague-Dawley (9)
Sacrifice
•	•	»

P0 Rat, Sprague-Dawley (9)
	•	~

Estrous Stages/ 21 Days
Lieder. 2009, 1578545
2 Generation Oral
F1 Rat, Sprague-Dawley (9)

M—•	•	~

P0 Rat, Sprague-Dawley (9)
~ » » •	~

Firsl Estrous - Litter N
Feng 2017. 3856465
20 Day Oral Gestation
F1 Mouse, ICR (9)
Beginning on PND24
M	A	A

Preputial .Separation
Lieder, 2009, 1578545

Fl Rat, Sprague-Dawley (Cf)
Day 29 (postpartum)





Vaginal Opening
Feng 2017. 3856465
20 Day Oral Gestation
F1 Mouse. ICR (9)
PND60
M	a	A

Vaginal Patency
Lieder. 2009. 1578545
2 Generation Oral
Fl Rat. Sprague-Dawley (9)
Postpartum Day 28
M—•	1	~

	1	1	1	1	1	1	1	1	1	1	1	
-100 0 100 200 300 400 500 600 700 800 900 1,000 1,100
Dose (mg/kg/day)
Figure 9. Effects to reproductive development and estrous cycling following PFBS exposure
(click to see interactive data graphic).
4.2.2.5 Reproductive organ weights and histopathology (female and male)
Studies have not consistently reported changes in reproductive organ weights (click to see
interactive graphic for Reproductive Organ Effects in HAWC). Reproductive organ weights,
9A marker of delayed reproductive development (Foster and Gray. 2013; U.S. EPA. 1996b).
36

-------
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
including testes, ovaries, and uterus, were unchanged in the two-generation reproductive study in
P0 and F1 males and females (Lieder et al.. 2009b) and following short-term and subchronic
exposure to K PFBS (NTP. 2019; Lieder et al.. 2009a: 3M. 2001. 2000d). F1 females
gestationally exposed to PFBS, however, exhibited decreased size and weight of the ovaries and
uterus (Feng et al.. 2017). In addition, the total uterine section diameter and endometrial and
myometrial thickness were significantly reduced. There were no significant histopathological
alterations in the male or female reproductive organs evaluated following exposure to K+PFBS
for 28 days (NTP. 2019) or in parental or offspring from the two-generation reproductive study
(Lieder et al.. 2009b).
4.3 Offspring Growth and Early Development
4.3.1 Human Studies
No human studies were available to inform the potential for PFBS exposure to cause effects on
the growth or early development of children.
4.3.2 Animal Studies
Evidence to inform organ-/sy stem-specific effects of PFBS in animals following developmental
exposure are discussed in the individual hazard sections (e.g., reproductive cycling after
developmental exposure is discussed in "Reproductive Effects"). This section is limited to
discussion of other, specific developmental effects commonly evaluated in guideline
developmental toxicity studies, including pup BW, developmental markers, and bone measures.
Four high- or medium-confidence studies examined potential alterations in offspring growth and
early development following PFBS exposure, including two gestational exposure studies in rats
(York. 2003a. 2002) and one gestational exposure study in mice (Feng et al.. 2017). as well as a
two-generation study in rats (Lieder et al.. 2009b; York. 2003c) (click to see interactive graphic
for Developmental Effects in HAWC).
None of the studies identified significant effects in either rats or mice on measures of fetal
morphology (i.e., malformations and variations). BW of female offspring of PFBS-exposed mice
at doses greater than 200 mg/kg-day was statistically significantly lower than control at PND 1,
and the pups remained underweight through weaning, pubertal, and adult periods, with decreases
of approximately 25% observable in pups nearing weaning (Feng et al.. 2017). At around
PND 16, Feng et al. (2017) also reported an ~1.5-day developmental delay in eye opening in
pups gestationally exposed to 200 mg/kg-day PFBS and greater. Importantly, no effects on the
health of the exposed dams (e.g., weight gain) were observed at any dose (Feng et al.. 2017).
Dose response graphics for eye opening, including effect size and variability, are included in
appendix E, Figure E-4. Fetal BWs (male and female) were also reduced (approximately 10%)
compared to controls following gestational exposure from GDs 6 to 20 at the highest tested dose
(1,000 mg/kg-day in York (2002)1 and 2,000 mg/kg-day in York (2003a)l). Parental BWs and
organ weights, however, were also affected to a similar degree at those doses (Lieder et al..
2009b; York. 2003 c. 2002). limiting the interpretation of the results. No statistically significant
changes in Fl - and F2-generation pups mean pup weight at birth and mean pup weight at
weaning were reported by Lieder et al. (2009b) or York (2003c).
37

-------
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
Several measures of thyroid hormone development and female reproductive development were
affected by gestational PFBS exposure in mice and are described in more detail in "Thyroid
Effects" and "Reproductive Effects," respectively.
4.4 Renal Effects
4.4.1 Human Studies
One low-confidence study (Pin et al. (2016). with additional details in Bao et al. (2014). selected
225 subjects ages 12-15 years old from a prior cohort study population in seven public schools
in northern Taiwan (Tsai et al.. 2010) and examined the association between PFBS exposure and
uric acid concentrations. There was no association between ln(PFBS) concentration and uric acid
concentrations in the total population (P = 0.0064 mg/dL increase in uric acid per 1 ln-|ig/L
increase in PFBS, 95% CI = -0.22, 0.23). EPA identified that a non-significant positive
association in boys was offset by a non-significant negative association in girls, and there is not
enough information to determine whether there is an interaction with sex. When PFBS exposure
was analyzed for high uric acid (> 6 mg/dL), the risk was somewhat elevated in boys
(OR = 1.53; 95% CI: 0.92, 2.54), but not in girls (OR = 0.99; 95% CI: 0.58, 1.73). The potential
for reverse causation (i.e., that renal function could influence the levels of PFBS in the blood)
tempers any conclusions that might be able to be drawn.
4.4.2 Animal Studies
Renal effects were evaluated in high-confidence short-term and subchronic-duration exposure
studies in rats (NTP. 2019: Lieder et al.. 2009a: 3M. 2001. 2000d) and in a high-confidence
two-generation reproductive study in rats (Lieder et al.. 2009b). Endpoints evaluated in these
studies include kidney weights, histopathological changes, and serum biomarkers of effect (see
Figure E-8 and Figure E-9). Dose-response graphics for histopathological effects, including
effect size and variability, are included in appendix E, Figure E-7.
Absolute and relative kidney weights of male and female S-D rats were unchanged in rats
exposed daily for 90 days to K+PFBS at doses up to 600 mg/kg-day compared to control rats
(Lieder et al., 2009a). This lack of effect on kidney weight was also observed in parental and F1
male and female rats of the same strain exposed to K+PFBS at doses up to 1,000 mg/kg-day
during a two-generation reproductive study (Lieder et al.. 2009b). Although none of the findings
reached statistical significance, however, an approximate 9% increase in absolute kidney weight
was observed in female S-D rats exposed to 1,000 mg/kg-day K+PFBS for 10 days (3M. 2000d):
relative-to-body kidney weights were also increased approximately 6%~9%. This organ-weight
effect was not observed in corresponding males of the study. In a follow-on 28-day study by the
same lab, a 9%—11% increase in absolute and relative-to-body kidney weight was observed in
female S-D rats exposed to 900 mg/kg-day K+PFBS (3M. 2001). although these changes were
not statistically significant. In this study, EPA also observed that smaller non-significant
increases in kidney weight occurred in male rats. In another 28-day study, K+PFBS exposure
significantly increased absolute and relative right kidney weights in high-dose male
(500 mg/kg-day) S-D rats (NTP, 2019). Only relative-to-BW kidney weights were altered in
female rats; but this effect was significant at all tested K+PFBS doses (> 62.6 mg/kg-day). Click
to see interactive graphic for Kidney Weight Effects in HAWC.
38

-------
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
After 90 days of exposure, Lieder et al. (2009a) observed increased incidences of
histopathological alterations of the kidneys of male and female rats of the high-dose group
(600 mg/kg-day). Increased incidence of hyperplasia of the epithelium of renal papillary tubules
and ducts was observed in rats of both sexes (see Figures E-7 and E-8). A single incidence of
papillary necrosis in both kidneys was observed in one male in the high-dose group. Further,
focal papillary edema was observed in 3/10 rats of both sexes of the high-dose groups compared
to no evidence of this effect in control rats. Similar histopathological alterations were observed
in parental and F1 male and female rats in the two-generation reproduction study (Lieder et al..
2009b). Compared to control rats, increased incidences of hyperplasia of the renal tubular and
ductal papillary epithelium, and focal papillary edema were observed in parental male and
female rats at PFBS doses > 300 mg/kg-day. Hyperplastic foci in the same locations of the
kidney were also observed in male and female F1 rats exposed to > 300 mg/kg-day PFBS across
life stages from gestation to adulthood (Lieder et al., 2009b). Focal papillary edema was
observed in male (> 1,000 mg/kg-day) and female (>300 mg/kg-day) F1 rats, although this
specific alteration did not appear to be dose-dependent in females. Although kidney alterations
such as hydronephrosis, mineralization, and tubular degeneration were observed in male or
female S-D rats after just 10 days of oral K+PFBS exposure, these effects were not significant
compared to control and/or did not appear to be dose-dependent (3M, 2000d). The same
histopathological lesions were noted in the 28-day rat study albeit with lack of significance
compared to control or dose-dependence (3M, 2001). In another 28-day oral gavage study in S-D
rats, chronic progressive nephropathy (CPN) was observed in all male and female PFBS
treatment groups and control rats, with no evidence of dose-dependence for this effect (NTP.
2019). Renal papillary necrosis was also observed in these rats but only at the highest exposure
dose (1,000 mg/kg-day).
Serum levels of biomarkers indicative of kidney injury and/or function, including blood urea
nitrogen (BUN) and creatinine, have been examined across multiple studies of varying exposure
durations, and were found to be unchanged in male and female rats treated with K+PFBS at doses
up to 1,000 mg/kg-day (Lieder et al.. 2009a; 3M. 2001. 2000d). After 28 days of oral gavage
exposure in S-D rats, however, NTP (2019) observed significantly increased levels of BUN in
males (> 250 mg/kg-day). This increased circulating BUN was not observed in female rats at
doses up to 1,000 mg/kg-day. Click to see interactive graphic for other Kidney Effects in
HAWC.
4.5 Hepatic Effects
4.5.1 Human Studies
No human studies were available to inform the potential for PFBS exposure to cause hepatic
effects.
4.5.2 Animal Studies
Hepatic effects were evaluated in high-confidence short-term and sub chronic-duration studies in
rats (NTP, 2019; Lieder et al., 2009a; 3M, 2001, 2000d) and in a high-confidence two-generation
reproductive study in rats (Lieder et al., 2009b). Endpoints evaluated in these studies include
liver weights, histopathological changes, and serum biomarkers of effect (see Figure E-10).
39

-------
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
Ten days of daily oral gavage exposure to K+PFBS significantly increased absolute,
relative-to-body, and relative-to-brain weights of liver in adult male and female S-D rats exposed
to 1,000 mg/kg-day (3MI_2000d). The absolute liver mass of male rats was increased by 36%
compared to females (22%). A similar profile of liver weight alteration in S-D rats was observed
following 28 days of exposure where absolute and relative liver weights of high-dose
(900 mg/kg-day) male rats were increased 25%—30% (3M. 2001). Female rats of the same
treatment dose did not experience a similar magnitude increase in absolute or relative liver
weights (4%—6%). In another 28-day study in S-D rats, K+PFBS exposure significantly increased
absolute and relative liver weights in males (> 125 and > 62.6 mg/kg-day, respectively) and
females (> 250 and > 125 mg/kg-day, respectively) (NTP. 2019). In contrast, the livers of male
and female S-D rats exposed to K+PFBS at doses up to 600 mg/kg-day for 90 days were not
significantly changed compared to respective controls (Lieder et al.. 2009a). In a two-generation
reproduction study using the same strain of rat, however, increased absolute and relative liver
weights were observed in male parental rats exposed to doses of K+PFBS > 300 mg/kg-day for
approximately 70 days (Lieder et al.. 2009b). In the F1 adult males, only relative liver weight
was significantly increased at the high dose (1,000 mg/kg-day), although terminal BW was
significantly decreased in this group compared to control.
Histopathological examination of the livers of S-D rats across three separate oral gavage studies
of increasing K+PFBS exposure duration (10-day (3M. 2000d)l; 28-day (3M. 2001)1; 90-day
(Lieder et al.. 2009a)l) did not reveal any significant dose-dependent alterations or lesions. For
example, focal/multifocal hepatic inflammation was observed in 3/10 male and 4/10 female rats
of the high-dose group (no incidence at the low- or mid-dose) compared to 6/10 male and female
rats in the control groups (Lieder et al.. 2009a). The Lieder et al. (2009b) two-generation
reproduction oral gavage study did identify increased incidences of hepatocellular hypertrophy in
parental and F1 adult male rats at > 300 mg/kg-day; however, this effect was absent in female
rats at doses of K+PFBS up to 1,000 mg/kg-day. NTP (2019) identified significantly increased
incidence of hepatocellular hypertrophy in male (> 125 mg/kg-day) and female (>
500 mg/kg-day) S-D rats after 28 days of K+PFBS exposure. Further, significantly increased
cytoplasmic alteration of hepatocytes was observed in these rats (male and female at >
500 mg/kg-day). Hepatic necrosis was also observed but was not significant compared to control
and only occurred at the high dose (1,000 mg/kg-day) in both sexes (NTP. 2019).
In general, serum biomarkers associated with altered liver function or injury, including alanine
aminotransferase (ALT) and aspartate aminotransferase (AST), were not significantly changed in
male and female S-D rats across multiple oral gavage studies of varying exposure durations up to
90 days, at K+PFBS doses up to 1,000 mg/kg-day (Lieder et al„ 2009a; 3M, 2001, 2000d). NTP
(2019). however, reported increased serum ALT and AST in male (500 mg/kg-day only) and
female (> 250 mg/kg-day for ALT; > 500 mg/kg-day for AST) rats exposed to K+PFBS for
28 days. Click to see interactive graphic for Liver Effects in HAWC.
4.6 Lipids and Lipoproteins
4.6.1 Human Studies
One low-confidence study (Zeng et al.. 2015) used the controls from the case-control study of
asthma described below (Dong et al.. 2013a) and examined the association between PFBS
exposure and serum lipids. There was a statistically significant increase in total cholesterol
40

-------
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
(P = 19.3 mg/dL increase per 1 jj.g/1 increase in PFBS, 95% CI = 0.6-38.0) but when PFBS
exposure was analyzed in quartiles, no exposure-response gradient was observed.
In addition, a medium confidence birth cohort in China examined associations with childhood
adiposity (Chen et al.. 2019). PFBS was measured in cord blood samples and several measures of
adiposity were collected at age 5. There was higher adiposity with higher exposure in girls, with
significant exposure-response relationships across tertiles with waist circumference, fat mass,
body fat percentage and waist to height ratio. No association with adiposity was observed in
boys. It is unlikely that the association in girls can be explained by confounding across the other
PFAS measured in this study as the associations were strongest for PFBS, but it is possible that
there is other unmeasured confounding.
4.6.2 Animal Studies
Beyond a single medium-confidence mouse study (Bijland et al.. 2011; 3M. 2010); summarized
below), PFBS studies have not particularly focused on perturbations in lipids or lipoproteins as a
potential health outcome, as studies have typically focused only on measures of serum
cholesterol and triglyceride as part of a broader panel of clinical chemistry measures in high- or
medium-confidence rat studies of 10, 28, and 90 days (see Figure E-ll) (3M (2000d)1; 3M
(2001)1; and Lieder et al. (2009a)"I. respectively). Circulating levels of cholesterol and
triglycerides were unchanged in male and female S-D rats following daily oral gavage exposure
to K+PFBS for 10 days at doses up to 1,000 mg/kg-day (3M, 2000d). In a similarly designed
study from the same laboratory, serum cholesterol and triglyceride levels were decreased in male
rats but at the high dose only, and, this effect was not statistically significant compared to control
nor was this effect observed in female rats of the same dose group (3M. 2001). Following
exposure for up to 90 days, cholesterol and triglycerides were unchanged in male and female rats
at doses up to 600 mg/kg-day (Lieder et al.. 2009a). PFBS was included in a multi-PFAS study
specifically designed to interrogate the mechanism of effect on lipid and lipoprotein metabolism
in a transgenic mouse line (APOE*3-Leiden CETP) that is highly responsive to fat and
cholesterol intake, consistent with human populations exposed to a western-type diet (containing
14% beef tallow, 1% corn oil, and 0.25% cholesterol) (Bijland et al.. 2011; 3M. 2010). Adult
male mice were fed a western-type, high-fat diet for 4 weeks prior to initiation of PFBS exposure
and throughout the 4-6 weeks PFBS exposure period (at approximately 30 mg/kg-day). This
study included several measures of lipid and lipoprotein synthesis, modification, and transport or
clearance such as circulating plasma levels, in vivo clearance of very low density lipoprotein
(VLDL)-like particles, fecal bile acid and sterol excretion, hepatic lipid levels, lipase activity,
VLDL-triglyceride and VLDL-apoB production, and gene expression profiles. After 4 weeks of
PFBS exposure, fasting plasma triglycerides, cholesteryl ester transfer protein, and glycerol were
significantly decreased compared to mice on the control diet. Further, the half-life of VLDL-like
particles and hepatic lipase activity, and hepatic cholesteryl ester and free cholesterol levels were
decreased (Bijland et al.. 2011; 3M. 2010). Hepatic uptake of VLDL-like particles (represents
fatty acid/lipid transport into hepatic tissue) was modestly, but significantly increased compared
to control mice. This increased hepatic lipid uptake in the liver was accompanied by increased
expression of genes associated with lipid binding, activation, and metabolism (e.g., P-oxidation).
41

-------
1	4.7 Other Effects
2	4.7.1 Human Studies
3	Two studies in China examined different immune outcomes in children (Chen et al.. 2018; Dong
4	et al.. 2013a).
5	One medium-confidence study reported in five publications (Pin et al.. 2017; Zhou et al.. 2017b;
6	Zhou et al.. 2017a; Zhu et al.. 2016; Dong et al.. 2013b) examined the association between PFBS
7	exposure and asthma, asthma symptoms, pulmonary function, and related immune markers (IgE,
8	absolute eosinophil count [AEC], eosinophilic cationic protein [ECP], T-helper cell-specific
9	cytokines, and 16-kDa club cell secretory protein). The primary finding was a statistically
10	significant (in the fourth quartile) positive association between incident asthma (i.e., diagnosis in
11	the previous year) and PFBS exposure (OR [95% CI] for Q2: 1.3 [0.7, 2.3], Q3: 1.2 [0.7, 2.2],
12	Q4: 1.9 [1.1, 3.4]). There were also increases in AEC and ECP with increased exposure (not
13	statistically significant with the exception of AEC in children with asthma). There was no clear
14	association with IgE or T-helper cell-specific cytokines. There was also no clear association with
15	asthma severity or control of asthma symptoms (Dong et al„ 2013a), or pulmonary function
16	measured with spirometry among children with asthma (Oin et al.. 2017). While pulmonary
17	function could be considered an outcome separate from asthma, the authors noted no associations
18	in pulmonary function (i.e., in nonasthmatics across the PFAS they studied), so for these
19	purposes, it was considered an indicator of asthma severity.
20	One medium-confidence study (Chen et al.. 2018) examined the association between PFBS
21	exposure and atopic dermatitis and reported a nonstatistically significant increase in atopic
22	dermatitis with increased exposure (OR: 1.23, 95% CI: 0.74-2.04).
23	In addition, two studies examined cardiovascular effects (Huang et al.. 2019b; Huang et al..
24	2018) but it is difficult to evaluate consistency across studies given the different outcomes in
25	each.
26	One medium confidence study (Huang et al.. 2018) using data from NHANES cycles for 1999-
27	2014 reported significantly higher odds of total cardiovascular disease with higher exposure (OR
28	(95%) CI) for above vs. below the LOD: 1.19 (1.06,1.32)) and elevated, though not statistically
29	significant odds of individual types of cardiovascular disease (congestive heart failure, coronary
30	heart disease, angina pectoris, heart attack, and stroke). There is potential in this study by
31	confounding across the PFAS, as PFBS was highly correlated with some other PFAS with
32	slightly stronger associations.
33	A medium confidence cross-sectional study (Huang et al.. 2019b) of hypertensive disorders of
34	pregnancy reported higher odds of all hypertensive disorders in pregnancy (in the third tertile)
35	(OR (95% CI) for tertile 2 vs. 1: 0.89 (0.39,2.44), tertile 3: 2.26 (1.02,5.02), p-trend 0.03) and
36	preeclampsia (tertile 2 vs. 1: 2.09 (0.51,8.53), tertile 3: 3.51 (0.94,13.2), p-trend 0.05), with both
37	trends being statistically significant after mutual adjustment of PFAS.
38	4.7.2 Animal Studies
39	Other effects were evaluated following exposure to PFBS, including outcomes related to the
40	spleen, hematological system, BW, neurotoxicity, and nonspecific clinical chemistry. These
42

-------
1	groups of outcomes were not synthesized due to inadequate available information, uncertain
2	biological relevance, and/or inconsistencies across studies and sexes.
3	4.8 Other Data
4	Other studies that used PFBS or K+PFBS are described in this section. These studies are not
5	adequate for the determination of RfD values and were considered supportive data. These data
6	might include acute duration exposures, genotoxicity, mechanistic, and other studies
7	(see Table 5).
43

-------
Table 5. Other studies
Test
Materials and methods
Results
Conclusions
References
Genotoxicity
Mutagenicity
test
Salmonella typhimurium (strains TA98 and TA100) and
Escherichia coli (E. coli) (strain pKMlOl) in the
presence or absence of S9. Concentrations of PFBS were
between 0-5,000 (ig/plate.
Test was negative for TA100 and pKMlOl
strains and equivocal for TA98 strain.
There is no in vitro
evidence of PFBS
mutagenicity.
NTP (2005)
Ames
S. typhimurium (strains TA98, TA100, TA1535, and
TA1537) and E. coli (strain WP2uvrA) were tested in
the presence or absence of S9 and with or without a
preincubation treatment. Concentrations of K+PFBS
were between 0-5,000 (ig/plate.
The results of both mutation assays
indicate that PFBS did not induce any
significant increase in the number of
revertant colonies for any of the tester
strains in the presence or absence of
induced rat liver S9.
There is no in vitro
evidence of PFBS
mutagenicity.
Pant (2001)
Genotoxicity
test
Human hepatoma (HepG2) cells were treated with
0.4 (iM to 2 mM PFBS. Intracellular ROS production
was measured by use of 2',7'-dichlorofluorescein
diacetate and DNA damage was measured with the
comet assay.
The amount of ROS and DNA strand
breaks remained unaffected by PFBS
treatment.
PFBS did not generate
ROS or DNA damage in
human liver cells.
Eriksen et al.
(2010)
CHO
chromosomal
aberration
Cultures of CHO cells were treated with K+PFBS at
concentrations ranging from 0 to 5,000 (ig/mL with or
without exogenous metabolic activation. The in vitro
exposure duration was 3 hr.
PFBS did not induce a statistically
significant increase in the percentage of
cells with aberrations at any of the
concentrations tested, either with or
without metabolic activation, in either
assay when compared to the solvent
controls.
Based on the negative
results in the in vitro CA
assay in CHO cells, PFBS
is not considered to be a
clastogenic agent.
Xu (2001)
Micronucleus
assay
Male and female S-D rats (5/group) were exposed twice
daily to K+PFBS by oral gavage at doses of 31.3, 62.5,
125, or 250 mg/kg for 28 d.
PFBS did not induce a statistically
significant increase in the frequency of
micronucleated polychromatic
erythrocytes.
PFBS was negative for
micronuclei in the blood
of male and female rats,
indicating a lack of
genotoxic potential.
NTP (2012)
Acute duration and other routes of exposure
Acute
10 rats/group, young adult male rat (strain not
specified), administered PFBS by gavage, single dose,
50, 100, 300, 600, or 800 (iL/kg and observed for 14-d
postexposure.
Mortality: 0%, 20%, 60%, 80%, and 100%
at 50, 100, 300, 600, and 800 (iL/kg PFBS,
respectively.
Acute oral PFBS rat LD5o
in male rats is 236 (iL/kg
(corresponding to
430 mg/kg).
Bomhard and
Loser (1996)
Low
confidence
44

-------
Test
Materials and methods
Results
Conclusions
References
Acute dermal
Adult (8 wk of age) male and female S-D rats (5/group)
were exposed dermally (10% of body surface area) to
500, 1,000, or 2,000 mg/kg K+PFBS for 24 hr and then
observed for 15-d postexposure for signs of clinical
toxicity, mortality, BW changes, or gross pathology
(terminus of study).
No treatment-related observations were
noted.
PFBS is not acutely toxic
via the dermal route of
exposure in rats.
3M (2000b)
Dermal irritation
Adult (14-wk of age) female NZW rabbits (3 rabbits
total for study) were exposed dermally (6 cm2 of skin) to
500 mg K+PFBS for approximately 4 hr and then
observed for 9-d postexposure for signs of clinical
toxicity, mortality, or BW changes.
Draize scoring was performed on the patch
site immediately following the exposure
period and 24, 48, and 72 hr postexposure.
No signs of dermal irritation were
observed. No signs of clinical toxicity or
mortality occurred. No treatment-related
alterations in BW were noted.
PFBS did not induce
erythema, edema, or other
possible dermal findings
during the scoring periods,
indicating a lack of dermal
irritant properties in
rabbits.
3M (2000a)
Ocular
sensitivity
Adult (16-wk of age) female NZW rabbits (3 rabbits
total for study) were exposed to approximately 80 mg
K+PFBS via ocular installation in the left eye for 2 sec.
Eyes were flushed with 0.9% saline after 24 hr and then
observed and scored for up to 21-d postexposure. The
rabbits were also followed for clinical signs of toxicity
or mortality/moribundity.
Excessive lacrimation of the left eyes
noted throughout study postexposure.
Based on the laboratory scoring system,
PFBS was "moderately" irritating at 24
and 72 hr postexposure.
PFBS is a moderate ocular
irritant in rabbits.
3M (2000c)
Contact
hypersensitivity
Adult male (10-12 wk old) and female (9 wk old)
CRL:(HA)BR Hartley guinea pigs were injected
intradermally with sterile water, Freund's adjuvant, or
adjuvant containing 125 mg/mL K+PFBS (induction
phase). Day 7 after induction, a petrolatum paste
containing 0.5 g K+PFBS was applied to the previous
injection site of the guinea pigs for 48 hr (topical
induction phase). Day 22, a challenge dose of 0.5 g
K+PFBS (petrolatum paste) was applied to the shaved left
cranial flank (right flanks were treated with petrolatum
paste only) (challenge phase). This challenge procedure
was repeated on Day 29. Challenge sites were observed
and scored following each challenge period (days 24-25
males and females and days 31-32 males only). Guinea
pigs were also followed for signs of clinical toxicity,
mortality/moribundity, or alterations in B W.
No mortalities, clinical signs of toxicity, or
changes in BW associated with PFBS
exposure. Dermal scores were zero (no
response) in females and did not exceed 1
in males (discreet or patchy edema), which
was not considered significant compared
to control guinea pigs exposed to Freund's
adjuvant alone.
PFBS is not considered an
allergen in the guinea pig
maximization test.
3M (2002a)
CA = chromosomal aberration; CHO = Chinese hamster ovary; cm2 = square centimeters; d = day(s); DNA = deoxyribonucleic acid; LD50 = median lethal dose; |xg/plate =
microgram per plate; |xM = micromol; mM = millimol; NZW = New Zealand White; ROS = reactive oxygen species; wk = weeks(s).
45

-------
4.8.1	Tests Evaluating Genotoxicity and Mutagenicity
Genotoxic, mutagenic, and clastogenic effects of PFBS have been tested in mammalian and
prokaryotic cells in vitro (Eriksen et al.. 2010; NTP. 2005; Pant 2001; Xu. 2001). and in rats in
vivo (NTP, 2019). PFBS was negative for mutagenicity in Escherichia coli (E. coli) strain
pKMlOl and Salmonella typhimurium strain TA100 (NTP. 2005). Mutagenicity test results were
equivocal in S. typhimurium strain TA98. Pant (2001) tested PFBS at concentrations up to
5,000 [j,g/plate in E. coli strain WP2uvrA and S. typhimurium strains TA98, TA100, TA1535,
and TA1537 in the presence or absence of exogenous metabolic activation and found no
evidence of mutagenic activity. In mammalian cells in vitro, PFBS did not generate reactive
oxygen species (ROS) or oxidative deoxyribonucleic acid damage in HepG2 cells (Eriksen et al..
2010). PFBS also failed to induce chromosomal aberrations in Chinese hamster ovary cells,
suggesting a lack of clastogenic activity (Xu. 2001). Adult male and female S-D rats exposed
twice daily to oral PFBS at doses up to 250 mg/kg for 28 days did not experience any significant
increases in micronucleated polychromatic erythrocytes, indicating a lack of genotoxic activity
(see Table 5) (NTP. 2012).
4.8.2	Acute Duration and Other Routes of Exposure
Limited data are available to evaluate acute toxicity and effects from dermal exposure to PFBS
(Table 5). One low-confidence acute oral toxicity study reported a median lethal dose (LD50) in
male rats of 236 [xL/kg (corresponding to 430 mg/kg) administered PFBS by oral gavage
(Bomhard and Loser. 1996). One acute dermal toxicity study concluded PFBS is not acutely
toxic via the dermal route of exposure in rats, with no treatment-related observation at doses up
to 2,000 mg/kg (3M. 2000b). PFBS was not reported to induce erythema, edema, or other
possible dermal findings during the scoring periods, indicating a lack of dermal irritant properties
in rabbits exposed to 500 mg K+PFBS for approximately 4 h (3M. 2000a). PFBS was found to be
a moderate ocular irritant in rabbits exposed to 80 mg K+PFBS via ocular installation (3M.
2000c). PFBS did not induce skin sensitization in the guinea pig maximization test with an
intradermal injection of 12.5 mg and topical induction of 50 mg K+PFBS (3M, 2002a).
46

-------
5.0 Evidence Integration and Hazard Characterization
The epidemiology database of studies of PFBS exposure and health effects consists of 19
epidemiologic studies (described in 22 publications), summarized in the previous section. The
experimental animal database of all repeated-dose oral toxicity studies for PFBS and the related
compound K+PFBS includes a short-term range finding study in rats (3Mi_2000d), two 28-day
studies in rats (NTP, 2019; 3M, 2001). one subchronic-duration study in rats (Lieder et al..
2009a), one subchronic-duration lipoprotein metabolism study in mice (Bijland et al., 2011; 3M,
2010), three gestational exposure studies in mice and rats (Feng et al„ 2017; York, 2003a, 2002),
and one two-generation reproductive toxicity study in rats (Lieder et al„ 2009b). Health
outcomes evaluated across available studies included effects on the thyroid, reproductive organs
and tissues, developing offspring, kidneys, liver, and lipids/lipoproteins following oral exposure
to PFBS. Table 6 provides an overview of this database of potentially relevant studies and
effects. This table includes only the high- and medium-confidence animal studies (a single, low-
confidence animal study was not considered informative to drawing judgments on potential
health hazard[s]); the available epidemiology studies are not included as their ability to inform
conclusions about associations was limited due to the small number of studies (typically one) per
outcome and poor sensitivity resulting from low exposure levels.
Following the summary of the available database in Table 6, narrative summaries describe the
evidence integration judgments and the primary rationales supporting these decisions for each
health effect. These narratives are supported by an evidence profile table that succinctly lays out
the various factors that were judged to increase or decrease the support for hazard. While the
epidemiology studies were not influential to drawing evidence integration judgments (i.e., they
were judged as equivocal for all outcomes) or the derivation of toxicity values (i.e., these studies
are not discussed in the next section), the general findings are summarized below to provide
context to the animal study findings and identify potential areas of future research.
47

-------
Table 6. Summary of noncancer data for oral exposure to PFBS (CASRN 375-73-5) and the related compound
K+PFBS (CASRN 29420-49-3)
Exposure
duration3
Reference
Study
confidence
Number of
male/female, strain,
species, study type,
study duration
Doses tested
(mg/kg-d)
Effects observed at LOAEL
NOAEL
(mg/kg-d)
LOAEL
(mg/kg-d)
Short-term
3M (2000d)
Medium
confidence
5/5, S-D rat, K+PFBS
administered by
gavage, 10 d
0, 100, 300,
1,000
Increased absolute and relative liver weight.
300
1,000
Short-term
3M (2001)
High
confidence
10/10, S-D rat,
K+PFBS administered
by gavage, 28 d
0, 100, 300,
900
Increased liver weight (male) and kidney
weight (female).
300
900
Short-term
NTP (2019s)
High
confidence
10/10, S-D rat, PFBS
administered by
gavage, twice/d, 28 d
0, 62.6, 125,
250, 500,
l,000b
Decreased T3, free T4, total T4 in males and
females. Increased relative liver weight in
females, and increased relative right kidney
weight in males.
ND
62.6
Subchronic
Lieder et al.
(2009a): York
(2003b)
High
confidence
10/10, S-D rat,
K+PFBS administered
by gavage, 7 d/wk,
90 d
0, 60, 200,
600
Increased incidence of renal hyperplasia in
males and females.
200
600
Subchronic
Biiland et al.
(2011): 3M
(2010)
Medium
confidence
6-8/0, Apoe*3-Leiden
CETP mice, K+PFBS
in diet, 4-6 wk
0, 30
Alterations in lipid homeostasis (e.g., decreased
hepatic lipase, triglycerides) is of uncertain
biological significance.
ND
ND
Developmental
Fens et al.
(2017)
High
confidence
0/10, ICR mice,
K+PFBS administered
by gavage, GDs 1-20
0, 50, 200,
500
Decreased T3, free T4, and total T4 in dams
and PND 1, 30, and 60 offspring. Increased
TSH in maternal and offspring (PND 30 only).
Delayed eyes opening, vaginal opening, and
final estrous and decreased BW in pups.
50
200
Developmental
York (2003a)
High
confidence
0/8, S-D rat, K+PFBS
administered by
gavage, GDs 6-20
0, 100, 300,
1,000, 2,000
Decreased maternal feed consumption, BW
gain, and gravid uterine weight. Decreased pup
BW at doses where maternal health was
affected limiting the interpretation of the
results; thus developmental effect levels were
not determined. (Limited endpoints
evaluated—pilot study).
P0: 1,000
Fl: ND
P0: 2,000
Fl: ND
48

-------
Exposure
duration3
Reference
Study
confidence
Number of
male/female, strain,
species, study type,
study duration
Doses tested
(mg/kg-d)
Effects observed at LOAEL
NOAEL
(mg/kg-d)
LOAEL
(mg/kg-d)
Developmental
York ("20021
High
confidence
0/25, S-D rat, K+PFBS
administered by
gavage, GDs 6-20
0, 100, 300,
1,000
Decreased maternal feed consumption and BW
gain. Decreased pup BW at doses where
maternal health was affected limiting the
interpretation of the results; thus developmental
effect levels were not determined.
P0:300
Fl: ND
P0: 1,000
Fl: ND
Reproductive
Lieder et al.
(2009b1; York
("2003c1; York
(2003dl; York
("2003e1
High
confidence
30/30, S-D rat,
K+PFBS administered
by gavage,
two-generation
reproductive study
P0 adults: 0,
30, 100, 300,
1,000
F1 adults: 0,
30, 100, 300,
1,000
P0 and F1 adults: increased incidence of
hyperplasia and focal papillary edema in the
kidneys of males and females.
F2 pups: no dose-related effects at the highest
dose tested (1,000 mg/kg-d).
P0, Fl: 100
F2: 1,000
P0, Fl:
300
F2: ND
Notes'. ND = no data; ICR = Institute of Cancer Research
a Duration categories are defined as follows: Acute = exposure for < 24 hours; short term = repeated exposure for 24 hours to < 30 days; long term (subchronic) = repeated
exposure for > 30 days < 10% lifespan for humans (> 30 days up to approximately 90 days in typically used laboratory animal species); chronic = repeated exposure for >
10% lifespan for humans (>~ 90 days to 2 years in typically used laboratory animal species) ("U.S. EPA. 2002).
bRats were gavaged twice daily at administered doses of 0, 31.3, 62.6,125,250, and 500 mg/kg in NTP ("20191.
49

-------
5.1 Thyroid Effects
PFBS-induced perturbation of the thyroid was consistently observed across two species, sexes,
life stages, and exposure durations in two independent, high-confidence studies. These
perturbations involved a coherent pattern of hormonal changes. Significant changes in tissue
weight or histopathology were not observed.
Similar patterns of decreases in total T3. total T4. and free T4 were observed in PFBS-exposed
pregnant mice, nonpregnant adult female and adult male rats from a 28-day study, and
gestationally exposed female mouse offspring (NTP, 2019; Feng et al., 2017). These decreases
were statistically significant (-20% in dams and -50% in offspring), and shown to persist at least
60 days after gestational exposure in offspring, and exhibited dose-dependence in both studies.
Development of numerous organ systems, including neuronal, reproductive, hepatic, and
immune systems, are affected by altered thyroid homeostasis since adequate levels of thyroid
hormones are necessary for normal growth and development in early life stages (Forhead and
Fowden. 2014; Gilbert and Zoeller, 2010; Hulbert. 2000). Thus, the observed effects of PFBS
exposure on thyroid hormone economy are biologically consistent with the reported delays and
abnormalities in organ/system development discussed below. It is well-established that the
presence of sufficient thyroid hormones during the gestational and neonatal period is essential
for brain development and maturation. Studies specifically evaluating the effect of PFBS on
neurodevelopment were not identified, leaving uncertainty as to the potential for adverse
developmental effects. Nonetheless, the coherence of these PFBS findings, in addition to the
large number of xenobiotic exposure studies demonstrating associations between thyroid
hormone economy and decrements in early life stage growth, development, and survival,
provides support for thyroid hazard.
Taken together, the evidence in animals for thyroid effects supports a hazard. The single
available study in humans did not report an association with thyroid hormones, but had severe
limitations hindering its interpretation. This low confidence cross-sectional study was conducted
in a highly selected population (i.e., women with premature ovarian insufficiency), had poor
sensitivity, and methodological limitations (Zhang et al.. 2018). The limited evidence for thyroid
effects in human studies is equivocal. Although there are some differences in hypothalamic-
pituitary-thyroid (HPT) regulation across species (e.g., serum hormone-binding proteins,
hormone turnover rates, and timing of in utero thyroid development), rodents are generally
considered to be a good model for evaluating the potential for thyroid effects of chemicals in
humans (Zoeller et al.. 2007). For more details pertaining to HPT dynamics and the similarities
and differences associated with thyroid hormone economy between rodents and humans, please
refer to 'A Literature Review of the Current State of the Science Regarding Species Differences
in the Control of, and Response to, Thyroid Hormone Perturbations. Part 1: A Human Health
Perspective' (Regulatory Science Associates, 2018). The pattern of decreased thyroid hormones
in the absence of a coordinated reflex increase in TSH and commensurate alterations in thyroid
tissue weight and/or histology, observed in PFBS studies (e.g., Feng et al. (2017)). is consistent
with the human clinical condition referred to as "hypothyroxinemia", which is commonly
associated with pregnancy in humans. Hypothyroxinemia has been defined as a low percentile
value of FT4 (ranging from the 2.5th percentile to the 10th percentile of FT4), with a TSH level
within the normal reference range (Hales et al., 2018; Alexander et al„ 2017; Lazarus et al..
50

-------
2012; Negro et al., 2011). Overall, based on findings in animal models considered to be
informative for evaluating the potential for thyroid effects in humans, the available evidence
supports a hazard and the thyroid is considered a potential target organ for PFBS toxicity in
humans.
5.2 Developmental Effects
Overt effects on birth parameters and early development have generally not been observed in
either rats or mice after PFBS exposure. Specifically, the available studies do not provide
evidence of effects on endpoints relating to pregnancy loss, fetal survival, or fetal morphology
(Feng et al.. 2017; Lieder et al.. 2009a; York. 2003a. c, 2002). While one mouse study indicated
pronounced decreases in female offspring BW at several ages after gestational exposure (Feng et
al.. 2017). several other studies either did not observe decreases in offspring BW or only detected
these changes when parental BWs were similarly affected (Feng et al.. 2017; Lieder et al.. 2009a;
York. 2003a. c, 2002).
Delays in development have been reported following gestational PFBS exposure in mice,
including delayed development of the female reproductive organs (i.e., ovaries, uterus, and
vaginal patency), delayed and abnormal estrous cycling (i.e., first estrous and prolongation of
diestrus), and delayed eye opening (Feng et al.. 2017). Age at vaginal patency and ovarian
follicle counts (i.e., in F1 rat offspring after delivery of the F2 generation) were unaffected at
1,000 mg/kg-day in a two-generation reproductive toxicity study (Lieder et al.. 2009a). This
observed lack of effects (i.e., on vaginal patency) is inconsistent with the findings in mice.
However, Feng et al. (2017) also noted changes in reproductive hormones that might be relevant
to the delays in female sexual development, including a decrease in serum estradiol and
increased luteinizing hormone in pubertal offspring (i.e., PND 30 [Note: progesterone was
decreased at a later age, PND 60, but not PND 30], As the changes reported in mice by Feng et
al. (2017) were observed in parallel with effects on thyroid hormone levels (discussed above), it
is plausible that these developmental delays and hormonal changes could represent sequalae of
reduced thyroid function, although that was not directly tested.
For the most part, developmental effects have been reported in a single study and species
(mouse); however, the findings are coherent with one another as well as with the consequences
of decreased thyroid hormone levels. Due to the coherence across effects on the thyroid and
several interrelated developmental effects in mice (i.e., delays and hormonal changes), the
evidence in animals for developmental effects supports a hazard. There is no reason to expect
that the specific developmental delays observed in mice would not be directly relevant to similar
processes in humans. Thus, based on findings in animals that are presumed to be relevant to
humans, the available evidence supports a hazard and the developing offspring is considered a
potential target for PFBS toxicity in humans. As no studies in humans were available that
investigated these endpoints, this represents an area deserving of additional research.
5.3 Reproductive Effects
Reproductive outcomes, including male and female fertility, pregnancy outcomes, hormone
levels, markers of reproductive development, and reproductive organ weights and
histopathology, have been evaluated in a number of high-confidence studies in mice (Feng et al.,
2017) and rats (NTP. 2019; Lieder et al.. 2009a; Lieder et al.. 2009b). In addition, five low-
51

-------
confidence human studies evaluated potential associations between PFBS exposure and
reproductive effects (Yao et al.. 2019; Song et al.. 2018; Zhang et al.. 2018; Zhou et al.. 2017a;
Zhou et al.. 2016).
PFBS exposure has resulted in no significant changes in male mating and fertility parameters,
reproductive organ weights, or reproductive hormones. While there were some slight,
statistically significant effects on male reproductive endpoints in two rat studies (specifically,
altered sperm parameters such as percentage of abnormal sperm or testicular sperm count (NTP.
2019; Lieder et al.. 2009a) and delayed preputial separation at 1,000 mg/kg-day (Lieder et al..
2009a)), these findings were observed only at the highest doses and the levels of change were of
questionable biological significance. No significant reproductive effects in men were noted
across two human studies (Song et al.. 2018; Zhou et al.. 2016). though EPA noted a non-
significant inverse association with testosterone and estradiol in male infants in one study (Yao
et al.. 2019).
In general, PFBS exposure in adults has also resulted in no significant alterations in female fertility
or pregnancy outcomes in rats or mice (NTP. 2019; Feng et al.. 2017; Lieder et al.. 2009a; Lieder
et al.. 2009b) or in two human studies (Yao et al.. 2019; Zhang et al.. 2018; Zhou et al.. 2017a;
Zhou et al„ 2016), and inconsistent changes in rodent reproductive organ weights were reported
across studies regardless of duration and timing of exposure. However, changes in normal
estrous cyclicity, specifically prolongation of the diestrus stage, have been reported in both
nonpregnant adult rats exposed to PFBS (NTP, 2019) and adult mouse offspring exposed
gestationally from GDs 1 to 20 (Feng et al., 2017). PFBS exposures in NTP (2019) began
between 8 and 10 weeks of age; although the exposures might overlap with some aspects of
reproductive development or changes in function during adolescence, these rats were sexually
mature and thus the endpoints are considered in the context of reproductive, rather than
developmental, effects. The mouse offspring in the study by Feng et al. (2017) also displayed
delayed vaginal patency and histopathological markers of decreased fertility (i.e., decreased
follicles and corpora lutea); however, the reproductive function of those offspring was not tested.
While adult rat offspring (Fl) in a two-generation toxicity study also exhibited variable changes
in estrous cyclicity (Lieder et al., 2009b), including prolonged diestrus at 100 mg/kg-day, this
effect was not observed at higher doses, limiting interpretation, and no effects on vaginal patency
were observed. Female reproductive hormones can inform the potential for effects on
reproductive organ development, estrous cyclicity, and fertility. Changes in serum hormones
included increased testosterone after exposure of female rats as adults (NTP, 2019), increased
luteinizing hormone and decreased estradiol in pubertal mice after gestational exposure (Feng et
al., 2017), and decreased estradiol and progesterone when these gestationally exposed mice were
assessed as adults. Overall, the pattern and timing of hormonal changes after PFBS exposure is
difficult to interpret and likely incomplete. However, the hormonal alterations after gestational
PFBS exposure in mice are most relevant to conclusions about female reproductive health.
Taken together, the evidence indicates that the developing reproductive system, particularly in
females, might be a target for PFBS toxicity. However, the potential for reproductive effects in
adults was less clear, and significant impacts on mating or fertility parameters were not observed
across the available studies. Therefore, the evidence in developing animals is considered most
informative to conclusions relating to potential developmental effects (see above) and the
evidence for reproductive effects (i.e., in adults) is equivocal. In the three studies of potential
52

-------
reproductive effects in humans, no clear associations were observed, and so the evidence in
human studies is equivocal. Overall, based on equivocal human and animal evidence, the
available evidence for reproductive effects is equivocal.
5.4 Renal Effects
Renal effects associated with oral exposure to PFBS have been observed in adult or developing
rats across high- or medium-confidence gavage studies of various duration (NTP, 2019; Lieder et
al.. 2009a; Lieder et al., 2009b; 3M. 2001. 2000d).
Statistically significant increases in kidney weights have been observed in male and female rats
after short-term exposure in one study (NTP, 2019), with strong dose-dependence for changes in
relative weights in female rats at doses as low as 62.5 mg/kg-day. This study was likewise the
only study to observe changes in serum markers of renal injury, specifically increased BUN in
males at > 250 mg/kg-day. However, while several other studies noted slight increases in weights,
typically at higher PFBS doses (> 500 mg/kg-day), EPA found that these non-significant changes
were not consistently observed across the set of available studies and no other studies reported
changes in serum markers of renal injury (Lieder et al„ 2009a; Lieder et al„ 2009b; 3M, 2001,
2000d).
Several kidney histopathology lesions (i.e., CPN, hydronephrosis, tubular degeneration, and
tubular dilation) were unaffected by PFBS exposure in rats, although each of these endpoints was
not assessed across several studies (NTP, 2019; Lieder et al„ 2009a; 3M, 2000d). Mixed results
were reported for mineralization and necrosis. Both of these endpoints were noted in females, but
not males, after subchronic exposure to 600 mg/kg-day (Lieder et al„ 2009a), whereas
mineralization was unaffected in male or female rats after short-term exposure (3M, 2000d) and
necrosis was unaffected in male or female rats in short-term and 2-generation (in both
generations) studies (NTP, 2019; Lieder et al., 2009b). Multiple markers of inflammatory
changes were consistently noted in the two longest exposure duration studies, which were the
only studies to report on these endpoints. Specifically, increases in chronic pyelonephritis,
tubular basophilia, and mononuclear cell infiltration were observed in female, but not male, rats
following subchronic exposure to 600 mg/kg-day (Lieder et al., 2009a). Similarly, increases in
papillary edema and hyperplasia were observed in male and female rats after subchronic
exposure to 600 mg/kg-day (Lieder et al., 2009a), and in both generations of rats in the
two-generation study at > 300 mg/kg-day (Lieder et al., 2009b), with female rats being more
sensitive than males.
Overall, the evidence in animals suggests an increased sensitivity of female rats (i.e., based on
histopathology and organ weight changes). Due primarily to the consistency and coherence in
renal effects observed in the subchronic-duration study by Lieder et al. (2009a) and the
reproductive toxicity study by Lieder et al. (2009b) in male and female rats, the evidence in
animals supports a hazard. There is insufficient evidence in epidemiology studies of PFBS to
inform the human relevance of these findings. Taken together, the renal histopathology evidence
in rodents identifies a toxicologically significant spectrum of effects that is presumed to be
relevant to similar changes known to occur in humans. Renal effects (i.e., uric acid) were
evaluated in one low-confidence human study and no clear association was observed, and so the
evidence in human studies is equivocal. Overall, based on findings in animals that are presumed
53

-------
to be relevant to humans, the available evidence supports a hazard and indicates the kidney as a
target organ of PFBS toxicity.
5.5 Hepatic Effects
Hepatic effects, including organ-weight changes and histopathology associated with oral
exposures to PFBS, have been observed in high- or medium-confidence studies in adult or
developing rats following short-term and subchronic durations (NTP, 2019; Lieder et al.. 2009a;
3M, 2001, 2000d) and in a two-generation reproductive study in rats (Lieder et al., 2009b).
Increased absolute and/or relative liver weights were consistently observed in male and female
rats after short-term and multigenerational exposure (NTP, 2019; Lieder et al., 2009b; 3M, 2001,
2000d). In some studies, the magnitude of the liver weight changes and the doses at which
effects occurred differed across sexes of rat, although the pattern across studies was unclear and
did not consistently indicate one sex as more sensitive. Liver histopathology, including necrosis
and inflammation, was not consistently observed across PFBS studies. One possible exception is
increases in hepatocellular hypertrophy in male rats observed across two studies (NTP, 2019;
Lieder et al., 2009b), although female rats were unaffected in the multigenerational study and
this lesion was not observed at up to 600 mg/kg-day in the subchronic study by Lieder et al.
(2009a). The only study to observe changes in serum markers of liver injury was NTP (2019), at
> 250 mg/kg-day in females and > 500 mg/kg-day in males. The biological relevance or
significance of the observed liver effects is not clear. In particular, the adversity of the variable
changes in liver weight and observations of cellular hypertrophy is unclear. Further, the observed
lesions either occurred in only one sex of rat, were not dose-dependent compared to control,
and/or occurred only at the highest PFBS dose tested. Thus, the evidence in animals is equivocal.
Overall, based on equivocal animal evidence and a lack of human studies, the available evidence
for hepatic effects is equivocal.
5.6 Effects on Lipid or Lipoprotein Homeostasis
Few studies have examined the effects of PFBS on circulating or hepatic lipid or lipoprotein
homeostasis. It is recognized that increased circulating levels of lipids and lipoprotein products
and/or increased hepatic lipid load are clinical observations of concern in humans. However, the
lack of effect on lipid dynamics in most studies of rats exposed to high oral K+PFBS doses for up
to 90 days and the generally modest effects in transgenic mice, designed to interrogate
mechanisms of lipid transport and metabolism, fed a high-fat, western-type diet renders this
potential health outcome of unclear toxicological significance at this time. Thus, given the
inconsistent, modest effects and the unclear biological relevance of these changes in isolation
(i.e., lipids/lipoproteins were decreased, not increased) the evidence in animals is equivocal.
Effects on serum lipids were evaluated in one low-confidence human study and childhood
adiposity was evaluated in one medium-confidence study. Although an association was observed
between increased PFBS exposure and increased total cholesterol and higher adiposity, this
evidence in humans is equivocal due to lack of additional supportive evidence. Overall, based on
equivocal evidence in both animal and human studies, the available evidence for effects on lipid
or lipoprotein homeostasis is equivocal.
5.7 Immune Effects
Immune effects were observed in two human studies, including associations with asthma (Dong
et al„ 2013a) and atopic dermatitis (Chen et al„ 2018). Exposure of human peripheral blood
54

-------
leukocytes or human promyelocytic THP-1 cells to PFBS, in culture, decreased cytokine (e.g.,
TNFa and IL-10) secretion following antigen challenge (Corsini et al.. 2012). Because of the
lack of additional evidence and some concerns about potential for residual confounding by other
PFAS, the evidence in human studies is equivocal. Overall, based on equivocal evidence in
human studies and a lack of animal studies, the available evidence for immune effects is
equivocal.
5.8 Cardiovascular Effects
Cardiovascular effects were observed in two human studies, including associations with
cardiovascular disease in adults (Huang et al.. 2018) and hypertensive disorders in pregnancy
(Huang et al.. 2019b). The results are compelling, but as with the evidence for immune effects,
there is a lack of additional supportive evidence and some concerns about potential for
confounding, thus the evidence in human studies is equivocal. Overall, based on equivocal
evidence in human studies and a lack of animal studies, the available evidence for cardiovascular
effects is equivocal.
5.9 Evidence Integration and Hazard Characterization Summary
Based on the evidence integration judgments regarding the potential for PFBS exposure to cause
health effects (the narrative above is summarized in Table 7), the animal studies informing the
potential effects of PFBS exposure on thyroid function, renal function, and development were
concluded to support hazard. Thus, for the purposes of this assessment, the animal data
supporting these outcomes were considered for use in dose-response analysis, and other data
were considered no further.
55

-------
Table 7. Summary of hazard characterization and evidence integration judgments
Studies and confidence
Factors that increase support
for hazard
Factors that decrease
support for hazard
Summary of findings
Overall
evidence
integration
judgment and
basis
Thyroid effects
Human studies
Supports a
• Low confidence
case-control study
(Zhane et al..
2018)
• No factors noted.
• Single study of low
confidence and poor
sensitivity.
No association of PFBS with free T3, free T4, or
thyroid stimulating hormone, but the study had
poor sensitivity and other methodological
limitations that hinder interpretability.
hazard
(animal
evidence
supports a
hazard; human
evidence is
Animal studies (all oral pavape)
Mouse Studies:
•	High-confidence
gestational
(GDs 1-20)
exposure study
(Fene et al.. 2017)
Rat Studies:
•	High-confidence
short-term (28-d)
toxicity study
(NTP. 2019)
•	Consistent thyroid
hormone decreases
(i.e., for total T3, total
T4, and free T4) across
two high-confidence
studies of varied
design. The findings
were consistent across
two species, sexes, life
stages, and exposure
durations.
•	Dose-response
gradients were
observed for those
thyroid hormones.
•	Large magnitudes of
effect (e.g., up to -50%
reductions in offspring
serum hormones) were
reported for those
thyroid hormones.
• No factors noted.
Similar patterns of decreases in thvroid hormones
(i.e.. for total T3. total T4. and free T4) were
observed in PFBS-exposed pregnant mice and
gestationally exposed female mouse offspring at
> 200 me/ke-d (Fens et al.. 2017) and in adult
female and male rats at > 62.6 me/ke-d (NTP.
2019).
Increased TSH was reported in mouse dams and
in pubertal (PND 30) offspring following
eestational exposure (Fene et al.. 2017). but no
changes were noted in rats exposed as adults
(NTP. 2019).
Thvroid weieht and historatholoev were not
equivocal).
The primary
basis for this
judgment is
thyroid
hormone
decreases in
mice and rats
at>
62.6 mg/kg-d.


changed after short-term exposure in adult male or
female rats (NTP. 2019).

Developmental effects
56

-------
Studies and confidence
Factors that increase support
for hazard
Factors that decrease
support for hazard
Summary of findings
Overall
evidence
integration
judgment and
basis
Human studies
Supports a
hazard
(animal
evidence
supports a
hazard; human
evidence is
equivocal).
The primary
basis for this
judgment is a
set of persistent
developmental
delays and
alterations in
reproductive
system
maturation in
female mice,
generally at >
200 mg/kg-d.
No studies available to
evaluate
--
--
--
Animal studies (all oral gavage)
Mouse Studies:
• High-confidence
gestational
(GDs 1-20)
exposure study
(Fens et al.. 2017)
•	Biologically consistent
spectrum of
developmental effects
in female offspring in a
high-confidence mouse
study at doses not
causing maternal
toxicity, including
pronounced and
persistent effects on
BW, delays in
developmental
milestones and sexual
maturation, concordant
effects on reproductive
organs, and altered
serum hormones.
•	Concerning magnitude
of effect (e.g., -25%
change in pup weight)
and dose-dependence
for several parameters.
•	Coherence of effects
with thyroid hormone
insufficiency (see
above).
Note: these effects were also
coherent with effects on estrous
•	Developmental
effects were limited
to changes in one
study, sex, and
species.
•	A high-confidence
rat study reported
some inconsistent
evidence, including
lack of a delay in
vaginal patency and
lack of clear effects
on estrous cyclicity
or ovarian
morphology,
although the latter
endpoint was
assessed in much
older animals. These
potential differences
across species are
not explainable
based on
toxicokinetics alone.
In the onlv mouse studv (Fens et al.. 2017).
developmental effects and altered markers of
female reproductive development or function
were observed in female offspring after
gestational PFBS exposure, including decreased
BW. delaved eve openine. delaved vasinal
Rat Studies:
•	Two high-
confidence
gestational
exposure
(GDs 6-20)
studies: a range
finding study and
a follow-up study
(York. 2003c.
2002)
•	High-confidence
2-generation study
(Licdcr et al..
2009b)
ODcnina. altered estrous cvclicitv (includins
prolonsed diestrus). altered reproductive
hormones (e.s.. decreased estradiol and
progesterone), and effects on reproductive organs
(e.e.. weisht and ovarian morpholosv). Most
effects were observed at > 200 mg/kg-d, with
several changes noted at PND 60.
Endooints relatins to fertility, presnancv.
survival, and fetal alterations were unchanged in
both rats and mice across the four available
studies, although this was not tested in mouse
offspring (Fens et al.. 2017).
Developmental BW chanses in rat offsprins were
either unchanged (Lieder et al.. 2009b) or
observed only at doses causing parental toxicity
(York. 2003c. 2002).
In a rat two-generation study, while some
statistically significant findings were noted for
markers of female reproductive development or
function, thev were not dosc-dcDcndcnt or were of
questionable biological relevance; thus, no clear
changes in F1 offspring were noted at doses up to
1.000 ms/ks-d resardins vasinal patencv or
57

-------
Studies and confidence
Factors that increase support
for hazard
Factors that decrease
support for hazard
Summary of findings
Overall
evidence
integration
judgment and
basis

cyclicity observed after
short-term exposure in adult rats
(NIP. 2019). but this was
categorized as a reproductive
effect (see below).

estrous cvclins at comparable ases to (Fens et al..
2017). or in ovarian morpholosv after the F1
females gave birth to the F2 pups.

Reproductive effects
Human studies
Equivocal
Male reproductive effects
(equivocal
human and
animal
evidence).
Note: As the
strongest
evidence for
female
reproductive
effects was in
offspring that
were
•	Low-confidence
cohort study
(Zhou et al.. 2016)
•	Low-confidence
cross-sectional
studv (Sons et al..
2018)
•	Low confidence
cross-sectional
studv (Yao et al..
2019)
• No factors noted.
• Lack of clear
association in studies
of low confidence
with poor sensitivity
(i.e., due to low
exposure levels,
range).
No clear association between PFBS exposure and
male reproductive hormones (Zhou et al.. 2016) or
semen parameters (Sons et al.. 2018) in adults. A
study in newborns reported non-significant
inverse associations between PFBS exposure and
testosterone and estradiol (Yao et al., 2019).
Female reproductive effects
gestationally
exposed, these
findings were
considered
most relevant
to
developmental,
not
reproductive,
effects.
•	Low-confidence
cross-sectional
studv (Zhou et al..
2017a)
•	Low-confidence
cohort study
(Zhou et al.. 2016)
•	Low confidence
cross-sectional
studv (Yao et al..
2019)
• No factors noted.
•	Lack of clear
association in studies
of low confidence
with poor sensitivity
(i.e., due to low
exposure levels,
range).
•	Potential for reverse
causation for
menstrual cycle
characteristics and
No clear association between PFBS exposure and
female reproductive hormones (Zhou et al.. 2016)
or menstrual cvcle characteristics (Sons et al..
2018).
58

-------
Studies and confidence
Factors that increase support
for hazard
Factors that decrease
support for hazard
Summary of findings
Overall
evidence
integration
judgment and
basis
• Low confidence
case-control study
(Zhane et al..
2018)

premature ovarian
insufficiency.


Animal studies (all oral pavape)

Male reproductive effects

Rat Studies:
•	High-confidence
short-term (28-d)
toxicity study
(NTP. 2019)
•	High-confidence
2-generation study
(Licdcr et al..
2009b)
•	High-confidence
subchronic study
(Licdcr et al..
2009a)
• No factors noted.
•	A few small,
statistically
significant changes
were not
dose-dependent or
were of questionable
biological relevance.
•	Lack of effects on
male mating and
fertility, hormones,
or reproductive
organs in rats.
Statistically significant effects on sperm health
(NTP. 2019; Lieder et al.. 2009a) and delayed
DrcDutial separation at 1.000 me/ke-d (Lieder et
al.. 2009b) were not observed at lower doses,
were within the normal range of historical
controls for the laboratory, and/or were no longer
significantly changed after correcting for other
variables (e.g., BW).
Other relevant parameters (e.e.. orean weiehts.
mating success, and so forth) were unchanged in
the three studies.

Female reproductive effects

Mouse Studies:
•	High-confidence
gestational
(GDs 1-20)
exposure study
(Fene et al.. 2017)
Rat Studies:
•	High-confidence
short-term (28-d)
toxicity study
(NTP. 2019)
•	Effects on markers of
female reproductive
function (i.e., estrous
cyclicity) were
observed in high-
confidence studies in
rats and mice.
•	Changes in
reproductive serum
hormones were
observed in female rats
(i.e., increased
•	Lack of similar
effects on
reproductive
function (i.e., estrous
cyclicity) in a second
high-confidence rat
study.
•	Lack of effects on
female fertility or
pregnancy measures,
although this was
untested in
See "Developmental effects" (above) for findings
from (Fens et al.. 2017) and (Lieder et al.. 2009b).
Altered estrous cvclicitv (includine proloneed
diestrus) and increased serum testosterone were
observed in female rats after short-term exposure,
primarily at > 250 me/ke-d (NTP. 2019).
Female reproductive orean weiehts were reduced
in eestationallv exposed mouse offsorine (Fene et
al.. 2017). but were unchaneed after short-term,
subchronic. or 2-eenerational exposure (NTP.
2019; Lieder et al.. 2009a: Lieder et al.. 2009b).

59

-------
Studies and confidence
Factors that increase support
for hazard
Factors that decrease
support for hazard
Summary of findings
Overall
evidence
integration
judgment and
basis
•	High-confidence
subchronic study
(Licdcr et al..
2009a)
•	High-confidence
2-generation study
(Licdcr et al..
2009b)
testosterone) and mice
(e.g., decreased
estradiol and
progesterone).
Although the pattern of
change is difficult to
interpret and likely
incomplete, there were
no conflicting data.
prenatally exposed
female mouse
offspring.
• Lack of organ
weight changes in
three rat studies.
Note: The lack of effects on
ovarian follicles in rats did
not decrease the support for
hazard provided by findings
in mice, as the age at endpoint
assessment was not
comparable.


Renal effects
Human studies
Supports a
hazard.
(animal
evidence
supports a
hazard; human
evidence is
equivocal).
The primary
basis for this
judgment is
kidney
histopathology
in rats,
primarily
females, at
>300 mg/kg-d.
• Low-confidence
cross-sectional
studv (Oin et al..
2016)
• No factors noted.
•	Inconsistency across
subpopulations in
single study.
•	Single study of low
confidence with
concern for potential
reverse causality.
Overall, there was no clear association for PFBS
and uric acid. No association observed between
PFBS and uric acid in the total population.
Increase in uric acid with increased exposure in
boys, but decrease for girls (neither was
statistically significant).
Animal studies (all oral savage)
Rat Studies:
• One high-
confidence
subchronic study
(Licdcr et al..
2009a)
• Two high-confidence
studies with the longest
exposure durations
reported consistent
effects on kidney
histopathology in male
and female rats
•	Inconsistency in
kidney weight
changes across
studies.
•	Findings are from a
single laboratory and
species.
Increases in kidnev weisht in male and female
rats were observed in one short-term study at >
62.5 mg/kg-d, but clear changes were not
observed in the other short-term, subchronic, or
two-generation rat studies.
Kidnev histopatholoev for some effects
• Two high-
confidence study
(i.e., CPN, hydronephrosis, tubular degeneration,
and tubular dilation) was unchanged in
single-study evaluations, and mixed results across
60

-------
Studies and confidence
Factors that increase support
for hazard
Factors that decrease
support for hazard
Summary of findings
Overall
evidence
integration
judgment and
basis
(NTP. 2019; 3M.
2001s) and one
medium-
confidence (3M.
2000d) short-term
(10-28 d) study
• One high-
confidence
2-generation study
(Lieder et al..
2009b)
(females were more
sensitive).
• The histopathological
effects related to
inflammation were
largely dose-dependent
and of a concerning
magnitude, although
primarily at high doses
(300 or 600 mg/kg-d).
Note: The general lack of
effects on other pathology
endpoints in the shorter term
studies was not considered to
decrease support for hazard,
as this was not interpreted as
inconsistent.
studies were reported for mineralization and
necrosis (NTP. 2019; Lieder et al.. 2009a: Lieder
et al.. 2009b: 3M. 2000d). Multiple markers
potential lv related to inflammation and most
notably papillary edema and hyperplasia were
increased in the two longest duration studies
(Lieder et al.. 2009a: Lieder et al.. 2009b).
without contrary evidence.
Other markers of renal iniurv. including BUN and
creatinine, were mostly unaffected across studies
(NTP. 2019: Lieder et al.. 2009a: Lieder et al..
2009b: 3M. 2001. 2000d). although the NTP
study did observe effects on BUN in males at >
250 mg/kg-d.

Hepatic effects
Human studies
Equivocal
(equivocal
human and
animal
evidence).
No studies available to
evaluate
--
--
--
Animal studies (all oral pavape)
Rat Studies:
•	One high-
confidence
subchronic study
(Lieder et al..
2009a)
•	Two high-
confidence study
fNTP. 2019; 3M.
2001) and one
medium-
confidence (3M.
• Consistent changes in
liver weights in rats of
both sexes across four
studies. Although the
pattern (e.g., by sex
and dose) and
magnitude of changes
varied across studies,
weights were
consistently increased.
•	Other than
liver-weight
changes, there were
notable unexplained
inconsistencies in the
findings across
studies.
•	One high-confidence
study was entirely
inconsistent.3
Absolute or relative liver weights were increased
in all studies except the 90-d exposure component
of the studv bv Lieder et al. (2009a). which tested
doses up to 600 mg/kg-d.
Note: 70 d of exposure in this study did elicit
effects.
Effects generally occurred at > 300 mg/kg-d,
although one study reported effects at lower doses
(NTP. 2019: 3M. 2001). and two others (3M.
2001. 2000d) observed changes at > 900 mg/kg-d.
Serum markers of liver iniurv were unchanged in
three studies (Lieder et al.. 2009a: 3M. 2001.
61

-------
Studies and confidence
Factors that increase support
for hazard
Factors that decrease
support for hazard
Summary of findings
Overall
evidence
integration
judgment and
basis
2000d) short-term
(10-28 d) study
• One high-
confidence
2-generation study
(Lieder et al..
200%)


2000d) and increased in one short-term studv at >
250 me/ke-d (NTP. 2019).
Liver histopatholoev. specificallv hepatocellular

hypertrophy and cytoplasmic alterations in males
and females (NTP. 2019) or hvDcrtroDhv in
females onlv (Lieder et al.. 2009a). were noted in
two studies, but not in the others.
Lipid or lipoprotein homeostasis
Human studies
Equivocal
(equivocal
human and
animal
evidence).
•	Low-confidence
cross-sectional
studv (Zens et al..
2015)
•	Medium
confidence study
rChenetal.. 2019)
•	Statistically significant
association in medium
confidence study of
adiposity.
•	Exposure response
gradient observed
across tertiles for
adiposity.
•	Single study per
outcome.
•	Potential for residual
confounding.
Increase in total cholesterol (statistically
significant, (3 = 19.3 mg/DL increase per unit
increase inPFBS) (Zens et al.. 2015). Hisher
adiposity in 5-year-old children associated with
hieher levels of PFBS in cord blood (Chen et al..
2019).
Animal studies
Mouse Studies (diet):
• Medium-
confidence
short-term
(4-6 wk) study
(Biiland et al..
2011); transgenic
mice (human-like
lipid metabolism)
were fed a high-fat
diet
Rat Studies (all oral
savage):
• Decreases in serum
cholesterol and
triglycerides were
observed in male rats
and mice.
•	Inconsistent
evidence in other rat
studies and across
sexes.
•	Small effect
magnitudes and
unclear direction
(decreases) of
changes are of
questionable
biological relevance
and could not be
informed by
evaluating
Serum lipids, specifically cholesterol and
triglyceride levels, were slightly decreased
(-20%) at 900 mg/kg-d in males, but not females,
in one rat studv (3M. 2001). but not in two other
rat studies at up to 1,000 mg/kg-d. Serum and
hepatic lipids and lipoproteins were also
decreased in male mice exposed to ~30 mg/kg-d
in diet.

62

-------
Studies and confidence
Factors that increase support
for hazard
Factors that decrease
support for hazard
Summary of findings
Overall
evidence
integration
judgment and
basis
•	One high-
confidence
subchronic study
(Licdcr et al..
2009a)
•	One high-
confidence study
(3M. 2001s) and
one medium-
confidence (3M.
2000d) short-term
(10-28 d) study

dose-dependency
(i.e., only
single-dose or
high-dose effects
were observed).


Immune effects
Human studies
Equivocal
Asthma
(equivocal
human and
animal
evidence).
• Medium-
confidence
case-control study
(Zhou et al.. 2016;
Zhuetal.. 2016;
Dons et al..
2013b)
• Statistically significant
association in a
medium-confidence
study.
Note: Increases in eosinophil
markers were not interpreted to
increase support for hazard, as
they were not statistically
significant and other markers
important to asthma etiology
(e.g., IgE) were unchanged.
• Association was
observed in a single
study with concern
regarding the
potential for residual
confounding
(e.g., with other
PFAS chemicals).
Statistically significant increase in odds of asthma
diagnosis in the previous year (OR = 1.2-1.9)
with increased PFBS exposure.
Eosinophil markers (i.e., AEC and ECP) were
increased with increased PFBS exposure in
asthmatics and nonasthmatics; however, this did
not reach statistical significance. IgE and T-helper
cell-specific cytokines were unchanged fZlui et
al.. 2016).
Atopic dermatitis

• Medium-
confidence cohort
studv (Chen et al..
2018)
• No factors noted.
• Slight associations
were not statistically
significant in a
single study with
Nonstatistically significant increase in odds of
atopic dermatitis (OR =1.2) with increased PFBS
exposure.

63

-------
Studies and confidence
Factors that increase support
for hazard
Factors that decrease
support for hazard
Summary of findings
Overall
evidence
integration
judgment and
basis


concern regarding
the potential for
residual confounding
(e.g., with other
PFAS chemicals).


Animal studies

No studies available to
evaluate
--
--
--

Cardiovascular effects
Human studies
Equivocal
•	Medium-
confidence cross-
section study
(H nana et al..
2018)
•	Medium-
confidence cross-
sectional study
(Hnana et al..
2019b)
• Statistically significant
associations in
medium-confidence
studies.
• Single study per
outcome.
Higher odds of cardiovascular disease (total and
individual types of disease) with PFBS exposure
(Hnana et al.. 2018). Hisher odds of hypertensive
disorders in pregnancy with higher PFBS
exposure (Huans et al.. 2019b). There is DOtcntial
for residual confounding that decreases
confidence in the evidence.
(equivocal
human and
animal
evidence).
Animal studies

No studies available to
evaluate
--
--
--

Notes:
a The lack of liver effects in the subchronic study was not interpreted to significantly reduce support for hazard, as the maximum tolerated dose was 600 mg/kg-d, and other studies
reported only liver effects at > 900 mg/kg-d.
T3 = triiodothyronine; T4 = thyroxine.
64

-------
6.0 Derivation of Values
The hazard and dose-response database for PFBS and the potassium salt is primarily associated
with the oral route of exposure. There are a limited number of dermal studies (see Table 5) and
no known inhalation studies. There are no known studies evaluating potential cancer effects of
PFBS. As such, only noncancer reference values are derived in this assessment for the oral route.
6.1 Derivation of Oral Reference Doses
The hazards of potential concern for oral PFBS exposure include thyroid, developmental, and
kidney effects. Overall, the evidence supports a hazard for thyroid, developmental, and kidney
effects based on the evidence from animal studies. The limited evidence for thyroid or renal
effects in human studies is equivocal, and no studies evaluating developmental effects following
PFBS exposure in humans were available. Thus, data in humans were not considered further and
the available animal studies that evaluated these effects are considered in the derivation of oral
RfDs.
6.1.1 Derivation of Subchronic RfD
6.1.1.1 Estimation of Points of Departure (PODs)
Effects in the thyroid were considered when determining potential PODs for derivation of a
subchronic RfD. Similar patterns of decreases in total T3. total T4. and free T4 were observed in
PFBS-exposed pregnant mice, nonpregnant adult female rats, adult male rats, and gestationally
exposed female mouse offspring (NTP. 2019; Feng et al.. 2017). These decreases were
significant (-20% in dams and -50% in offspring), were shown to persist at least 60 days after
gestational exposure in offspring, and they exhibited a clear dose dependence in both studies.
Reflex increases in TSH in response to decreased T4 or T3 were not observed in male or female
rats following 28 days of exposure (NTP. 2019). Such an increase in TSH was observed in
pregnant mice (measured at GD20) and their corresponding female offspring, at PND 30 only,
with an irregular dose-response or time course (Feng et al.. 2017). This pattern of decreased
thyroid hormone without a concomitant increase in TSH is consistent with a human clinical
condition referred to as "hypothyroxinemia" (Negro et al.. 2011). Importantly, it has been noted
that milder forms of thyroid perturbation are up to 10 times more prevalent in human populations
than overt gestational hypothyroidism (Korevaar et al.. 2016; Stagnaro-Green et al.. 2011).
Hypothyroxinemia has been associated with impairments in neurodevelopment and/or cognition
later in life (Thompson et al.. 2018; Min et al.. 2016). As the single available study in humans
had severe limitations hindering the interpretation of the relationship between PFBS exposure
and thyroid hormone alterations, at this time the available evidence in humans is not able to
inform the potential for thyroid effects in humans. This hypothyroxinemia, rather than overt or
subclinical hypothyroidism, is further supported by the lack of effect on thyroid weight or tissue
architecture in rats after 28 days of PFBS exposure (NTP. 2019).
Developmental effects were considered in the determination of potential PODs for derivation of
a subchronic RfD. Specifically, in Feng et al. (2017). developmental delays or abnormalities in
growth (i.e., BW and eye opening), reproductive organs (i.e., ovaries, uterus, and vaginal
opening), and reproductive cycling (i.e., first estrous and prolongation of diestrus) were observed
65

-------
in mouse offspring. These effects were observed in mice from litters in which thyroid hormone
deficiency occurred at PND 1 and was sustained through pubertal and adult periods (i.e., PND 30
and PND 60, respectively). These interrelated developmental effects in mice (i.e., delays and
hormonal changes) are coherent with effects on the thyroid and presumed to be directly relevant
to similar processes in humans; however, studies evaluating these outcomes in humans are not
available.
Effects in the kidney were considered when determining potential PODs for derivation of a
subchronic RfD. Mild-to-moderate hyperplasia was reported in the kidneys of male and female
rats following subchronic-duration exposure to PFBS by Lieder et al. (2009a) and in the P0- and
F1-generation animals of the reproductive toxicity study by Lieder et al. (2009b). Other studies
evaluating effects in the kidney were of shorter duration and thus less suitable as a candidate
principal study. Additional histopathological alterations accompanied the hyperplasia observed
in the kidney, including papillary edema and inflammatory changes, specifically increases in
chronic pyelonephritis, tubular basophilia, and mononuclear cell infiltration (Lieder et al.. 2009a;
Lieder et al., 2009b). Across kidney histopathological effects reported following PFBS exposure,
in general, female rats were more sensitive than males.
Selected data sets from studies with multiple exposure levels for thyroid, developmental, and
kidney effects were modeled using the EPA's Benchmark Dose Software (BMDS) Version 2.7.
Consistent with the EPA's Benchmark Dose Technical Guidance Document (U.S. EPA. 2012).
the BMD and 95% lower confidence limit on the BMD (BMDL) were estimated using a
benchmark response (BMR) to represent a minimal, biologically significant level of change.
Based on BMD guidance, in the absence of information regarding the level of change that is
considered biologically significant, a BMR of 1 SD from the control mean for continuous data or
a BMR of 10% extra risk for dichotomous data is used to estimate the BMD and BMDL, and to
facilitate a consistent basis of comparison across endpoints, studies, and assessments. For some
types of effects (e.g., frank effects, developmental effects), biological considerations may
warrant the use of a BMR of 0.5 SD or lower.
For effects in developing offspring, including thyroid hormone changes, a BMR of 0.5 SD
change from the control mean is used for continuous data to account for effects occurring in a
sensitive life stage. A 1 SD BMR is also presented as the basis for model comparison as directed
in the EPA BMD Technical Guidance (U.S. EPA. 2012).
For thyroid hormone effects in offspring, a biological level of concern was considered in the
identification of a BMR. Multiple lines of evidence regarding degree of thyroid hormone
disruption and developmental outcomes in offspring were evaluated. During developmental life
stages such as gestational/fetal and postnatal/early newborn, thyroid hormones are critical in a
myriad of physiological processes associated with somatic growth and maturation and survival
mechanisms such as thermogenesis, pulmonary gas exchange, and cardiac development (Sferruzzi-
Perri et al.. 2013; Hillman et al.. 2012). Further, thyroid hormones are critically important in early
neurodevelopment as they directly influence neurogenesis, synaptogenesis, and myelination (Rovet.
2014; Puig-Domingo and Vila. 2013; Stenzel and Huttner. 2013; Patel et al.. 2011). It should be
noted that evidence from human epidemiological studies examining the association between thyroid
hormone economy in pregnant mothers and neurodevelopment in their offspring is inconsistent.
Several human epidemiologic studies have demonstrated key relationships between decreased
66

-------
levels of thyroid hormones such as FT4 in a pregnant woman and in utero and early postnatal life
neurodevelopmental status. For example, children born euthyroid but who were 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)
and/or concomitant abnormalities in brain imaging (Levie et al.. 2018; Korevaar et al.. 2016;
Henrichs et al.. 2010; Lavado-Autric et al.. 2003; Mirabella et al.. 2000). Maternal
hypothyroxinemia was also associated with adverse motor function and teacher-reported problems
of behavior in offspring at five years of age (Andersen et al.. 2018). Other human epidemiologic
studies have not reported significant associations between thyroid hormone status during pregnancy
and neurodevelopmental outcomes in offspring. For example, there was no statistically significant
association between thyroid status and IQ decrements or neuropsychological parameters in children
born to mothers screened and diagnosed with subclinical hypothyroidism (Hales et al.. 2018;
Lazarus et al., 2012) or mothers undergoing treatment for hypothyroxinemia during gestation
(Casey et al.. 2017). In these studies, the timing of maternal hypothyroxinemia during pregnancy
may be a critical consideration for developmental health outcomes in offspring. Studies have
observed a relationship between low free T4 levels in women at 12 weeks gestation, but not 32
weeks gestation, and impaired psychomotor development in their offspring (Kooistra et al.. 2006;
Pop et al.. 2003). In addition, differences in the type of maternal disruption of thyroid homeostasis
may affect the interpretation of the human epidemiologic study results. Specifically, aside from
overt primary hypothyroidism, there are two primary subcategories of hypothyroidism: (1)
subclinical hypothyroidism; and (2) hypothyroxinemia. Subclinical hypothyroidism is characterized
by elevated TSH levels with normal serum T4 and T3 concentrations. In contrast, hypothyroxinemia
is characterized by decreased T4 with normal serum concentrations of TSH and T3 (Alexander et
al.. 2017; Choksi et al.. 2003). As maternal T4 is the primary source of thyroid hormone for a
developing human fetus in the first trimester (i.e., little if any maternal T3 is transferred across the
placenta primarily due to high levels of deiodinase 3 activity that catabolizes T3 to a biologically
inactive form), and the first trimester is a critical window for central nervous system development
(e.g., neural tube, spinal cord, medulla, pons, thalamus/hypothalamus, etc.), it stands to reason that
the health implications for early in utero development associated with a condition where maternal
T4 (and T3) concentrations are normal (subclinical hypothyroidism) versus a condition involving
decreased levels of T4 (hypothyroxinemia) may be different.
With regard to what level of decrease in thyroid hormone (e.g., T4) is sufficient for anatomical
and/or functional alterations, particularly in neurodevelopment in developing fetuses or newborns,
several studies have identified a range of T4 decrements associated with neurodevelopmental
health outcomes across humans or experimental rodents. 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 (Haddow et al.. 1999). In other studies, mild-to-moderate
thyroid insufficiency in pregnant women was defined as having serum T4 levels below the 10th
percentile for the study population, which was associated with a 15%—30% decrease relative to the
corresponding median (Finken et al.. 2013; Julvez et al.. 2013; Roman et al.. 2013; Henrichs et
al.. 2010). In experimental animals, decreases in mean maternal T4 levels of ~10%-17% during
pregnancy and lactation have been found to elicit neurodevelopmental toxicity in rat offspring
(Gilbert et al.. 2016; Gilbert. 2011). With regard to a general diagnostic criterion to delineate
hypothyroxinemia from other types of clinical hypothyroidism, the Controlled Antenatal Thyroid
Study (CATS), conducted in a large cohort of pregnant women in Europe, resulted in the
identification of a condition referred to as 'isolated hypothyroxinemia' and is defined as the
67

-------
presence of free thyroxine (FT4) below the 2.5th percentile with a thyrotropin (TSH) level within
the reference range (Hales et al.. 2018; Lazarus et al.. 2012; Negro et al.. 2011). However, as
there is no clear or consistent biological threshold for T4 changes specifically associated with
untoward developmental health outcomes, a BMR of 0.5 SD was identified as a default when
performing BMD modeling on thyroid hormone alterations in offspring, consistent with EPA
BMD Technical Guidance (U.S. EPA. 2012). Further, while total T4 (TT4), free T4 (FT4), and
TSH dose-response data are BMD modeled (see Table 9), important biological considerations
are presented in section 6.1.1.2 that delineates total T4 (TT4) as the key hormone metric for a
developing fetus/neonate.
Significantly decreased thyroid hormone (e.g., T4 and T3) was observed in adult rats exposed
twice daily to oral K+PFBS (NTP. 2019) for 28-day s, as well as the P0 (maternal) mice of the
Feng et al. (2017) study. No overt signs of traditional hypothyroidism such as increased TSH and
increased thyroid tissue weight or histopathology were observed in either adult population. Adult
rodents have a considerable reserve thyroid hormone capacity, compared to the developing
offspring that depend on the supply from maternal T4. While there is concern over decreases in
thyroid hormone (i.e., hypothyroxinemia) in developmental life stages due to critical endocrine
dependency of in utero and neonatal development, the levels at which there is concern for
hypothyroxinemia in euthyroid adults is unclear. As such, for euthyroid adult rats and mice, a
biologically significant level of change was not determined for the BMR as it is unclear what
magnitude of hormone perturbation would be considered adverse. Therefore, for thyroid
hormone effects in adult rodents, a default BMR of 1 SD from control mean was applied. Section
6.1.1.2 presents critical distinctions between perturbations in thyroid hormone economy in adults
versus developing fetus/neonates, resulting in the use of different BMRs across lifestages (e.g., 1
SD for adults, 0.5 SD for newborns).
For kidney hyperplasia data from the subchronic-duration study by Lieder et al. (2009a) and two-
generation reproductive toxicity study by Lieder et al. (2009b). a BMR of 10% extra risk was
used because it is the recommended approach for dichotomous data in the absence of information
on the minimally significant level of change.
Approach for Animal-Human Extrapolation of Perfluorobutane Sulfonic Acid (PFBS) Dosimetry
As discussed in Section 1.3, toxicokinetic data exists for PFBS in relevant animal species (i.e., rats
and mice) and humans, such that a data-informed adjustment approach for estimating the
dosimetric adjustment factor (DAF) can be used. In Recommended Use of Body Weight4 as the
Default Method in Derivation of the Oral Reference Dose (U.S. EPA. 201 lb), the EPA endorses a
hierarchy of approaches to derive human equivalent oral exposures from data from laboratory
animal species, with the preferred approach being physiologically based toxicokinetic modeling.
Other approaches might include using chemical-specific information, without a complete
physiologically based toxicokinetic model. In the absence of chemical-specific models or data to
inform the derivation of human equivalent oral exposures, the EPA endorses BW3/4 as a default to
extrapolate toxicologically equivalent doses of orally administered agents from all laboratory
animals to humans for the purpose of deriving an RfD under certain exposure conditions.
The EPA concluded that data for PFBS are adequate to support derivation of data-informed
dosimetric adjustment. Briefly, the ratio of the clearance (CL) in humans to animals, CLh/CLa,
can be used to convert an oral dose-rate in experimental animals (mg/kg/d) to a human
68

-------
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:
r (m9 / \ - fabs ¦ dose (mg/kg/hr) /
avg \ i ml)	/CL (ml/kg/hr)
where fabs is the fraction absorbed and dose is the average dose rate expressed at an hourly rate.
Assuming equal toxicity given equal Cavg in humans as in mice or rats, and that fabs is the same
in humans as animals, the equitoxic dose, human equivalent dose (HED) (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:
POD CLh
HED = 77;	= POD x —
CLa/	cla
/clh
Thus, the DAF could be calculated as simply CLh/CLa, the ratio of clearance in humans to
clearance in the animal from which the POD is obtained. However, clearance values are not
reported for humans in the available toxicokinetic studies for PFBS (Xu et al.. 2020: 01sen et al..
2009). As clearance is a measure of average elimination, in order to calculate clearance in the
absence of the information, one also needs to evaluate a companion variable, the volume of
distribution (Vd). Neither 01 sen et al. (2009) nor Xu et al. (2020) reported the Vd for humans.
However, there is evidence suggesting that Vd for PFBS is relatively similar across species
including rodents (e.g., 0.12-0.29 L/kg across male and female rats following 10 mg/kg i.v. dose)
and monkeys (e.g., 0.21-0.25 L/kg across male and female cynomolgus macaques following 10
mg/kg i.v. dose) (Chengelis et al.. 2009: 01sen et al.. 2009). Therefore, it is reasonable to assume
Vd for humans is approximately equivalent to Vd for animals (i.e., Vd h = Vd a), in which case
clearance and half-life are inversely related as follows:
Clearance (ml/kg/hr) = ln(2) x		—- x Vd
^1/2 \hr) ^ K9>
Since reliable measures of half-life in humans and animals are available for PFBS, the ratio of
elimination half-life in animals from which the POD is obtained to that in humans, to.5,A/to.5,H, can
be used to calculate the DAF, and the human equivalent dose (HED) can be calculated as
follows:
tl/2
HED = POD X	4
11/2h
As described in Section 1.3, two studies evaluated the elimination of human serum K+PFBS in
human populations with previous occupational exposure (Xu et al.. 2020: Olsen et al.. 2009). Initial
blood concentrations of PFBS in the population examined by Xu et al. (2020) are more
representative of environmental exposure and the population was larger including eleven male
69

-------
and six female employees when compared to Olsen et al. (2009). While the estimated serum
half-life of PFBS reported by Olsen et al. (2009) overlapped with those by Xu et al. (2020)
(mean=43.8 d, range = 21.9-87.6 d), there is a statistically significant difference between these
two studies. As such, the two data sets will not be combined and the half-life estimated by Xu et
al. (2020) is presumed to better predict human dosimetry at environmental levels. The average
half-life reported by Xu et al. (2020) (mean = 43.8 d = 1,050 h) was assigned for tv^H.
One study evaluated the elimination of serum PFBS in mice. Lau et al. (2020) reported serum
terminal half-lives of 5.8 hours in male mice and 4.5 hours in female mice. Since the half-life
estimates did not vary significantly between the doses (i.e., 30 and 300 mg/kg), these parameter
estimates were combined. However, there was a statistically significant difference in the half-life
estimates between sexes (female mice [4.5 h] had a slightly shorter half-life compared to males
[5.8 h]), so sex-specific half-lives were assigned for to.5Afor mice.
Two studies were used to calculate serum half-life estimates for dosimetric adjustment in rats
(Huang et al.. 2019a; Olsen et al.. 2009). A numerical average of the terminal half-lives (ti/2,p)
measured in rats after oral and i.v. doses is identified in Olsen et al. (2009) as 4.6 hours in males
and 5.7 hours in females. Olsen et al. (2009) reports sex-specific elimination differences in half-
life values in rats. A numerical average of the terminal half-lives (ti/2,p) measured in male rats
after oral and i.v. doses in Huang et al. (2019a) is 4.9 hours. In male rats, half-life values
reported in Olsen et al. (2009) and Huang et al. (2019a) are consistent, thus were averaged for
use in dosimetric adjustment resulting in a geometric mean terminal serum half-life of 4.8 hours.
The terminal half-life value reported by Huang et al. (2019a) in female rats after a 4 mg/kg i.v.
dose of PFBS is 0.95 hours. Following oral, exposure Huang et al. (2019a) was not able to fit the
data to a two-compartment model, thus do not report a terminal half-life (ti/2,p). For this reason,
the mean female terminal half-life (ti/2,p) value from Olsen et al. (2009) was used for dosimetric
adjustment.
Table 8 presents the DAFs for converting rat and mice PODs to HEDs for PFBS.
Table 8. Mouse, Rat, and Human half-lives and data-informed dosimetric adjustment
factors
Species
Sex
Animal ti/2 (h)
Human ti/2 (h)
DAF (tl/2,A/tl/2,H)
Mouse
Male
5.81

0.0055
Female
4.52
1,0505
0.0043
Rat
Male
4.83
0.0046
Female
5.T

0.0054
'Terminal serum half-life of combined doses for male mice from Lau et al. f2020")
2Terminal serum half-life of combined doses for female mice from Lau etal. ("20201
3Geometric mean of terminal serum half-lives (ti/2,p) measured after all oral and i.v. doses for male rats from Olsen etal.
("20091 and Huang et al. (~2019al
4Mean of terminal serum half-lives (ti/2,p) measured after oral and i.v. doses for female rats from Olsen etal. ("20091
5Mean serum elimination half-life for humans (combined sexes) from Xu etal. ("20201
Where modeling was feasible, the estimated BMDLs were identified as PODs (summarized in
Table 9). Further details, including the modeling output and graphical results for the model
selected for each endpoint, can be found in HAWC and are discussed in Appendix F. Where
70

-------
dose-response modeling was not feasible, NOAELs or LOAELs were identified (summarized in
Table 9).
71

-------
Table 9. PODs considered for the derivation of the subchronic RfD for K+PFBS (CASRN 29420-49-3)
Endpoint/reference
Species/life stage—
sex
PODiin)1
(mg/kg-d)
Comments*
Thyroid effects
Total T4—Fens et al. (2017)
Mouse/Po—Female
BMDLisd — 0.093
Adequate model fit
Free T4—Fens et al. (2017s)
Mouse/Po—Female
NOAEL = 0.21
No models provided adequate statistical or visual fit to mean
responses
TSH—Fens etal. (2017)
Mouse/Po—Female
NOAEL = 0.21
No models provided adequate statistical or visual fit to mean
responses
Total T4 PND 1 (fetal n)b —Fens et al.
(2017s)
Mouse/Fi—Female
NOAEL = 0.21
No models provided adequate fit to the data, specifically variance
Total T4 PND 1 (litter n)b —Fens et al.
(2017s)
Mouse/Fi—Female
BMDLo.5sd = 0.095
(BMDLisd = 0.25)
Adequate model fit
Total T4 PND 30—Fens et al. (2017s)
Mouse/Fi—Female
NOAEL = 0.21
No models provided adequate statistical or visual fit to mean
responses
Total T4 PND 60—Fens et al. (2017s)
Mouse/Fi—Female
NOAEL = 0.21
No models provided adequate fit to the data, specifically variance
TSH PND 30—Fens et al. (2017s)
Mouse/Fi—Female
NOAEL = 0.21
No models provided adequate statistical or visual fit to mean
responses
Total T4—NTP (2019s)
Rat—Male
LOAEL = 0.34
No models provided adequate statistical or visual fit to mean
responses
Rat—Female
BMDLisd = 0.037
Adequate model fit
Free T4—NTP (2019s)
Rat—Male
LOAEL = 0.34
No models provided adequate statistical or visual fit to mean
responses
Rat—Female
BMDLisd = 0.027
Adequate model fit
Developmental effects
Eves openins (fetal n)h—Fens et al. (2017s)
Mouse/Fi—Female
NOAEL = 0.21
No models provided adequate fit to the data, specifically variance
Eves ODcnins (litter n)h —Fens et al. (2017s)
Mouse/Fi—Female
BMDLo.5sd = 0.073
(BMDLisd = 0.16)
Adequate model fit
Vasinal ODcnins (fetal n)b —Fens et al.
(2017s)
Mouse/Fi—Female
BMDLo.5sd = 0.15
(BMDLisd = 0.35)
Adequate model fit
Vasinal ODcnins (litter n)b —Fens et al.
(2017s)
Mouse/Fi—Female
BMDLo.5sd = 0.094
(BMDLisd = 0.22)
Adequate model fit
First estrous (fetal n)b —Fens et al. (2017s)
Mouse/Fi—Female
NOAEL = 0.21
No models provided adequate statistical or visual fit to mean
responses
72

-------
Endpoint/reference
Species/life stage—
sex
PODmi)1
(mg/kg-d)
Comments*
First estrous (litter n)b —Fens et al. (20171
Mouse/Fi—Female
NOAEL = 0.21
No models provided adequate statistical or visual fit to mean
responses
Kidney effects

Kidney histopathology—papillary
epithelial tubular/ductal
hyperplasia—Lieder et al. (2009a1
Rat—Male
BMDLio = 0.49
Adequate model fit
Rat—Female
BMDLio ~ 0.30
Adequate model fit
Kidney histopathology—papillary
epithelial tubular/ductal
hyperplasia—Lieder et al. (2009b1
Rat/Po—Male
BMDLio = 0.35
Adequate model fit
Rat/Po—Female
BMDLio = 0.27
Adequate model fit
Kidney histopathology—papillary
epithelial tubular/ductal
hyperplasia—Lieder et al. (2009b1
Rat/Fi—Male
BMDLio ~ 0.78
Adequate model fit
Rat/Fi—Female
BMDLio = 0.48
Adequate model fit
Notes:
BMDLo.5sd = benchmark dose lower confidence limit for 0.5 SD change from the control, BMDLio = 10% benchmark dose lower confidence limit; BMDLisd = benchmark dose lower confidence
limit for 1 SD change from the control.
a Following U.S. EPA ^201 lb') and (U.S. EPA. 2014d1 guidance, animal doses from candidate principal studies were converted to HEDs through the application of a dosimetric adjustment factor
(DAF), where FEED = dose x DAF.
b Fetal endpoints from Feng et al. ('2017') were modeled alternatively using dose group sizes based either on total number of fetuses or dams. Given that it appears that
Feng etal. ("20171 did not use the litter as the statistical unit of analysis, it is unclear if the study-reported standard errors pertain to litters or fetuses. Alternatively, modeling fetal endpoints using litter
n or fetal n provides two modeling results that bracket the "true" variance among all fetuses in a dose group (i.e., using the fetal n will under-estimate the true variance while using the litter n will
over-estimate the true variance). Individual animal data were requested from study authors but were unable to be obtained,
i BMD modeling methods and links to modeling inputs and results in F1AWC are found in appendix F.
F1AWC visualization: Candidate POPs for Subchronic and Chronic RfD
73

-------
6.1.1.2 Considerations for Selection of Critical Effect for Derivation of RfDs
The evidence for the thyroid, developmental, and kidney effect domains support a hazard via the
oral exposure route (Table 7). However, qualitative and quantitative differences in the strength of
evidence between these effect domains are present (Table 9). PFBS-induced perturbation of the
thyroid was consistently observed across two species, sexes, life stages, and exposure durations
in two independent, high-confidence studies. These perturbations involved a coherent pattern of
hormonal changes with similar sensitivity in the POD ranges across lifestages (e.g., maternal and
PND1/newborn BMDLoss of 0.093 and 0.095 mg/kg-day, respectively). Developmental effects
(e.g., delayed eyes opening, vaginal opening, or first estrous) were observed in mouse litters in
which decrements in thyroid hormone occurred and with similar sensitivity in the ranges of POD
estimates (i.e., 0.073-0.21 mg/kg-day) (Feng et al.. 2017). However, these developmental effects
have been reported in a single study and species (mouse). Kidney effects in adult animals (Lieder
et al.. 2009a; Lieder et al.. 2009b) were observed in adult or developing rats across high- or
medium-confidence gavage studies of various duration; however, were less sensitive at 0.27
mg/kg-day and above.
In the derivation of a subchronic RfD, the Feng et al. (2017) and NTP (2019) studies were both
considered for potential principal study due to the observed sensitivity of thyroid hormone
decrements. However, the biological significance of hypothyroxinemia (i.e., decreased T4) in
adult euthyroid animals, absent additional signs of overt thyroid toxicity (e.g., reflex increase in
TSH and/or alterations in tissue weight or histology), is unclear; therefore, the thyroid effects
from the NTP (2019) rat study were not selected as a critical effect. The gestational exposure
study in mice was selected as the principal study for derivation of the subchronic RfD based on
thyroid effects. The gestational exposure study conducted by Feng et al. (2017) reports
administration of K+PFBS by gavage in ICR mice (10/dose) from GDs 1 to 20. This study was of
good quality (i.e., high confidence) with adequate reporting and consideration for appropriate
study design, methods, and conduct (click to see risk of bias analysis in HAWC). Feng et al.
(2017) reported statistically significantly decreased total T3, total T4, and free T4, as well as
increased TSH in dams and offspring (increased TSH PND 30 only) gestationally exposed to
PFBS.
The critical effect from the Feng et al. (2017) study is decreased serum total thyroxine (T4) in
newborn (PND 1) mice. T4 and T3 are essential for normal growth of developing offspring
across animal species (for review see Forhead and Fowden (2014)). And, previous studies have
shown that exposure to other PFAS during pregnancy results in lower T4 and T3 levels in
pregnant women and fetuses or neonates (Yang et al.. 2016; Wang et al.. 2014). The selection of
total T4 as the critical effect is based on a number of key considerations (see below) that account
for cross-species correlations in thyroid physiology and hormone dynamics particularly within
the context of a developmental life stage.
A key consideration for selection of total T4 is that this represents the aggregate of potential
thyroid endocrine signaling (i.e., free T4 + protein bound T4) at any given time. In humans, FT4
represents approximately 0.03% of circulating hormone, indicating that as much as 99.97% of all
T4 is protein bound (e.g., albumin; TBG). While T3 is the active hormone form in respondent
somatic tissues, the formation of T3 is contingent upon the deiodination of free T4. A critical
consideration in pregnant females is that T4, not T3, is the thyroid hormone that crosses the
placenta of humans and rodents. Although free T4 might be considered a suitable measure of
74

-------
thyroid hormone status in non-developmental (e.g., adult) life stages, there are some important
factors associated with maintenance of the microenvironment for developing offspring in utero that
lends credence to the use of total T4 as the critical effect. A tightly regulated transfer of maternal
thyroid hormone to a fetus is paramount to proper development of multiple tissues and organ
systems (e.g., nervous system), especially during the early trimesters. The placenta has
transporters and deiodinases that collectively act as a gatekeeper to maintain an optimal T4
microenvironment in the fetal compartment (Fisher, 1997; Koopdonk-Kool et al., 1996). For
example, deiodinase 3 (D3) is highly expressed in human uterus, placenta, and amniotic
membrane, where it serves a critical role of regulating thyroid hormone transfer to the fetus
through the deiodination of T4 to transcriptionally inactive reverse triiodothyronine (rT3) or T3
to inactive 3,5-diiodo-L-thyronine (T2). Similarly, Wasco et al. (2003) showed that D3 is highly
expressed in rodent uterus and is highly induced during pregnancy. Further, the Dio3 gene that
encodes D3 has been shown to be imprinted in the mouse (Hernandez et al., 2002), suggesting a
pivotal role for this specific deiodinase in the mouse as well. Indeed, the human and rodent
placenta have been shown to be similarly permeable to T4 and T3 (Fisher, 1997; Calvo et al.,
1992). Due to placental barrier functionality, free T4 levels in a pregnant dam might not be
entirely representative of actual T4 status in a developing fetus. Further, the American Thyroid
Association published a Guidelines document in 2017 in which they stated "Current uncertainty
around FT4 estimates in pregnancy has led some to question the wisdom of relying on any FT4
immunoassays during pregnancy. In contrast, measurement of TT4 and the calculated FT4 index
do show the expected inverse relationship with serum TSH. This finding suggests that TT4
measurements may be superior to immunoassay measurement of FT4 measurements in pregnant
women " (Alexander et al., 2017) Thus, decreased total T4 in offspring (and dams during
pregnancy/at delivery) is expected to be more representative of PFBS-mediated thyroid effects
and potentially associative developmental effects.
There are some differences in HPT development and functional maturation and regulation during
early life stages (e.g., timing of in utero and early postnatal thyroid development) between
humans and rodents (for a comprehensive overview see (Regulatory Science Associates, 2019)).
Human thyroid development occurs in three phases in utero which entails initial development of
the gland between embryonic day 10 to gestational week 11 (Phase I), maturation of the fetal
thyroid system from gestational weeks 11-35 (Phase II), and further refinement of hypothalamic-
pituitary-thyroid axis functionality during the latter portion of gestation up to approximately 4
weeks into the postnatal period (Phase III) (Klein et al„ 1982; Fisher and Klein, 1981).
Importantly, in utero development of the rodent thyroid gland occurs in the same phases and
order as humans, the difference being that rodents are essentially born during Phase II with Phase
III occurring almost exclusively postnatally; whereas in humans, Phase III is well underway in
utero and completes postnatally. As such, rodent neurodevelopment in the early postnatal phase
is analogous to the third trimester of human development in utero (Gilbert et al„ 2012). Further,
fetal development of rodents in utero is entirely dependent on maternal thyroid hormone until
approximately gestational day 17-18, whereas in humans fetal development transitions from
complete reliance on maternal thyroid hormone during the first trimester (i.e., thyroid
development Phase I) to a mix of fetal thyroid hormone synthesis and maternal transplacental
hormone transfer beginning in the second trimester (i.e., thyroid development Phase II) through
the in utero portion of Phase III (Fisher and Klein, 1981).
75

-------
Within the context of early developmental life stages, there are several commonalities in HPT
dynamics between humans and rodents such as similar profiles of (1) thyroid hormone binding
proteins, (2) hormone functional reserve, and (3) placental deiodinase. For example, two carrier
proteins—thyroid binding globulin (TBG) and transthyretin (TTR)—are primarily responsible
for storage and transit of T4 in mammals (Rabah et al.. 2019). TBG is the primary carrier of T4
in humans across all life stages (Savu et al.. 1991). Importantly, in fetal and infant rats, TBG is
also the primary carrier of T4 (Savu et al., 1989). As rats transition to adulthood, TTR takes over
as the primary carrier of T4. In addition, as a relatively highly abundant carrier protein, albumin
also plays a role in thyroid hormone binding and transit in humans and rodents; however, the
relative affinity for binding is lower than either TBG or TTR.
Life stage-specific differences in thyroid hormone reserve capacity between adults and neonates
have been noted. On average, intrathyroidal thyroglobulin stores in adults are on the order of
months whereas in neonates the functional reserve is approximated at less than 1 day (Gilbert
and Zoeller. 2010; Savin et al.. 2003; Van Den Hove et al.. 1999). This suggests that the adult
thyroid has compensatory abilities not present in early life stages, making fetal/neonatal
populations particularly sensitive to perturbations in thyroid hormone economy (e.g.,
hypothyroxinemia). And, although the timing of thyroid development can vary between species
(Forhead and Fowden. 2014). the dynamic reserve capacity of T4 between humans and rodents
near birth and in early postpartum might not be significantly different. For example, human
neonates have a serum half-life of T4 of approximately 3 days (Vulsma et al.. 1989). and thyroid
tissue stores of T4 are estimated to be less than 1 day (Van Den Hove et al„ 1999). As the
developing rodent thyroid does not begin producing its own hormone until late in gestation
(> GD 17), newborn rodent T4 levels are primarily a reflection of transplacental^ translocated
maternal hormone; and adult rats have been shown to have a serum T4 half-life of 0.5-1 day
(Choksi et al.. 2003). As such, significant differences in functional thyroid reserve capacity
between human and rodent neonates is not anticipated.
Accounting for the information presented above, the range of values for the subchronic RfD,
based on the BMDLo.ssd (HED) of 0.095 mg/kg-day for decreased serum total T4 in newborn
(PND 1) mice, is derived as follows:
Subchronic RfD range for K+PFBS= BMDLo.ssd (HED) UFc
= 0.095 mg/kg-day 100 or 30
= 0.00095 to 0.0032 mg/kg-day
= 1 x 10"3 to 3* 10"3 mg/kg-day
Table 10 summarizes the UFs for the subchronic RfD for K+PFBS based on effects in the
thyroid.. In the process of developing the subchronic and chronic RfDs, scientific rationales
were provided for assigning a value for the database uncertainty factor (UFD)of 1 and of 3. Each
argument was considered by EPA to have merit. Therefore, EPA has presented RfDs for K+
PFBS and for PFBS (free acid) derived using both an UFd of 1 and an UFd of 3. Risk assessors
may evaluate the justifications for application of either UFd and decide whether the risk scenario
under consideration warrants use of the higher or lower RfD considering the purpose and scope
76

-------
of their risk assessment and the decision-making it supports, i.e., which is fit-for-purpose of the
specific risk assessment10.
Table 10. UFs for the subchronic RfD for thyroid effects for K+PFBS (CASRN 29420-49-3)
UF
Value
Justification
UFa
3
A UFa of 3 (10°5) is applied to account for uncertainty in characterizing the toxicokinetic and
toxicodynamic differences between mice and humans following oral K+PFBS/PFBS exposure. Some
aspects of the cross-species extrapolation of toxicokinetic and toxicodynamic processes have been
accounted for by calculating a HED by applying a DAF as outlined in the EPA's Recommended Use
ofBodv Weight3/4 as the Default Method in Derivation of the Oral Reference Dose ('U.S. EPA.
201 lb). However, some residual uncertainty remains in the relative cross-soccics sensitivity in
toxicodynamics (e.g., thyroid signaling). Thus, in the absence of chemical-specific data to quantify
these uncertainties, EPA's guidance recommends use of a UFA of 3.
UFh
10
A UFh of 10 is applied to account for interindividual variability in the human populations because of
both intrinsic (toxicokinetic, toxicodynamic, genetic, life stage, and health status) and extrinsic (life
style) factors that can influence the response to dose. In the absence of chemical-specific data to
quantify this variability in the toxicokinetics and toxicodynamics of K+PFBS/PFBS in humans, EPA
recommends use of a UFH of 10.
UFd
3 or 1
A UFd of 3 or 1 is may be applied due to database deficiencies. The oral exposure database contains
multiple short-term and subchronic-duration toxicity studies of laboratory animals (NTP. 2019;
Biiland et al.. 2011; 3M. 2010; Lieder et al.. 2009a; 3M. 2001. 2000d). a two-generation reproductive

toxicity study in rats (Lieder et al.. 2009b). and multiple developmental toxicity studies in mice and
rats (Fens et al.. 2017; York. 2002). The observation of decreased thyroid hormone is known to be a
crucial element during developmental life stages, particularly for neurodevelopment, and the database
is limited by the lack of developmental neurotoxicity studies, which would warrant a UFd value of 3.
However, deficits in thyroid hormone are a precursor event to the potential for adverse effects on the
developing brain. Therefore, selecting a critical study and endpoint that would protect against the
thyroid effects would protect against potential adverse effects on the developing brain, thereby
justifying a reduced UFD value of 1. In addition, as other health effect domains such as
immunotoxicity and mammary gland development are effects of increasing concern across several
members of the lareer PFAS family (Grandiean. 2018; Liew et al.. 2018; White et al.. 2007); however
studies evaluating these outcomes following PFBS exposure exist for subchronic exposures.
UFl
1
A UFl of 1 is applied for LOAEL-to-NOAEL extrapolation because the POD is a BMDL and the
BMR was selected based on evidence that it represented a minimal biologically significant response
level in susceptible populations such as developing offspring.
UFs
1
A UFS of 1 is applied because the POD comes from a developmental study in mice. The
developmental period is recognized as a susceptible life stage in which exposure during certain time
windows (e.g., gestational) is more relevant to the induction of developmental effects than lifetime
exposure (U.S. EPA. 1991a).
UFC
100 or
30
Composite UF = UFA x UFH x UFD x UFL x UFS
10 Uncertainty factors (UFs) were a consideration during peer review. Within the context of the scientifically-
justifiable UFd, the choice about which of the two UFDs to use is a policy judgment that has been delegated to the
risk assessor. The choice of the UFd is a decision best made within the context of a fit-for-purpose risk assessment,
which includes an understanding of flexibility and necessary degree of certainty.
77

-------
The data for K+PFBS can be used to derive a subchronic RfD for the free acid (PFBS), as
K+PFBS is fully dissociated in water at the environmental pH range of 4-9 (NICNAS, 2005). To
calculate the subchronic RfD for the free acid, the subchronic RfD for the potassium salt is
adjusted to compensate for differences in MW between K+PFBS (338.19) and PFBS (300.10).
The range of values for the subchronic RfD for PFBS (free acid) is calculated as follows:11
Subchronic RfD range = RfD for K+PFBS salt x (MW free acid ^ MW salt)
for PFBS (free acid) = 0.00095 to 0.0032 mg/kg-day x (300.10 - 338.19)
= 0.00095 to 0.0032 mg/kg-day x (0.89)
= 0.00085 to 0.0028 mg/kg-day
= 9 x 10"4to 3 x 10"3 mg/kg-day
Confidence in the range of values for the subchronic RfD for PFBS and K+PFBS for thyroid
effects is medium, as explained in Table 11.
Table 11. Confidence descriptors for the subchronic RfD for PFBS (CASRN 375-73-5) and
the related compound K+PFBS (CASRN 29420-49-3)
Confidence categories
Designation
Discussion
Confidence in study
H
Confidence in the principal study is high because the overall
study design, performance, and characterization of exposure
was sood. Studv details and risk of bias analvsis can be found
inHAWC.
Confidence in database
M
Confidence in the oral toxicity database for derivation of the
candidate subchronic RfD for thyroid effects is medium
because although there are multiple developmental toxicity
studies in mice and rats, no studies are available that have
specifically evaluated neurodevelopmental, immunological, or
mammary gland effects. In addition, available toxicokinetic
studies are limited (e.g., one mouse toxicokinetic study) and
toxicokinetic data do not exist for PFBS at all life stages,
including neonates, infants, and children. Additionally,
studies are not available to estimate the relative cross-species
sensitivity in toxicodynamics (e.g., thyroid signaling).
Confidence in candidate subchronic
RfD
M
The overall confidence in the candidate subchronic RfD for
thyroid effects is medium.
Notes'. H = high; M = medium
11 The subchronic RfD for PFBS (free acid) is provided as a range defined by either the use of an UFD of 1 or an
UFd of 3. Risk assessors may evaluate the justifications for application of either UFD and decide whether the risk
scenario under consideration warrants use of the higher or lower RfD considering the purpose and scope of their risk
assessment and the decision-making it supports, i.e., which is fit-for-purpose of the specific risk assessment.
78

-------
The range of values for the subchronic RfD is derived to be protective of all types of effects
across studies and species following oral subchronic exposure and is intended to protect sensitive
subpopulations and life stages.
6.1.2 Derivation of the Chronic RfD
There are no chronic-duration studies available for PFBS and K+PFBS. Therefore, based on the
same database and similar considerations as the subchronic RfDs, the range of values for the
noncancer chronic RfD is derived, based on the same BMDLo.ssd (HED) of 0.16 mg/kg-day for
decreased serum total T4 in newborn (PND 1) mice (Feng et al.. 2017). as follows:
Chronic RfD range for K+PFBS = BMDLo.ssd (HED) UFc
= 0.095 mg/kg-day 300 or
100
= 0.00032 to
0.00095 mg/kg-day
3 x 10"4 to 1 x 10"3
mg/kg-day
Table 12 summarizes the UFs for the chronic RfD for K+PFBS based on effects in the thyroid.
In the process of developing the subchronic and chronic RfDs, scientific rationales were
provided for assigning a value for the database uncertainty factor (UFD)of 1 and of 3. Each
argument was considered by EPA to have merit. Therefore, EPA has presented RfDs for K+
PFBS and for PFBS (free acid) derived using both an UFd of 1 or an UFd of 3. Risk assessors
may evaluate the justifications for application of either UFd and decide whether the risk scenario
under consideration warrants use of the higher or lower RfD considering the purpose and scope
of their risk assessment and the decision-making it supports, i.e., which is fit-for-purpose of the
specific risk assessment12.
12 Uncertainty factors (UFs) were a consideration during peer review. Within the context of the scientifically-
justifiable UFd, the choice about which of the two UFDs to use is a policy judgment that has been delegated to the
risk assessor. The choice of the UFd is a decision best made within the context of a fit-for-purpose risk assessment,
which includes an understanding of flexibility and necessary degree of certainty.
79

-------
Table 12. UFs for the chronic RfD for thyroid for K+PFBS (CASRN 29420-49-3)
UF
Value
Justification
UFa
3
A UFa of 3 (10°5) is applied to account for uncertainty in characterizing the toxicokinetic and
toxicodynamic differences between mice and humans following oral K+PFBS/PFBS exposure. Some
aspects of the cross-species extrapolation of toxicokinetic and toxicodynamic processes have been
accounted for by calculating a HED by applying a DAF as outlined in the EPA's Recommended Use
ofBodv Weieht3/4 as the Default Method in Derivation of the Oral Reference Dose ('U.S. EPA.
201 lb). However, some residual uncertainty remains in the relative cross-soccics sensitivity in
toxicodynamics (e.g., thyroid signaling). Thus, in the absence of chemical-specific data to quantify
these uncertainties, EPA's guidance recommends use of a UFA of 3.
UFh
10
A UFh of 10 is applied to account for interindividual variability in the human populations because of
both intrinsic (toxicokinetic, toxicodynamic, genetic, life stage, and health status) and extrinsic (life
style) factors that can influence the response to dose. In the absence of chemical-specific data to
quantify this variability in the toxicokinetics and toxicodynamics of K+PFBS/PFBS in humans, EPA
recommends use of a UFH of 10..
UFd
10 or
3
A UFd of 10 or 3 may be appropriate to account for database deficiencies. The oral exposure database
contains multiple short-term and subchronic-duration toxicity studies of laboratory animals (NTP.
2019; Biiland et al.. 2011; Lieder et al.. 2009a: 3M. 2001. 2000d). a two-generation reproductive
toxicity study in rats (Lieder et al.. 2009b). and multiple developmental toxicity studies in mice and
rats (Fens et al.. 2017; York. 2002). As thyroid hormone is known to be critical during developmental
life stages, particularly for neurodevelopment, the database is limited by the lack of developmental
neurotoxicity studies. However, deficits in thyroid hormone are a precursor event to the potential for
adverse effects on the developing brain. Therefore, selecting a critical study and endpoint that would
protect against the thyroid effects would protect against potential adverse effects on the developing
brain. Due to the lack of chronic duration studies, there is additional uncertainty regarding how
longer-term exposures might impact hazard identification and dose-response assessment for PFBS via
the oral route (e.g., potentially more sensitive effects), which warrant application of a UFD value of
either 10 or 3. Lastly, as immunotoxicity and mammary gland development are effects of increasing
concern across several members of the larger PFAS family (Grandican. 2018; Liew et al.. 2018;
White et al.. 2007); however, studies evaluating these outcomes following PFBS exposure exist for
subchronic exposures.
UFl
1
A UFl of 1 is applied for LOAEL-to-NOAEL extrapolation because the POD is a BMDL and the
BMR was selected based on evidence that it represented a minimal biologically significant response
level in susceptible populations such as developing offspring.
UFS
1
A UFS of 1 is applied because the POD comes from a developmental study of mice. The
developmental period is recognized as a susceptible life stage in which exposure during certain time
windows (e.g., gestational) is more relevant to the induction of developmental effects than lifetime
exposure (U.S. EPA. 1991b). The additional concern over potential hazards following longer-term
(chronic) exposures is accounted for under the UFD above.
UFC
300 or
100
Composite UF = UFA x UFH x UFD x UFL x UFS
The data for K+PFBS can be used to derive a chronic RfD for the free acid (PFBS), as K+PFBS is
fully dissociated in water at the environmental pH range of 4-9 (NICNAS. 2005). In order to
calculate the chronic RfD for the free acid, the chronic RfD for the potassium salt is adjusted to
compensate for differences in molecular weight between K+PFBS (338.19) and PFBS (300.10).
80

-------
The chronic RfD for PFBS (free acid) for thyroid effects is the same as the value for the K+PFBS
salt. The chronic RfD for PFBS (free acid) is calculated as follows:13
Chronic RfD range = RfD for K+PFBS salt x (MW free acid MW salt)
for PFBS (free acid) = 0.00032 to 0.00095 mg/kg-day x (300.10 - 338.19)
= 0.00032 to 0.00095 mg/kg-day x (0.89)
= 0.00028 to 0.00084 mg/kg-day
= 3 x 10"4 to 1 x 10"3 mg/kg-day
Confidence in the range of values for the chronic RfD for PFBS and K+PFBS for thyroid effects
is low, as explained in Table 13 below.
Table 13. Confidence descriptors for chronic RfD for PFBS (CASRN 375-73-5) and the
related compound K+PFBS (CASRN 29420-49-3)
Confidence categories
Designation
Discussion
Confidence in study
H
Confidence in the principal study is high because the overall study
design, performance, and characterization of exposure was good. Study
details and risk of bias analysis can be found in HAWC.
Confidence in database
L
Confidence in the oral toxicity database for derivation of the chronic
RfD is low because, although there are multiple short-term studies and
a subchronic-duration toxicity study in laboratory animals, one
acceptable two-generation reproductive toxicity study in rats, and
multiple developmental toxicity studies in mice and rats, the database
lacks any chronic duration exposure studies or studies that have
evaluated neurodevelopmental, immunological, or mammary gland
effects. In addition, available toxicokinetic studies are limited (e.g., one
mouse toxicokinetic study) and toxicokinetic data do not exist for
PFBS at all life stages, including neonates, infants, and children.
Additionally, studies are not available to estimate the relative cross-
species sensitivity in toxicodynamics (e.g., thyroid signaling).
Confidence in candidate
chronic RfD
L
The overall confidence in the candidate chronic RfD for thyroid effects
is low.
Notes'. H = high; L = low
The range of values for the chronic RfD is derived to be protective of all types of effects across
studies and species following oral chronic exposure and is intended to protect the population as a
whole, including potentially susceptible populations and life stages (U.S. EPA. 2002). The
individual value applied will depend on the needs of the program office and in the type of risk
assessment being performed (e.g., general population). Decisions concerning averaging
exposures over time for comparison with the RfDs should consider the types of toxicological
effects and specific life stages of concern. For example, fluctuations in exposure levels that result
13 The chronic RfD for PFBS (free acid) is provided as a range defined by either the use of an UFD of 1 or an UFD of
3. Risk assessors may evaluate the justifications for application of either UFD and decide whether the risk scenario
under consideration warrants use of the higher or lower RfD considering the purpose and scope of their risk
assessment and the decision-making it supports, i.e., which is fit-for-purpose of the specific risk assessment.
81

-------
in elevated exposures during development could potentially lead to an appreciable risk, even if
average levels over the full exposure duration were less than or equal to either of the RfD values.
6.2 Derivation of Inhalation Reference Concentrations
No published studies investigating the effects of subchronic- or chronic-duration inhalation
toxicity of PFBS and the related compound K+PFBS in humans or animals have been identified.
6.3 Cancer Weight-of-Evidence Descriptor and Derivation of Cancer Risk Values
No studies evaluating the carcinogenicity of PFBS or K+PFBS in humans or animals were
identified. In accordance with the Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005).
the EPA concluded that there is "inadequate evidence to assess carcinogenic potential" for PFBS
and K+PFBS by any route of exposure. Therefore, the lack of data on the carcinogenicity of
PFBS and the related compound K+PFBS precludes the derivation of quantitative estimates for
either oral (oral slope factor) or inhalation (inhalation unit risk) exposure.
6.4 Susceptible Populations and Life Stages
Early life stages as well as pregnant women are potentially susceptible to PFBS exposure. PFBS
has been detected in blood serum of nursing mothers, which might indicate a potential for
lactational exposure (Glynn et al.. 2012); however, information on the kinetics of lactational
transfer are lacking, and represents a key data gap for future research.
The available information suggests sex-specific variation in the toxicokinetics of PFBS in
rodents. Studies in mice and rats generally report clearance and elimination half-life times to be
faster for females than for males (see the "Toxicokinetics" section). For example, Lau et al.
(2020) reports statistically significant differences in half-life between the sexes with female mice
exhibiting a shorter half-life compared to males. Similar sex-specific variation in elimination has
been reported in rats. 01 sen et al. (2009) reported a statistically significant difference in the
urinary clearance rates (p < 0.01) with female rats (469 ± 40 mL/hour) having faster clearance
rates than male rats (119 ± 34 mL/h). Huang et al. (2019a) also reported higher clearance in
female rats compared to male rats given the same dose (26.0-75.5 mL/h/kg in males, 152-259
mL/h/kg in females). Chengelis et al. (2009) reported that the mean apparent clearance of PFBS
from the serum was approximately eightfold higher for female rats (0.311 L/h/kg) than for male
rats (0.0394 L/h/kg). Statistically significant sex-related differences in half-life or clearance were
not observed between male and female monkeys (01 sen et al.. 2009). Differences in the
toxicokinetics in rodents could result in sex-specific differences in toxicity studies.
In vivo toxicity studies report that PFBS exposure can alter thyroid hormone levels in parental
and F1 generation animals (see "Thyroid Effects"). Thyroid hormones play a critical role in
coordinating complex developmental processes for various organs/systems (e.g., reproductive
and nervous system), and disruption of thyroid hormone production/levels in a pregnant woman
or neonate can have persistent adverse health effects for the developing offspring (Ghassabian
and Trasande. 2018; Foster and Gray. 2013; Julvez et al.. 2013; Roman et al.. 2013).
Animal studies also provide evidence that gestationally exposed females might be a susceptible
subpopulation because of potential effects on female reproduction, including evidence of altered
ovarian follicle development and delayed vaginal opening (see "Reproductive Effects").
82

-------
Furthermore, gestationally exposed females also had significantly reduced BWs and delayed eye
opening. These findings suggest that developmental landmarks indicative of adverse responses
can be affected after PFBS exposure (see "Offspring Growth and Early Development").
83

-------
Appendix A: Literature Search Strategy
This appendix presents the full details of the literature search strategy used to identify primary,
peer-reviewed literature pertaining to perfluorobutane sulfonic acid (PFBS) (Chemical Abstracts
Service Registry Number [CASRN] 375-73-5) and/or the potassium salt (K+PFBS) (CASRN
29420-49-3) and the deprotonated anionic form of PFBS (i.e., PFBS"; CASRN 45187-15-3).
Initial database searches were conducted on July 18, 2017 using four online scientific databases
(PubMed, Web of Science [WOS], Toxline, and TSCATS via Toxline) and updated on
February 28, 2018 and May 1, 2019. The literature search focused on chemical name and
synonyms (see Table A-l) with no limitations on publication type, evidence stream (i.e., human,
animal, in vitro, and in silico) or health outcomes. Beyond database searches, references were
also identified from studies submitted under TSCA and from review of other government
documents (e.g., Agency for Toxic Substances and Disease Registry [ATSDR]) and combined
with the results of the database search. Search results are retained in the EPA's Health and
Environmental Research Online (HERO) database.
Table A-l. Synonyms and MESH terms
Cliem II)
375-73-5
1,1,2,2,3,3,4,4,4-Nonafluoro-l-butanesulfonic acid
1-Perfluorobutanesulfonic acid
Nonafluoro-l-butanesulfonic acid
Nonafluorobutanesulfonic acid
Perfluorobutanesulfonic acid
PFBS
1,1,2,2,3,3,4,4,4-Nonafluorobutane-l-sulphonic acid
PubMed (new only)
Perfluorobutane sulfonic acid
Perfluorobutanesulfonate
Perfluorobutane sulfonate
EPA Spreadsheet
1,1,2,2,3,3,4,4,4-Nonafluoro-l-butanesulfonic acid
1-Butanesulfonic acid, 1,1,2,2,3,3,4,4,4-nonafluoro-
1-Butanesulfonic acid, nonafluoro-
1-Perfluorobutanesulfonic acid
Nonafluoro-l-butanesulfonic acid
Nonafluorobutanesulfonic acid
PFBS
Perfluoro-1 -butanesulfonate
Perfluorobutane Sulfonate
Perfluorobutanesulfonate
Perfluorobutanesulfonic acid
Perfluorobutylsulfonate
45187-15-3
Note: MESH = Medical subject headings
A.l. Literature Search Strings
PubMed
375-73-5[rn] OR 45187-15-3[rn] "nonafluorobutane-l-sulfonic acid"[nm] OR
"1,1,2,2,3,3,4,4,4-Nonafluoro-l-butanesulfonic acid"[tw] OR "1-Perfluorobutanesulfonic
A-l

-------
acid"[tw] OR "Nonafluoro-l-butanesulfonic acid"[tw] OR "Nonafluorobutanesulfonic acid"[tw]
OR "Perfluorobutanesulfonic acid"[tw] OR "1,1,2,2,3,3,4,4,4-Nonafluorobutane-l-sulphonic
acid"[tw] OR "Perfluorobutane sulfonic acid"[tw] OR "Perfluorobutanesulfonate"[tw] OR
"Perfluorobutane sulfonate"[tw] OR "1-Butanesulfonic acid, l,l,2,2,3,3,4,4,4-nonafluoro-"[tw]
OR "1-Butanesulfonic acid, nonafluoro-"[tw] OR "Perfluoro-l-butanesulfonate"[tw] OR
"Perfluorobutylsulfonate"[tw] OR "Eftop FBSA"[tw] OR (PFBS[tw] AND (fluorocarbon*[tw]
OR fluorotelomer*[tw] OR polyfluoro* [tw] OR perfluoro-* [tw] OR perfluoroa*[tw] OR
perfluorob*[tw] OR perfluoroc*[tw] OR perfluorod*[tw] OR perfluoroe*[tw] OR
perfluoroh*[tw] OR perfluoron*[tw] OR perfluoroo*[tw] OR perfluorop*[tw] OR
perfluoros*[tw] OR perfluorou*[tw] OR perfluorinated[tw] OR fluorinated[tw] ORPFAS[tw]
OR PFOS[tw] OR PFOA[tw]))
wos
TS="l,l,2,2,3,3,4,4,4-Nonafluoro-l-butanesulfonic acid" OR TS="1-Perfluorobutanesulfonic
acid" OR TS="Nonafluoro-l-butanesulfonic acid" OR TS="Nonafluorobutanesulfonic acid" OR
TS="Perfluorobutanesulfonic acid" OR TS="l,l,2,2,3,3,4,4,4-Nonafluorobutane-l-sulphonic
acid" OR TS="Perfluorobutane sulfonic acid" OR TS="Perfluorobutanesulfonate" OR
TS="Perfluorobutane sulfonate" OR TS=" 1-Butanesulfonic acid, 1,1,2,2,3,3,4,4,4-nonafluoro-"
OR TS=" 1-Butanesulfonic acid, nonafluoro-" OR TS="Perfluoro-l-butanesulfonate" OR
TS="Perfluorobutylsulfonate" OR TS="Eftop FBSA" OR (TS=PFBS AND TS=(fluorocarbon*
OR fluorotelomer* OR polyfluoro* OR perfluoro-* OR perfluoroa* OR perfluorob* OR
perfluoroc* OR perfluorod* OR perfluoroe* OR perfluoroh* OR perfluoron* OR perfluoroo*
OR perfluorop* OR perfluoros* OR perfluorou* OR perfluorinated OR fluorinated OR PFAS
OR PFOS OR PFOA))
Toxline
( ( 375-73-5 [rn] OR 45187-15-3 [rn] OR "1 1223344 4-nonafluoro-l-butanesulfonic acid"
OR "1-perfluorobutanesulfonic acid" OR "nonafluoro-l-butanesulfonic acid" OR
"nonafluorobutanesulfonic acid" OR "perfluorobutanesulfonic acid" OR "1 1223344
4-nonafluorobutane-l-sulphonic acid" OR "perfluorobutane sulfonic acid" OR
"perfluorobutanesulfonate" OR "perfluorobutane sulfonate" OR "1-butanesulfonic acid 112 2 3
3 4 4 4-nonafluoro-" OR "1-butanesulfonic acid nonafluoro-" OR "perfluoro-1-butanesulfonate"
OR "perfluorobutylsulfonate" OR "eftop fbsa" OR (pfbs AND (fluorocarbon* OR
fluorotelomer* OR polyfluoro* OR perfluoro* OR perfluorinated OR fluorinated OR pfas OR
pfos OR pfoa ) ) ) ) AND ( ANEUPL [org] OR BIOSIS [org] OR CIS [org] OR DART [org] OR
EMIC [org] OR EPIDEM [org] OR HEEP [org] OR HMTC [org] OR IPA [org] OR RISKLINE
[org] ORMTGABS [org] ORNIOSH [org] ORNTIS [org] ORPESTAB [org] ORPPBIB [org]
) AND NOT PubMed [org] AND NOT pubdart [org]
TSCATS
375-73-5[rn] AND tscsats[org]
45187-15-3 [rn] AND tscsats[org]
A-2

-------
Appendix B: Detailed PECO Criteria
Table B-l. Population, exposure, comparator, and outcome criteria
PECO
element
Evidence
Population
Human: Any population (occupational; general population including children, pregnant women, and
other sensitive populations). The following study designs will be considered most informative:
controlled exposure, cohort, case-control, or cross-sectional. Note: Case reports and case series are
not the primary focus of this assessment and will be tracked as supplemental material during the study
screening process.
Animal: Nonhuman mammalian animal species (whole organism) of any life stage (including
preconception, in utero, lactation, peripubertal, and adult stages).
In vitro models of genotoxicity: The studies will be considered PECO-relevant. All other in vitro
studies will be tagged as "not-PECO relevant, but supplemental material."
Nonmammalian model systems/;'/? vitro/in silico NOT related to genotoxicity: Nonmammalian
model systems (e.g., fish, amphibians, birds, and C. elegans); studies of human or animal cells,
tissues, or biochemical reactions (e.g., ligand binding assays) with in vitro exposure regimens;
bioinformatics pathways of disease analysis; and/or high throughput screening data. These studies
will be classified as non-PECO-relevant, but have supplemental information.
Exposure
Human: Studies providing qualitative or quantitative estimates of 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, job title
or other relevant occupational information. Human "mixture" studies are considered PECO-relevant
as long as they have the per- and polyfluoroalkyl substances (PFAS) of interest.
Animal: Studies providing qualitative and quantitative estimates of exposure based on administered
dose or concentration. Oral and inhalation studies are considered PECO-relevant. Nonoral and
noninhalation studies are tagged as supplemental. Experimental mixture studies are included as
PECO-relevant only if they include a perfluorobutane sulfonic acid- (PFBS-) only arm. Otherwise,
mixture studies are tagged as supplemental.
All studies must include exposure to PFBS, CASRN 375-73-5. Studies of precursor PFAS that
identify any of the targeted PFAS as metabolites will also be included.
Comparator
Human: A comparison or reference population exposed to lower levels (or no exposure/exposure
below detection levels) or for shorter periods of time. For D-R purposes, exposure-response
quantitative results must be presented in sufficient detail such as regression coefficients presented
with statistical measure of variation such as RR, HR, OR, or SMR or observed cases vs. expected
cases (common in occupational studies); slope or linear regression coefficient (i.e., per unit increase
in a continuous outcome); difference in the means; or report means with results of t-test, mean
comparison by regression, or other mean-comparing hypothesis test.
Animal: Quantitative exposure versus lower or no exposure with concurrent vehicle control group.
Outcome
Cancer and noncancer health outcomes. In general, endpoints related to clinical diagnostic criteria,
disease outcomes, histopathological examination, genotoxicity, or other apical/phenotypic outcomes
will be prioritized for evidence synthesis. Based on preliminary screening work and other
assessments, the systematic review is anticipated to focus on liver (including serum lipids),
developmental, reproductive, neurological, developmental neurotoxicity, thyroid disease/disruption,
immunological, cardiovascular, and musculoskeletal outcomes.
Notes'. D-R = Dose-Response; HR = hazard ratio; OR = odds ratio; PECO = population, exposure, comparator, and outcome;
RR = risk ratio; SMR = standardized mortality ratio
A-l

-------
Appendix C: Study Evaluation Methods
For each outcome in a study, in each domain, reviewers reached a consensus judgment of good,
adequate, deficient, not reported, or critically deficient. Questions used to guide the development
of criteria for each domain in epidemiology studies are presented in Table C-l and experimental
animal toxicology studies in Table C-3. These categories were applied to each evaluation domain
for each study as follows:
• Good represents a judgment that the study was conducted appropriately in relation to the
evaluation domain and any deficiencies, if present, are minor and would not be expected
to influence the study results.
• Adequate indicates a judgment that there are methodological limitations relating to the
evaluation domain, but that those limitations are not likely to be severe or to have a
notable impact on the results.
• Deficient denotes identified biases or deficiencies that are interpreted as likely to have
had a notable impact on the results or that prevent interpretation of the study findings.
• Not reported indicates that the information necessary to evaluate the domain was not
available in the study. Generally, this term carries the same functional interpretation as
deficient for the purposes of the study confidence classification. Depending on the
number and severity of other limitations identified in the study, it may or may not be
worth reaching out to the study authors for this information.
• Critically deficient reflects a judgment that the study conduct introduced a serious flaw
that makes the observed effect(s) uninterpretable. Studies with a determination of
critically deficient in an evaluation domain will almost always cause the study to be
considered overall "uninformative".
Once the evaluation domains were rated, the identified strengths and limitations were considered
to reach a study confidence rating of high, medium, low, or uninformative for a specific health
outcome. This was based on the reviewer judgments across the evaluation domains and included
consideration of the likely impact the noted deficiencies in bias and sensitivity, or inadequate
reporting, have on the results. The ratings, which reflect a consensus judgment between
reviewers, are defined as follows:
• High. A well-conducted study with no notable deficiencies or concerns were identified;
the potential for bias is unlikely or minimal, and the study used sensitive methodology.
High confidence studies generally reflect judgments of good across all or most evaluation
domains.
• Medium: A satisfactory (acceptable) study in which deficiencies or concerns were noted,
but the limitations are unlikely to be of a notable degree. Generally, medium confidence
studies will include adequate or good judgments across most domains, with the impact of
any identified limitation not being judged as severe.
• Low. A substandard study in which 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. Typically, low confidence studies would have a deficient evaluation
for one or more domains, although some medium confidence studies could have a
deficient rating in domain(s) considered to have less influence on the magnitude or
direction of effect estimates. Generally, low confidence results are given less weight than
A-l

-------
high or medium confidence results during evidence synthesis and integration and are
generally not used as the primary sources of information for hazard identification or
derivation of toxicity values unless they are the only studies available. Studies rated as
low confidence only because of sensitivity concerns about bias towards the null require
additional consideration during evidence synthesis. Observing an effect in these studies
could increase confidence, assuming the study was otherwise well-conducted.
• Uninformative: An unacceptable study in which serious flaw(s) make the study results
unusable for informing hazard identification. Studies with critically deficient j udgments
in any evaluation domain will almost always be classified as uninformative (see
explanation above). Studies with multiple deficient judgments across domains might also
be considered uninformative. Uninformative studies will not be considered further in the
synthesis and integration of evidence for hazard identification or dose response but might
be used to highlight possible research gaps.
Table C-l. Questions used to guide the development of criteria for each domain in
epidemiology studies
Core question
Prompting questions
Follow-up questions
ExDOSure
For all:
•	Does the exposure measure capture the variability in exposure
among the participants, considering intensity, frequency, and
duration of exposure?
•	Does the exposure measure reflect a relevant time window? If
not, can the relationship between measures in this time and the
relevant time window be estimated reliably?
•	Was the exposure measurement likely to be affected by a
knowledge of the outcome?
•	Was the exposure measurement likely to be affected by the
presence of the outcome (i.e., reverse causality)?
For case-control studies of occupational exposures:
•	Is exposure based on a comprehensive job history describing
tasks, setting, time period, and use of specific materials?
For biomarkers of exposure, general population:
•	Is a standard assay used? What are the intra- and inter-assay
coefficients of variation? Is the assay likely to be affected by
contamination? Are values less than the limit of detection
dealt with adequately?
What exposure time period is reflected by the biomarker? If the
half-life is short, what is the correlation between serial measurements
of exposure?
Is the degree of
exposure
misclassification likely
to vary by exposure
level?
If the correlation
between exposure
measurements is
moderate, is there an
adequate statistical
approach to ameliorate
variability in
measurements?
If there is a concern
about the potential for
bias, what is the
predicted direction or
distortion of the bias on
the effect estimate (if
there is enough
information)?
measurement
Does the
exposure
measure reliably
distinguish
between levels
of exposure in a
time window
considered most
relevant for a
causal effect
with respect to
the development
of the outcome?
A-2

-------
Core question
Prompting questions
Follow-up questions
Outcome
For all:
•	Is outcome ascertainment likely to be affected by knowledge
of, or presence of, exposure (e.g., consider access to health
care, if based on self-reported history of diagnosis)?
For case-control studies:
•	Is the comparison group without the outcome (e.g., controls in
a case-control study) based on objective criteria with little or
no likelihood of inclusion of people with the disease?
For mortality measures:
•	How well does cause of death data reflect occurrence of the
disease in an individual? How well do mortality data reflect
incidence of the disease?
For diagnosis of disease measures:
•	Is diagnosis based on standard clinical criteria? If based on
self-report of diagnosis, what is the validity of this measure?
For laboratory-based measures (e.g., hormone levels):
•	Is a standard assay used? Does the assay have an acceptable
level of inter-assay variability? Is the sensitivity of the assay
appropriate for the outcome measure in this study population?
Is there a concern that
any outcome
misclassification is
nondifferential,
differential, or both?
What is the predicted
direction or distortion of
the bias on the effect
estimate (if there is
enough information)?
ascertainment
Does the
outcome
measure reliably
distinguish the
presence or
absence (or
degree of
severity) of the
outcome?
ParticiDant
For longitudinal cohort:
•	Did participants volunteer for the cohort based on knowledge
of exposure and/or preclinical disease symptoms? Was entry
into the cohort or continuation in the cohort related to
exposure and outcome?
For occupational cohort:
•	Did entry into the cohort begin with the start of the exposure?
•	Was follow-up or outcome assessment incomplete, and if so,
was follow-up related to both exposure and outcome status?
•	Could exposure produce symptoms that would result in a
change in work assignment/work status ("healthy worker
survivor effect")?
For case-control study:
•	Were controls representative of population and time periods
from which cases were drawn?
•	Are hospital controls selected from a group whose reason for
admission is independent of exposure?
•	Could recruitment strategies, eligibility criteria, or
participation rates result in differential participation relating to
both disease and exposure?
For population-based survey:
•	Was recruitment based on advertisement to people with
knowledge of exposure, outcome, and hypothesis?
Were differences in
participant enrollment
and follow-up evaluated
to assess bias?
If there is a concern
about the potential for
bias, what is the
predicted direction or
distortion of the bias on
the effect estimate (if
there is enough
information)?
Were appropriate
analyses performed to
address changing
exposures over time in
relation to symptoms?
Is there a comparison of
participants and
nonparticipants to
address whether
differential selection is
likely?
selection
Is there
evidence that
selection into or
out of the study
(or analysis
sample) was
jointly related to
exposure and to
outcome?
A-3

-------
Core question
Prompting questions
Follow-up questions
Confounding
Is confounding adequately addressed by considerations in...
a.	... participant selection (matching or restriction)?
b.	... accurate information on potential confounders, and
statistical adjustment procedures?
c.	... lack of association between confounder and outcome, or
confounder and exposure in the study?
d.	... information from other sources?
Is the assessment of confounders based on a thoughtful review of
published literature, potential relationships (e.g., as can be gained
through directed acyclic graphing), minimizing potential overcontrol
(e.g., inclusion of a variable on the pathway between exposure and
outcome)?
If there is a concern
about the potential for
bias, what is the
predicted direction or
distortion of the bias on
the effect estimate (if
there is enough
information)?
Is confounding
of the effect of
the exposure
likely?
Analysis
•	Are missing outcome, exposure, and covariate data recognized
and, if necessary, accounted for in the analysis?
•	Does the analysis appropriately consider variable distributions
and modeling assumptions?
•	Does the analysis appropriately consider subgroups of interest
(e.g., based on variability in exposure level or duration, or
susceptibility)?
•	Is an appropriate analysis used for the study design?
•	Is effect modification considered, based on considerations
developed a priori?
•	Does the study include additional analyses addressing
potential biases or limitations (i.e., sensitivity analyses)?
If there is a concern
about the potential for
bias, what is the
predicted direction or
distortion of the bias on
the effect estimate (if
there is enough
information)?
Do the analysis
strategy and
presentation
convey the
necessary
familiarity with
the data and
assumptions?
Sensitivity
•	Is the exposure range adequate?
•	Was the appropriate population included?
•	Was the length of follow-up adequate? Is the time/age of
outcome ascertainment optimal given the interval of exposure
and the health outcome?
•	Are there other aspects related to risk of bias or otherwise that
raise concerns about sensitivity?

Is there a
concern that
sensitivity of the
study is not
adequate to
detect an effect?
Selective
•	Are the results needed for the IRIS analysis presented (based
on a priori specification)? If not, can these results be obtained?
•	Are only statistically significant results presented?

reDorting
Is there reason
to be concerned
about selective
reporting?
Note: IRIS = Integrated Risk Information System
C.l. Exposure measurement evaluation criteria
The criteria used to evaluate exposure measurement for PFBS (Table C-2) are adapted from the
criteria developed by the National Toxicology Program (NTP) Office of Health Assessment and
Translation for their assessment of the association between perfluorooctane sulfonic acid (PFOS)
and perfluorooctanoic acid (PFOA) and immune effects (NTP. 2016. 2015) and were established
prior to beginning study evaluation. Standard analytical methods for evaluating individual
per- and polyfluoroalkyl substances (PFAS) in serum or whole-blood using quantitative
techniques such as liquid chromatography-triple quadrupole mass spectrometry are preferred
A-4

-------
(CDC. 2018; U.S. EPA. 2014b. e; AT SDR. 2009; CDC. 2009). The estimated serum half-life of
PFBS is approximately 1 month (Lau. 2015; 01 sen et al.. 2009). so unlike for some other PFAS
with longer half-lives, current exposure might not be indicative of past exposures. Little data is
available on repeated measures of PFBS in humans over time, so the reliability of a single
measure is unclear. The timing of the exposure measurement is considered in relation to the
etiologic window for each outcome being reviewed.
Table C-2. Criteria for evaluation of exposure measurement in epidemiology studies
Exposure
measurement
rating
Criteria
Good
All of the following:
• Evidence that exposure was consistently assessed using well-established methods that
directly measure exposure (e.g., measurement of PFAS in blood, serum, or plasma).
• Exposure was assessed in a relevant time window for development of the outcome (i.e.,
temporality is established and sufficient latency occurred prior to disease onset).
• There is evidence that a sufficient proportion of the exposure data measurements are
above the limit of quantification for the assay so that different exposure groups can be
distinguished based on the analyses conducted.
• The laboratory analysis included standard quality control measures with demonstrated
precision and accuracy.
• There is sufficient specificity /sensitivity and range or variation in exposure
measurements that would minimize potential for exposure measurement error and
misclassification by allowing exposure classifications to be differentiated (i.e., can
reliably categorize participants into groups such as high vs. low exposure).
Adequate
• Evidence that exposure was consistently assessed using well-established methods that
directly measure exposure (e.g., measurement of PFAS in blood, serum, or plasma),
but there were some minor concerns about quality control measures or other potential
for nondifferential misclassification.
OR
• Exposure was assessed using indirect measures (e.g., drinking water concentrations and
residential location/history, questionnaire, or occupational exposure assessment by a
certified industrial hygienist) that have been validated or empirically shown to be
consistent with methods that directly measure exposure (i.e., inter-methods validation:
one method vs. another) Note: This could be good if the validation was sufficient. All
studies for PFBS used direct measures.
And all of the following:
• Exposure was assessed in a relevant time window for development of the outcome.
• There is evidence that a sufficient proportion of the exposure data measurements are
above the limit of quantification for the assay.
• There is sufficient specificity /sensitivity and range or variation in exposure
measurements that would minimize potential for exposure measurement error and
misclassification by allowing exposure classifications to be differentiated (i.e., can
reliably categorize participants into groups such as high vs. low exposure), but there
might be more uncertainty than in good.
A-5

-------
Exposure
measurement
rating
Criteria
Deficient
Any of the following:
• Some concern, but no direct evidence, that the exposure was assessed using poorly
validated methods.
• There is insufficient information provided about the exposure assessment, including
precision, accuracy, and level of quantification, but no evidence for concern about the
method used.
• Exposure was assessed in a relevant time window for development of the outcome.
There could be concerns about reverse causation between exposure and outcome, but
there is no direct evidence that it is present.
• There is some concern over insufficient specificity/sensitivity and range or variation in
exposure measurements that may result in considerable exposure measurement error
and misclassification when exposure classifications are compared (i.e., data do not lend
themselves to reliably categorize participants into groups such as high vs. low
exposure, and/or there is considerable uncertainty in exposure values that do not allow
for confidence in the examination of small per unit changes in continuous exposures).
Critically
deficient
Any of the following:
• Exposure was assessed in a time window that is unknown or not relevant for
development of the outcome. This could be due to clear evidence of reverse causation
between exposure and outcome, or other concerns such as the lack of temporal
ordering of exposure and disease onset, insufficient latency, or having exposure
measurements that are not reliable measures of exposure during the etiologic window.
• Direct evidence that bias was likely, since the exposure was assessed using methods
with poor validity.
• Evidence of differential exposure misclassification (e.g., differential recall of self-
reported exposure).
• There is evidence that an insufficient proportion of the exposure data measurements are
above the limit of quantification for the assay.
A-6

-------
Table C-3. Questions used to guide the development of criteria for each domain in experimental animal toxicology studies
Evaluation
type
Domain-
core question
Prompting questions
Basic considerations
Reporting Quality
Reporting Quality -
Does the study report
information for evaluating the
design and conduct of the
study for the
endpoint(s)/outcome(s) of
interest?
Notes:
Reviewers should reach out to
authors to obtain missing
information when studies are
considered key for hazard
evaluation and/or dose-
response.
This domain is limited to
reporting. Other aspects of the
exposure methods,
experimental design, and
endpoint evaluation methods
are evaluated using the
domains related to risk of bias
and study sensitivity.
Does the study report the following?
• Critical information necessary to
perform study evaluation:
o Species; test article name; levels and
duration of exposure; route (e.g., oral;
inhalation); qualitative or quantitative
results for at least one endpoint of interest.
• Imnortant information for evaluating
the study methods:
o Test animal: strain, sex, source, and
general husbandry procedures,
o Exposure methods: source, purity, method
of administration,
o Experimental design: frequency of
exposure, animal age and life stage during
exposure and at endpoint/outcome
evaluation,
o Endpoint evaluation methods: assays or
procedures used to measure the
endpoints/outcomes of interest.
These considerations typically do not need to be refined by
assessment teams, although in some instances the
imnortant information mav be refined depending on the
endpoints/outcomes of interest or the chemical under
investigation.
A judgment and rationale for this domain should be given
for the study. Typically, these will not change regardless of
the endpoints/outcomes investigated by the study. In the
rationale, reviewers should indicate whether the study
adhered to GLP, OECD, or other testing guidelines.
• Good: All critical and imnortant information is
reported or inferable for the endpoints/outcomes of
interest.
• Adeauate: All critical information is rcDortcd but
some imnortant information is missine.
However, the missing information is not expected
to significantly impact the study evaluation.
• Deficient. All critical information is rcDortcd but
imnortant information is missine that is expected
to significantly reduce the ability to evaluate the
study.
• Critically Deficient. Study report is missing any
nieces of critical information. Studies that are
Critically Deficient for reporting are Uninformative
for the overall rating and considered no further for
evidence synthesis and integration.
A-7

-------
Evaluation
type
Domain-
core question
Prompting questions
Basic considerations
Risk of Bias
Selection and performance bias
Allocation -
Were animals assigned to
experimental groups using a
method that minimizes
selection bias?
For each study:
•	Did each animal or litter have an equal
chance of being assigned to any
experimental group (i.e., random
allocation)?
•	Is the allocation method described?
•	Aside from randomization, were any
steps taken to balance variables across
experimental groups during allocation?
These considerations typically do not need to be refined by
assessment teams.
A judgment and rationale for this domain should be given
for each cohort or experiment in the study.
•	Good: Experimental groups were randomized and
any specific randomization procedure was
described or inferable (e.g., computer-generated
scheme). [Note that normalization is not the same
as randomization (see response for 'Adequate') ]
•	Adequate: Authors report that groups were
randomized but do not describe the specific
procedure used (e.g., "animals were randomized").
Alternatively, authors used a nonrandom method to
control for important modifying factors across
experimental groups (e.g., body weight
normalization).
•	Not Reported (interpreted as Deficient): No
indication of randomization of groups or other
methods (e.g., normalization) to control for
important modifying factors across experimental
groups.
•	Critically Deficient. Bias in the animal allocations
was reported or inferable.
Observational
Bias/Blin ding-
Did the study implement
measures to reduce
observational bias?
For each endpoint/outcome or grouping of
endpoints/outcomes in a study:
•	Does the study report blinding or other
methods/procedures for reducing
observational bias?
•	If not, did the study use a design or
approach for which such procedures can
be inferred?
•	What is the expected impact of failure to
implement (or report implementation) of
these methods/procedures on results?
These considerations typically do not need to be refined by
the assessment teams. [Note that it can be useful for teams
to identify highly subjective measures of
endpoints/outcomes where observational bias may strongly
influence results prior to performing evaluations.]
A judgment and rationale for this domain should be given
for each endpoint/outcome or group of endpoints/outcomes
investigated in the study.
• Good: Measures to reduce observational bias were
described (e.g., blinding to conceal treatment
groups during endpoint evaluation; consensus-
based evaluations of histopathology lesions).3
A-8

-------
Evaluation
type
Domain-
core question
Prompting questions
Basic considerations




•	Adequate: Methods for reducing observational bias
(e.g., blinding) can be inferred or were reported but
described incompletely.
•	Not Reported. Measures to reduce observational
bias were not described.
o Interpreted as Adequate—The potential concern for
bias was mitigated based on use of
automated/computer driven systems, standard
laboratory kits, relatively simple, objective measures
(e.g., body or tissue weight), or screening-level
evaluations of histopathology.
o Interpreted as Deficient—The potential impact on the
results is major (e.g., outcome measures are highly
subjective).
•	Critically Deficient. Strong evidence for
observational bias that could have impacted results.
Confounding/
variable control
Confounding-
Are variables with the
potential to confound or
modify results controlled for
and consistent across all
experimental groups?
For each study:
•	Are there differences across the
treatment groups (e.g., co-exposures,
vehicle, diet, palatability, husbandry,
health status, and so forth) that could
bias the results?
•	If differences are identified, to what
extent are they expected to impact the
results?
These considerations may need to be refined by assessment
teams, as the specific variables of concern can vary by
experiment or chemical.
A judgment and rationale for this domain should be given
for each cohort or experiment in the study, noting when the
potential for confounding is restricted to specific
endpoints/outcomes.
•	Good: Outside of the exposure of interest, variables
that are likely to confound or modify results appear
to be controlled for and consistent across
experimental groups.
•	Adequate: Some concern that variables that were
likely to confound or modify results were
uncontrolled or inconsistent across groups, but are
expected to have a minimal impact on the results.
•	Deficient. Notable concern that potentially
confounding variables were uncontrolled or
A-9

-------
Evaluation
type
Domain-
core question
Prompting questions
Basic considerations




•
inconsistent across groups and are expected to
substantially impact the results.
Critically Deficient. Confounding variables were
presumed to be uncontrolled or inconsistent across
groups and are expected to be a primary driver of
the results.

Reporting and attrition bias
Selective Reporting and
Attrition-
Did the study report results for
all prespecified outcomes and
tested animals?
Note:
This domain does not consider
the appropriateness of the
analysis/results presentation.
This aspect of study quality is
evaluated in another domain.
For each study:
Selective reporting bias:
•	Are all results presented for
endpoints/outcomes described in the
methods (see note)?
Attrition bias:
•	Are all animals accounted for in the
results?
•	If there are discrepancies, do authors
provide an explanation (e.g., death or
unscheduled sacrifice during the study)?
•	If unexplained results, omissions, and/or
attrition are identified, what is the
expected impact on the interpretation of
the results?
These considerations typically do not need to be refined by
assessment teams.
A judgment and rationale for this domain should be given
for each cohort or experiment in the study.
•	Good: Quantitative or qualitative results were
reported for all prespecified outcomes (explicitly
stated or inferred), exposure groups and evaluation
timepoints. Data not reported in the primary article
is available from supplemental material. If results,
omissions, or animal attrition is identified, the
authors provide an explanation and these are not
expected to impact the interpretation of the results.
•	Adequate: Quantitative or qualitative results are
reported for most prespecified outcomes (explicitly
stated or inferred), exposure groups and evaluation
timepoints. Omissions and/or attrition are not
explained, but are not expected to significantly
impact the interpretation of the results.
•	Deficient. Quantitative or qualitative results are
missing for many prespecified outcomes (explicitly
stated or inferred), exposure groups and evaluation
timepoints and/or high animal attrition; omissions
and/or attrition are not explained and may
significantly impact the interpretation of the
results.
•	Critically Deficient. Extensive results omission
and/or animal attrition is identified and prevents
comparisons of results across treatment groups.
A-10

-------
Evaluation
type
Domain-
core question
Prompting questions
Basic considerations
s
QJ
J1
"5
c
zj
o
-C
K
£
o
s.
M
W
Chemical Administration
and Characterization-
Did the study adequately
characterize exposure to the
chemical of interest and the
exposure administration
methods?
Note:
Consideration of the
appropriateness of the route of
exposure is not evaluated at
the individual study level.
Relevance and utility of the
routes of exposure are
considered in the PECO
criteria for study inclusion and
during evidence synthesis.
For each study:
Does the study report the source and
purity and/or composition (e.g., identity
and percent distribution of different
isomers) of the chemical? If not, can the
purity and/or composition be obtained
from the supplier (e.g., as reported on
the website)?
Was independent analytical verification
of the test article purity and composition
performed?
Did the authors take steps to ensure the
reported exposure levels were accurate?
For inhalation studies: Were target
concentrations confirmed using reliable
analytical measurements in chamber air?
For oral studies: If necessary based on
consideration of chemical-specific
knowledge (e.g., instability in solution;
volatility) and/or exposure design (e.g., the
frequency and duration of exposure), were
chemical concentrations in the dosing
solutions or diet analytically confirmed?
Are there concerns about the methods
used to administer the chemical (e.g.,
inhalation chamber type, gavage volume,
etc.)?
It is essential that these criteria are considered and
potentially refined by assessment teams, as the specific
variables of concern can vary by chemical.
A judgment and rationale for this domain should be given
for each cohort or experiment in the study.
•	Good: Chemical administration and
characterization is complete (i.e., source, purity,
and analytical verification of the test article are
provided). There are no concerns about the
composition, stability, or purity of the administered
chemical or the specific methods of administration.
For inhalation studies, chemical concentrations in
the exposure chambers are verified using reliable
analytical methods.
•	Adequate: Some uncertainties in the chemical
administration and characterization are identified
but these are expected to have minimal impact on
interpretation of the results (e.g., source and
vendor- reported purity are presented, but not
independently verified; purity of the test article is
suboptimal but not concerning). For inhalation
studies, actual exposure concentrations are missing
or verified with less reliable methods.
•	Deficient Uncertainties in the exposure
characterization are identified and expected to
substantially impact the results (e.g., source of the
test article is not reported; levels of impurities are
substantial or concerning; deficient administration
methods such as use of static inhalation chambers
or a gavage volume considered too large for the
species and/or life stage at exposure).
•	Critically Deficient. Uncertainties in the exposure
characterization are identified, and there is
reasonable certainty that the results are largely
attributable to factors other than exposure to the
A-ll

-------
Evaluation
type
Domain-
core question
Prompting questions
Basic considerations




chemical of interest (e.g., identified impurities are
expected to be a primary driver of the results).


Exposure Timing,
Frequency and Duration-
Was the timing, frequency,
and duration of exposure
sensitive for the
endpoint(s)/outcome(s) of
interest?
For each endpoint/outcome or grouping of
endpoints/outcomes in a study:
•	Does the exposure period include the
critical window of sensitivity?
•	Was the duration and frequency of
exposure sensitive for detecting the
endpoint of interest?
Considerations for this domain are highly variable
depending on the endpoint(s)/outcome(s) of interest and
must be refined by assessment teams.
A judgment and rationale for this domain should be given
for each endpoint/outcome or group of endpoints/outcomes
investigated in the study.
•	Good: The duration and frequency of the exposure
was sensitive and the exposure included the critical
window of sensitivity (if known).
•	Adequate: The duration and frequency of the
exposure was sensitive and the exposure covered
most of the critical window of sensitivity (if
known).
•	Deficient. The duration and/or frequency of the
exposure is not sensitive and did not include the
majority of the critical window of sensitivity (if
known). These limitations are expected to bias the
results towards the null.
•	Critically Deficient. The exposure design was not
sensitive and is expected to strongly bias the results
towards the null. The rationale should indicate the
specific concern(s).

Outcome measures and
results display
Endpoint Sensitivity and
Specificity-
Are the procedures sensitive
and specific for evaluating the
endpoint(s)/outcome(s) of
interest?
Note:
Sample size alone is not a
reason to conclude an
For each endpoint/outcome or grouping of
endpoints/outcomes in a study:
•	Are there concerns regarding the
specificity and validity of the protocols?
•	Are there serious concerns regarding the
sample size (see note)?
•	Are there concerns regarding the timing
of the endpoint assessment?
Considerations for this domain are highly variable
depending on the endpoint(s)/outcome(s) of interest and
must be refined by assessment teams.
A judgment and rationale for this domain should be given
for each endpoint/outcome or group of endpoints/outcomes
investigated in the study.
Examples of potential concerns include:
•	Selection of protocols that are insensitive or
nonspecific for the endpoint of interest.
•	Use of unreliable methods to assess the outcome.
A-12

-------
Evaluation
type
Domain-
core question
Prompting questions
Basic considerations


individual study is critically
deficient.

•	Assessment of endpoints at inappropriate or
insensitive ages, or without addressing known
endpoint variation (e.g., due to circadian rhythms,
estrous cyclicity, etc.).
*	Decreased specificity or sensitivity of the response
due to the timing of endpoint evaluation, as
compared to exposure (e.g., short-acting depressant
or irritant effects of chemicals; insensitivity due to
prolonged period of nonexposure prior to testing).
Results Presentation-
Are the results presented in a
way that makes the data usable
and transparent?
For each endpoint/outcome or grouping of
endpoints/outcomes in a study:
•	Does the level of detail allow for an
informed interpretation of the results?
•	Are the data analyzed, compared, or
presented in a way that is inappropriate
or misleading?
Considerations for this domain are highly variable
depending on the outcomes of interest and must be refined
by assessment teams.
A judgment and rationale for this domain should be given
for each endpoint/outcome or group of endpoints/outcomes
investigated in the study.
Examples of potential concerns include:
•	Nonpreferred presentation such as developmental
toxicity data averaged across pups in a treatment
group when litter responses are more appropriate.
•	Failing to present quantitative results.
•	Pooling data when responses are known or
expected to differ substantially (e.g., across sexes
or ages).
•	Failing to report on or address overt toxicity when
exposure levels are known or expected to be highly
toxic.
•	Lack of full presentation of the data (e.g.,
presentation of mean without variance data;
concurrent control data are not presented).
A-13

-------
Evaluation
type
Domain-
core question
Prompting questions
Basic considerations
Overall Confidence
Overall Confidence-
Considering the identified
strengths and limitations, what
is the overall confidence rating
for the endpoint(s)/outcome(s)
of interest?
Note:
Reviewers should mark studies
that are rated lower than high
confidence only due to low
sensitivity (i.e., bias towards
the null) for additional
consideration during evidence
synthesis. If the study is
otherwise well-conducted and
an effect is obser\>ed, the
confidence may be increased.
For each endpoint/outcome or grouping of
endpoints/outcomes in a study:
• Were concerns (i.e., limitations or
uncertainties) related to the reporting
quality, risk of bias, or sensitivity
identified?
• If yes, what is their expected impact on
the overall interpretation of the
reliability and validity of the study
results, including (when possible)
interpretations of impacts on the
magnitude or direction of the reported
effects?
The overall confidence rating considers the likely impact of
the noted concerns (i.e., limitations or uncertainties) in
reporting, bias, and sensitivity on the results.
A confidence rating and rationale should be given for each
endpoint/outcome or group of endpoints/outcomes
investigated in the study.
• High Confidence: No notable concerns are
identified (e.g., most or all domains rated Good).
• Medium Confidence: Some concerns are identified,
but expected to have minimal impact on the
interpretation of the results (e.g., most domains
rated Adequate or Good; may include studies with
Deficient ratings if concerns are not expected to
strongly impact the magnitude or direction of the
results). Any important concerns should be carried
forward to evidence synthesis.
• Low Confidence: Identified concerns are expected
to significantly impact on the study results or their
interpretation (e.g., generally. Deficient ratings for
one or more domains). The concerns leading to this
confidence judgment must be carried forward to
evidence synthesis (see note).
• Uninformative: Serious flaw(s) that make the study
results unusable for informing hazard identification
(e.g., generally. Critically Deficient rating in any
domain; many Deficient ratings). Uninformative
studies are considered no further in the synthesis
and integration of evidence.
Notes: GLP =good laboratory practices; OECD = Organisation for Economic Cooperation and Development.
a For nontargeted or screening-level histopathology outcomes often used in guideline studies, blinding during the initial evaluation of tissues is generally not recommended as
masked evaluation can make "the task of separating treatment-related changes from normal variation more difficult" and "there is concern that masked review during the initial
evaluation may result in missing subtle lesions." Generally, blinded evaluations are recommended for targeted secondary review of specific tissues or in instances when there is a
predefined set of outcomes that is known or predicted to occur ("Crissman et al.. 2004).
A-14

-------
Appendix D: HAWC User Guide and Frequently Asked Questions
D.l. What is HAWC and What is its Purpose?
HAWC (Health Assessment Workspace Collaborative) is an interactive expert-driven content
management system for human health assessments that is intended to promote transparency,
trackability, data usability, and understanding of the data and decisions supporting an
environmental and human health assessment. Specifically, HAWC is an interface that allows the
data and decisions supporting an assessment to be managed in modules (e.g., study evaluation,
summary study data, etc.) that can be publicly accessed on-line (see #2 below and Figure D-l).
Following literature search and screening that are conducted using HERO and DistillerSR.
HAWC manages each study included in an assessment and makes the extracted information
available via a web link that takes a user to a web page displaying study specific details and data
(e.g., study evaluation, experimental design, dosing regime, endpoints evaluated, dose response
data, etc., described in further detail below in #s 3-6). Finally, all data managed in HAWC is
fully downloadable using the blue "Download datasets" link (highlighted in the red box below)
also located in the grey navigation bar located on the assessment home page (discussed in # 1
below). Note that a user may quickly navigate HAWC by clicking on the file path (highlighted in
orange dashed box below) given in the grey row below the HAWC icon and Login bar (Figure
D-l). HAWC aims to facilitate team collaboration by scientists who develop these assessments
and enhance transparency of the process by providing online access (no user account required) to
the data and expert decisions used to evaluate potential human health hazard and risk of chemical
exposures.
<- C 0 • Secure https//tiawcpfd.epa.gov/assessment/100000037/ 9 Q :
HAWtr Contact About Public Assessments Login
Public Assessments
Study list
VHuafcations
PFBS (2018)
Download datasets
I Hidden on puMc page?
Editable True
Public True
Figure D-l. HAWC homepage for the public PFBS assessment.
D.2. How Do I Access HAWC?
HAWC is an open-source online application that may be accessed using the following link:
https://hawcprd.epa.gov/assessment/public/ and then selecting an available assessment. The
following browsers are fully supported for accessing HAWC: Google chrome (preferred),
Mozilla Firefox, and Apple Safari. There are errors in functionality when viewed with Internet
Explorer. No user account is required for access to public HAWC assessments. The assessments
located in HAWC are meant to accompany a textual expert synthesis of the data managed in
HAWC. Each written assessment document contains embedded URL links to the evidence in
HAWC (e.g., study evaluation, summary study data, visualizations, etc) supporting the
A-l

-------
assessment text. The links embedded in an assessment document can be accessed by a mouse
click (or hover while pressing CTRL+right click).
D.3. What Can I Find in HAWC?
HAWC contains a comprehensive landscape of study details and data supporting an assessment.
Note that links are provided in the assessment text to guide the reader, but a user may also
navigate to the HAWC homepage for an assessment on their own. Once a user lands on an
assessment homepage all studies included in an assessment can be viewed by clicking the blue
"Study list" link (highlighted in the red box below) in the grey navigation pane (Figure D-2). By
clicking the study name listed in blue (under "Short citation") a user can view the full study
details, study evaluation, and experimental details and data. For example, in Figure D-2, a user
may click on 3M (2000d) (highlighted in orange dashed box below). This will take the user to
the 3M (2000d) study details page that includes a link to the study in FIERO along with study
details, study evaluation, and available experimental (animal) and study population
(epidemiologic) groups.
4- O O i Secure https hawcprd.epa.gov V jdy/assessmerrt/100000037/
Q 
-------
pie graph (Figure D-3). For full domain and rating details the user may click the blue
"View details" button (highlighted in the red box below). (Note that this example is given for the
3M (2000dV).
Risk of bias visualization
Reporting domain
Reporting of information necessary for study
evaluation
Good. Important information is provided for test
species, strain, source, sex, exposure methods,
experimental design, endpoint evaluations and the
presentation of results. The study was not conducted
under GLP guidelines; however, several reviews were
performed by the Quality Assurance Unit.
Reporting or AttritMariable Control
3M, 2000, 4289992 risk of bias summary
Figure D-3. Representative study evaluation pie chart with the reporting domain selected
and text populating to the right of pie chart.
D.5. How Do I Access Study Specific Information on Experimental and Study
Population Details, and Extracted Endpoint Data?
Specific information on experimental design, dosing (if animal bioassay), outcomes and
exposure (if epidemiology) and extracted endpoint data can be accessed from the study details
page by clicking on (for the 3M (2000d) study) Available animal bioassay experiments at the
bottom of the study details page. A user may click on the experiment name (highlighted in blue,
10 Day Oral) to view dosing/exposure details and available groups. Clicking on available animal
groups (e.g., Male Sprague-Dawlev or Female Sprague-Dawlev) will take the reader to a new
page with experimental group information (e.g., species/strain/sex , dosing regime information,
and available/additional endpoints information for animal studies; and outcome and exposure
information for epidemiologic studies. If a study reports data then the data are extracted and
managed as "available endpoints". If study authors include endpoints in the methods and results,
but do not report data the endpoint is listed under "additional endpoints" without dose-response
data. All endpoints are also clickable and contain an endpoint description, methods, and (if data
are reported) a clickable data plot (e.g., Alanine Aminotransferase (ALT)). The description of
endpoints, methods, and data are often copied directly from the study report and, therefore, can
contain study author judgments and may not necessarily include EPA judgments on the endpoint
data that would be included in the assessment.
D.6. What Are Visualizations and How Do I Access Them?
The data managed in HAWC is displayed using visualizations that are intended to support textual
descriptions within an assessment. All visualizations can be accessed using the blue
"Visualizations" link (highlighted in the red box below) also found in the grey navigation pane
A-3
-------
(Figure D-4A). Note that the available visualizations are at the discretion of the chemical
manager and are meant to accompany the assessment text. Visualizations are fully interactive.
Hovering and clicking on records in the rows and columns and data points on a plot will cause a
pop-up window to appear (Figure D-4B). This pop-up window is also interactive and clicking on
blue text within this pop-up will open a new web page with descriptive data.
HAwfr
Contact About Public Assessments Login
Pufcbc Assessments PFBS (2010) Visualizations PFBST
SELECTED ASSESSMENT X

AMUUUtlX UOOULES



Visualizations

DOWNLOADS
Download datasets

PFBS T4 (effect size, animal)
hp«flfl<»nl MpOM*	Unm
T'201®. 43MMI 28 Day o*' TWnuaoMhjfCir** , f#W. Hart«r,	<
Hit MrW. Scra9*-Onwto* < f. N-9
Figure D-4A. Visualization example for PFBS. (Note that the records listed under each
column (study, experiment endpoint, units, study design, observation dine, dose) and data
within the plot are interactive.)
NTP 2018,4309741 / 28 Day Oral / Male Harlan Sprague Dawley Rat / Tetraiodothyronine (T4), Free
Study Experiment Animal Group Endpoint
Endpoint name
Tetraiodothyronine (T4), Free
System
Endocrine
Organ Thyroid
Effect
Hormone
Effect subtype
Thyroid Hormone
Observation time
Day 28
Data reported?
*
Data extracted?

Values estimated?
~
Location in literature
R07-Hormone Summary
Expected response
any change from reference/control group
adversity direction

LOAEL
62.6 mg/kg-day
Monotonicity
not-reported
Statistical test description
Jonchkeere {trend) and Shirley or Dunn (pairwise) tests
Trend result
not reported
Power notes
Statistical analyses performed by Jonchkeere (trend) and Shirley or Dunn (pairwise) tests Statistical significance for a

treatment group indicates a significant pairwise test compared to the vehicle control group Statistical significance for the

control group indicates a significant trend test" Statistically significant at P <= 0.05 " Statistically significant at P <= 0.01
Ciose
Figure D-4B. Example pop-up window after clicking on interactive visualization links.
(In Figure D-4A the red circle for study NTP (2019); male at a dose of 500 mg/kg-day was
clicked leading to the pop-up shown above. Clicking on blue text will open a new window
with descriptive data.)
A-4
-------
D.7. How do I download datasets?
A user may download any available dataset by first clicking on the blue "Download datasets"
link (highlighted in the red box below) in the grey navigation pane on the assessment homepage.
This takes the user to a new page where the desired data set may be selected for download as an
excel file (See representative image in Figure D-5).
 ¦£{	Upgrade now
space and sha
Comae? About Public Assessments Login
Study lis?
Visualizations
Download ctaiaseis
PFBS (2018) downloads
it data from HAWC are exportable wo Excel. Devefoper exports in JSOW format are also available (please contact us Sor more infermatfon),
• Literature-review
Microsoft Excel spreadsheet
il bioassay data
Complete export ¦ Endpoint summary
Microsoft Excel spreadsheet
Epidemiology data
Microsoft Excel spreadsheet
In-vitro data
Microsoft Excel spreadsheet
VhWliMtlOBI
ViSMtftjatfOns con be exported into SVG. PNG. PDF. and PPTX formats; you can download each individual visualization when bent*) viewed.
Figure D-5. Representative data download page.
D.8. How Do I Access the Benchmark Dose Modeling Outputs?
Benchmark dose (BMD) modeling is performed on an endpoint by endpoint basis at the
discretion of the chemical manager. Those endpoints for which BMD modeling has been
completed are referenced in the assessment text and are available for viewing. To access BMD
modeling outputs the user can click on links included in the assessment text. Alternatively, the
user may navigate to the BMD modeling outputs by clicking on a study (e.g., Feng et al. (2017))
of interest from the Study list, an available animal bioassay experiment (in this example the 20
Day Oral Gestation), an available animal group (PO Female ICR Mice), and an endpoint of
interest (Tetraiodothyronine (T4). Free). Next navigate to the blue Actions button, click, and
scroll to "View session" (highlighted in the red box below) under BMD Modeling (Figure D-
6A). The BMD setup. Results, and Model recommendation and selection (highlighted in orange
dashed box below) are available for viewing (Figure D-6B). Selecting the BMD setup tab will
display the modeled dose-response data, the selected models and options, and all benchmark
modeling responses (BMRs). The results tab will display the BMD modeling output summary for
all models. A user may hover over a selected model row to visualize the model fit to the data. In
addition, a user may obtain the Benchmark Dose Software (BMDS) Output text by clicking the
"View" button under the "Output" column for each model that was run. The Model
recommendation and selection tab displays all models, warnings when appropriate, and the
recommendation for which models are valid, questionable, or failed to fit.
A-5
-------
4- C i Secure http5 ,,,,'1iawcprd.epa.gov,''arii/erKfpoint/l00002285/
Public Assessments PF0S (2018) Feng, 2017. 3856465 20 Day Oral Gestation PO Femafe ICR Mice Tetrafodottiyrc
Contact About Public Assessments Login
SELECTED ASSESSMENT
AVAILABLE MODULES
Study list
Visualizations
DOWNLOADS
Download cfaiasets
Tetraiodothyronine (T4), Free
Endpoint Details
Endpoint name
Tetralodothyrontne (T4), Free
System
Endocrine
Organ
Thyroid
Effect
Hormone
Effect subtype
Thyroid Hormone
Observation time
GD20
Additional tags
high confidence
Data reported?
<~
Data extracted?
~
Values estimated? —
Location in literature
Table 3
NOAEL
50 mg/Vg-day
Plot
BMP Modeling
Trtralodothyronli .
View session
100 200 300 400 500
Deae (myKgOay)	
Figure D-6A. Example BMD modeling navigation.
C i Secure https//hawcp*d.ep«.gov,,:;r»ir',/swion/100000019y'
Contact About Public Assessments Login
Public Assessments PFBS (2018) Feng, 2017.
20 Day Oral Gestation PO Female ICR Mice Telraioctothyronine (T4). Free BMD
SELECTED ASSESSMENT
BMD session
&MD setup Results Model recommendation and selection
Study list




Visualizations
Dose-response


DOWNLOADS
Dose (mg/kg-day



Download dataseis
HED) >
Number of Animals
Response (pg/ml)
Standard Error

0
8
16.81
0.7

7.8"
8
17.58
0.64

31b,s:
8
14.74
0.51

77b
8
14.95
0.46
* NOAEL (No OMcnwd Adverse Effecl Level*
6 Stgndcantly iliflWnl from control Ip < 0.05)
1LOAEL {Lowest Observed Advert# EtfecS Level)
Selected models and options
BMDS V2.7.0
Model options
Model nam* Non-default settings
Tetraiodothyronine (T4J, free
t
^Dose* in Study
0LOAEL
^NOAEL
T
I
0 20 40 60 60
Ooae (mg'k^-day HEO)
o
« •
J. *
Benchmark modeling responses
Confidence
Figure D-6B. Example BMD session.
A-6
-------
Appendix E. Additional Data Figures
Experiment Cndpoiul
Units Study Design
Observation Confidence
NTP20I8	28-day oral Free Tetraiodotbyronine 
-------
Studv
Stud) Design
Route
Exposure
Knd point

Confidence
1" nits
Dose
(mg/kg-day)
NTP, 2018,4309741
Rat, Harlan Sprague-Dawley (9- N=l-10)
oral gavage
28 days
Triiodothyronine (T3)
Ihigh confidence!
ng/dl.
0
62.6








125








250








500








1.000

Rat. Harlan Sprague-Dawley (d\ N=9-10)
oral gavage
28 days
Triiodothyronine (T3)

ng/dL
0
62.6
125
250
500
Peng, 2017. 3856465
F1 Mouse. ICR (9. N=10)

GDI to 20
Triiodothyroni
ine (13)

n8ta'
0
50
200








500








0








50








200








500

F1 Mouse. ICR (9- N=30)
oral gavage
GDI to 20
Triiodothyroni
ine(T3)

ng/ml
0








50








200
500

P0 Mouse. ICR (9, N=8)
oral gavage
GDI to 20
Triiodothyroni
ine (T3)

ng/ml
0
50
200
500

Fl Mouse. ICR (9. N=I0)
oral gavage
GDI to 20
Triiodothyronine (T3I - Litter N

ng/ml
0








50








200
-60 -40 -20 0 20 40 AO 80 100
percent control response
Figure E-2. Serum total triiodothyronine (T3) response in animals following K+PFBS
exposure (click to see interactive data graphic).
A-2
-------
Study
Study Design
Exposure
observation time
Units
Dose
(mg/kg-day)
NTP 2018.4309741
Rat. Harlan Spraguc-Dawley (9. N=l-10)
28 days
Day 28
ng/mL
0
62.6
125
250





500





1,000

Rat. Harlan Spraguc-Dawley (cf, N=9-10)
28 days
Day 28
ng/mL
0
62.6
125





250





500
Feng 2017. 3856465
F1 Mouse, ICR ($, N=10)
GDI to 20
PND1
ng/ml
0
50
200
500



PND30
ng/ml
0
50
200





500



PND60
ng/ml
0





50





200
500

P0 Mouse, ICR (9, N=8)
GDI to 20
C.D20
ng/ml
0
D statistically significant
£ percent control response
IH 95% CI
200
500













<






















H
1

H
!-•




















-100 -80 -60 -40 -20 0 20 40 60 80 100
percent control response
Figure E-3. Serum thyroid-stimulating hormone (TSH) response in animals following
K' PFBS exposure (click to see interactive data graphic).
Experiment Endpoint	Units Study Design
observation Confidence Dose
time	(mg/kg-day)
Feng. 2017. 3860465 20-day oral EyeGpenir^g	days F1 Mouse. ICR (2. M=50) Beginning on
PND12
Eye Opening - Litter N days F1 Mouse, SCR (S. N=10) Beginning on
PND12
200
500
200
50Q
~ statislicaSy significant
0 percent control, response
M 65% CF
	1	1	1	1	!	1	1	
-100 -30 -60 -40 -20 0 20 40 60
percent controf response
Figure E-4. Developmental effects (eye opening) following K+PFBS in rats
(click to see interactive data graphic).
A-3
-------
Study
Study Design
Route	Exposure Endpoint
Units Dose
(mg/kg-day)
O statistically signific
£ percent control resj
H 95% CI
Feng, 2017, 3856465 F1 Mouse. ICR (9. N=30) oral gavage GDI to 20 Estrous Cycle. Diestrus days 0
50
200
500
FI Mouse. ICR (9. N= 10) oral gavage GDllo20 Estrous Cycle, Proestrus days 0
50
200
500
M H
~I	1	1 I
©
~i	1	1	r~
-100 -80 -60 -40 -20 0 20 40 60 80 100
percent control response
Figure E-5. Developmental effects (first estrus) following K+PFBS in rats
(click to see interactive data graphic).
Study
Study Design
Route
Exposure
Kndpoint
Units
Dose
(mg/kg-day)
Lieder. 2009. 1578545
Fl Rat. Sprague-Dawley (9. N=30)
oral gavage

Vaginal Patency
days
0
30
100






300






1.000
Feng. 2017. 3856465
Fl Mouse. ICR (9. N=30)
oral gavage
GDI to 20
Vaginal Patency
days
0
50
200
500

Fl Mouse. ICR (9. N=I0)
oral gavage
GDI to 20
Vaginal Patency (litter)
days
0
50
200
500
O statistically significant
0 percent control response
H9S% CI
o
-100 -80 -60 -40 -20 0 20 40 60 80 100
percent control response
Figure E-6. Developmental effects (vaginal patency) following K+PFBS in rats
(click to see interactive data graphic).
A-4
-------
Animal de%cri(»tion
(considered
Lieder, 2009, 1578546 Ral. CrK'
c*»l ravage
reproductive
Kidney, Edema. Focal Papillary
0/34) (0.0%)
0





7/30 (23.3%)
300





4/30413.3%)
1.000
1



Kidney. Hyperplasia, Papillary
Tubtilar/Ducial Epithelium
2/3046.7%)
0
¦




13/30(43.3%)
300





15/30 (90.0%)
1.0(10

I I Rat, Sprague Dawlcy (Ct. N=30)
oral savage
reproductive
Kidney. Edema, Fucal Papillary
1/30(3.3%)
0
~




(V30<4X0%>
300





9/30430.0%)
1.4)00




Kidney. Hypoplasia, Papillary
Tahalar/Ductal Epithelium
3/30410.0%)
0
=¦




5/30416.7%)
300
i




21/30 (70-0%)
1.000

PO Rut. Sprnguc-Duwky 49. N»X»
oral jsvage
reproductive
Kidney. Edema. Focal Papillary
1/3043.3%)
8/30426.7%)
0
300
B




7/30(23.3%)
1,000
~i



Kidney. Hyperplasia, Papillary
Tolxilai/Ducial Epithelium
3/30410.0%)
0
1=1




16/30 (53.3%)
3(H)





21/30 (70.0%)
1.4100

Kl Rat. Sprague-Dawtcy (tf. N=30)
oral savage
reproductive
Kidney. Edema. Focal Papillary
1/3043-3%)
2/3046.7%)
0
300
~




6/30420.0%)
1.000
i



Kidney, Hyperplasia. Papillary
Tulxilar/Ductal Epithelium
0/30(0.0%)
0





9/30430.0%)
300
¦¦¦¦I




IW30 (63.3%)
1.000
—i—i—i—i—i—i—i—i—i—i—i—i—
M 16 IS 20
Figure E-7. Kidney histopathological effects following K+PFBS in rats
(click to see interactive data graphic).
A-5
-------
End point Name
Study Name
F.xpcrimont Name
Animal Description
Observation Time
Bl
Day 90
Creatinine (CREAT)
3M, 2000, 4289992
10 Day Oral
Rat, CH: Cd (Sd) lbs Br (cf)
Day 11



Ral. CH: Cd (Sd) lbs Br (9)
Day 11

NTP 2018. 4309741
28 Day Oral
Ral. Harlan Spraguc-Dawlcy ((f)
Day 28



Rat. Harlan Spraguc-Dawlcy 19)
Day 28

3M. 2001.4241246
28 Day Oral
Rat, Crl:CD(SD) (cf)
Day 29



Rat. Cri:CD(SD) (9)
Day 29

Lieder. 2009. 1578546
90 Day Oral
Rat. Cri:Cd@(Sd)lgs Br Vaf/Plustm icf)
Day 90



Rat. Crl:Cd@(Sd)Igs Br Vaf/Plustm id")
Day 90



Rat Crl:Cd@(Sd)Igs Br Vaf/PlusUn (9)
Day 90
Kidney Basophilia. Tubular, Multifocal
Lieder, 2009, 1578546
90 Day Oral
Rat, Crl:Cd@(Sd)lgs Br Vaf/Plustm (Cf)
Day 90



Rat. Crl:Cd»(Sd)lgs Br Vaf/Plustm <9>
Day 90
Subacute Inflammation. Cortex
3M. 2001.4241246
28 Day Oral
Rat. Crl:CD(SD) (Cf)
Day 29
Kidney. Mononuclear Cell Infiltrate
Lieder. 2009, 1578546
90 Day Oral
Rat. Crl:Cd(Sd)Tgs Br Vaf/Plustm (9)
Day 90
Kidney, Mineralization
3M, 2000, 4289992
10 Day Oral
Rat, CH: Cd (Sd) Ths Br (Cf)
Day II



Rat, Crl: Cd (Sd) lbs Br (9)
Day 15

Lieder. 2009, 1578546
90 Day Oral
Rat. Crlfd@(Sd)Igs Br Vaf/Plustm icf)
Day 90



Rat. Crl:Cd@(Sd)lgs Br Vaf/Plustm (9)
Day 90
Mineralization
3M. 2001,4241246
28 Day Oral
Rat, Crl:CD(SD) (9)
Day 29
Mineralization, Terminal
3M, 2001,4241246
28 Day Oral
Rat, Crl:CD(SD) (9)
Day 29
Kidney, Nephropathy, Chronic Progressive
NTP 2018, 4309741
28 Day Oral
Ral. Harlan Sprague-Dawley (cf)
Day 28



Rat, Harlan Spraguc-Dawlcy (9)
Day 28
Kidney. Hyaline Droplets, Cortical Tubules
Licdcr. 2009. 1578546
90 Day Oral
Rat, Crl;Cd(*(Sd)lgs Br Vaf/Plustm fcf)
Day 90



Ral, CH:Cd#(Sd)lgs Br Vaf/Plustm (9)
Day 90
Kidney Hislopalbolugy
3M. 2000,4289992
10 Day Oral
Ral, CH: Cd (Sd) lbs Br (Cf)
Day 11



Ral, CH: Cd (Sd) lbs Br (9)
Day II

3M, 2001,4241246
28 Day Oral
Rat, Crl:CD(SD) (cf)
Day 29



Rat, Crt:CD(SD) (9)
Day 29
FIBS Kidney Effects
Dow Range
Signdicanl Increase
.Significant Decrease
100 200 300 400 500 600 700 800 900
Dose (mg/kg/dav)
Figure E-8. Renal effects following K+PFBS in rats (click to see interactive data graphic).
A-6
-------
Kndpnint Name
Study Name
Experiment Name
Animal Description
Observation 'lime

PFBS Kidney Weight Effects



Kidney Weight. Absolute
Lieder, 2009. 1578545
2 Generation Oral
F2 Ral. Sprague-Dawley «f9)


~ * •	•	



• Doses

3M. 2000.4289992
10 Day Oral
Rat. Crl: Cd 
Day 120

M »	•	

—~




Fl Rat, Sprague-Dawley (9)
Day 120

• •	

—~


NTP 2018,4309741
28 Day Oral
Rat, Harlan Sprague-Dawley (Cf)
Day 28

~ > «	•	A	

—~




Rat, Harlan Sprague-Dawley (9)
Day 28

» • •	•	•	

—~

Kidney Weight. 1 -eft. Relative
Lieder. 2009,1578545
2 Generation Oral
P0 Rat. Sprague-Dawley (cf)


»• «	•	

—~




P0 Ral. Sprague-Dawley (9)
LD22

M—•	•	

—~




Fl Rat. Sprague-Dawley (cf)
Day 120

»	•	

—~




Fl Rat. Sprague Dawley (9)
Day 120

M » •

—~

Kidney Weight, Relative
Lieder, 2009, 1578545
2 Generation Oral
F2 Rat. Sprague-Dawley (cf 9)


H—•	•	

—~


3M. 2000.4289992
10 Day Oral
Rat. Crl: Cd (Sd) lbs Br (Cf)
Day 11

~	•	•	

—4




Rat. Crl: Cd (Sd) lbs Br (9)
Day 11

~	~	•	

—~


3M, 2001,4241246
28 Day Oral
Rat, Crl:CD(SO) (cf)
Day 29

~ • •	






Rat. Crl:CD(SD) (9)
Day 29

•	•	•	
—A



Lieder. 2009,1578546
90 Day Oral
Rat, CH:CdW(Sd)lgs Br Vaf/Plustm (cf)
Day 90

~—•	•	~






Rat. Crl:Cd@(Sd)Igs Br Var/Plustm (9)
Day 90

	•	~



Kidney Weight. Right. Relative
Lieder. 2009.1578545
2 Generation Oral
PO Ral. Sprague-Dawley (cf)


M » •	

—~




PO Rat. Sprague-Dawley (9)
LD22

M—•	•	

—~




FI Ral. Sprague-Dawley (cf)
Day 120

H—I	•	

—~




Fl Rat, Sprague-Dawley (9)
Day 120

M—•	•	

—~


NTP 2018.4309741
28 Day Oral
Rat, Harlan Sprague-Dawley (Cf)
Day 28

» • •	•	A	

—~




Rat, Harlan Sprague-Dawley (9)
Day 28

~ AA A	A	

-A






-1
DO 0 100 200 300 400 500 600 700 8
» 900
1.000 1.
00






Dose (nig/kg/day)



Figure E-9. Kidney weight effects following K+PFBS in rats (click to see interactive data
graphic).
A-7
-------
Endpoint Name
Study Name
Experiment Name
Animal Description
Observation Time


PFBS Liver Effects


Alanine Aminotransferase (ALT)
3M. 24 KM), 4289992
10 Day Oral
Rat, Crl: Cd (Sd) lbs Br(
Day 29
~	•—
	•—







Rat Crt:CD
Day 90
~—•	





Hcpatocytc Cytoplasmic Alteration
NTP 2018. 4309741
28 Day Oral
Rat. Harlan Spraguc-Dawlcy (9)
Day 28
» • •



—~




Rat. Harlan Sprague-Dawley (cf)
Day 28
• • •



—~

Cellular Infiltration
NTP 2018, 4309741
28 Day Oral
Rat. Harian Sprague-Dawley (9)
Day 28
~ * »



—~




Rat. Harlan Spraguc-Dawlcy (cf)
Day 28
• • •



—~

Mononuclear Cell Infiltrate
3M. 2000. 4289992
10 Day Oral
Rat, Crl; Cd (Sd) lbs Br (cf)
Day 11
~	•—



—~




Rat. Crl: Cd (Sd) lbs Br (9)
Day 15
~	•—



—~

Inflammation, Subacute
3M, 21X11, 4241246
28 Day Oral
Rat, Crl:CD(SD) (9)
Day 29
•—


—~





Rat, CrI:CD








Rat. Harlan Sprague-Dawley (Cf)
Day 28
~ • •






Licdcr. 2009. 1578545
2 Generation Oral
F1 Rat, Spraguc-Dawlcy (9)
Day 120
M—»—








Fl Rat. Sprague-Dawley (Cf )
Day 12(1
M—•—



—~




P0 Rat, Sprague-Dawley (9)
LD22
M—•—





Liver. Necrosis
NIP 2018, 4309741
28 Day Oral
Rat, Harlan Sprague-Dawley (9)
Day 28
~ • •








Rat. Harlan Spraguc-Dawlcy (*?)
Day 28
» • •





Necrosis
3M, 2001, 4241246
28 Day Oral
Rat, Crl:CD(SD) (Cf)
Day 29
~ *





Necrosis. Terminal
3M. 2001. 4241246
28 Day Oral
Rat Crl:CD(SD) (cf 1
Day 29
~	•—





Bile Duct Cyst
NTP 2018. 4309741
28 Day Oral
Rat, Harlan Sprague-Dawley (9)
Day 28
~ • •
—•—
•






Rat, Harlan Spraguc-Dawlcy (cf I
Day 28
~ • •
—•—
•	



llepatodiaphragmatic Nodule
NTP 2018, 4309741
28 Day Oral
Rat. Harlan Sprague-Dawley (9)
Day 28
~ > ~
—•—
	•	






Rat. Harlan Spraguc-Dawlcy (Cf)
Day 28
• • •
—•—
	•	



Hemorrhage
NTP 2018. 4309741
28 Day Oral
Rat, Harlan Sprague-Dawley (9)
Day 28
~ • •
•
	•	






Rat, Harlan Sprague-Dawley (o*i
Day 28
~ > •
•
•



Liver Hlstopathology
3M, 2000, 4289992
10 Day Oral
Rat, Crl: Cd (Sd) lbs Br (9)
Day 11
~	•—








Rat. Crl: Cd (Sd) lbs Br (cf)
Day 11
*	•—






3M, 2001, 4241246
28 Day Oral
Rat. Crl:CD •—
	A-


-A


3M. 2000. 4289992
10 Day Oral
Rat, Crl: Cd (Sd) lbs Br (9)
Day 11
~ •



-A




Rat. Cii: Cd (Sd) lbs Br(d")
Day 11
~—•—



-A


NTP 2018, 4309741
28 Day Oral
Rat, Harlan Sprague-Dawley (9)
Day 28
» • A-

	A	

—A




Rai. Harlan Sprague-Dawley (Cf)
Day 28
~-AA-
—A—
	~	




3M. 2001,4241246
28 Day Oral
Rat, Crl:CD(SD) (9)
Day 29
~—•—








Rat. Crt:CD(Sd)lgs Br Vaf/Plustm (9)
Day 90
	








Rat. Crl:Cd@'{Sd)Igs Br Vaf/Plustm (cf)
Day 90
•	






Lieder, 2009, 1578545
2 Generation Oral
Fl Rat, Sprague-Dawley (Cf)
Day 120
M—•—



-A





-1
» 0 100
200 300
400 500 600 700 8
0 'XXI
1,000 1,
X)







Dose (mg/kg/day)


Figure E-10. Liver effects following K+PFBS in rats (click to see interactive data graphic).
A-8
-------
Endpolnt Name	Study Name Experiment Name	Animal Description	Observation Time
Apolipopiotein A1 (ApoAl)
Bijlund, 2011.1578502
Subchronic Oral
Mouse. Apoe"3-Leklen.Celp (cf)
Week 4
M
|—| Dose Range
Bile Acid Secretion
Bijbuid.2011, 1578502
Subchronic Oral
Mouse. Apoc*3-Leidcn.Cclp (Cf)
Week 4

• Doses
Chnbwml KsterTrunsfer Protein (OK
IT) Bijland, 2011. 1578502
Suhchmnic Oral
Mouse. Apoe"3-i.cklen.Cclp 
Day II

V Significant Decrease



Ral. Crl; Cd (Sd) lbs Br (cf)
Day 11
•—a	a	•


3M. 2001.4241246
28 Day Oral
Rat. CrtCD(SD) (Cf)
Day 29



Bijland, 2011,1578502
Subchronic Oral
Mouse. Apoc"3-Leiden.Cclp (Cf)
Week 1



l.foder, 2009, 1578546
90 Day Oral
Ral. C'rK'd W(SdHgs Br VaPI'lusIm (9)
Day 90





Rat. CrlCd®(Sd)lgs Br VaPPhuim («f)
Day 90
~—•	•	~

Cholesterol Estet (CE)
Bijland, 2011. IS78502
Subchronic Oral
Mouse. Apoe*3-Leiden,Cetp (0"l
Week 4


ChulesleiuL Free
Bijlund, 2011.1578502
Subcluunk- Oral
Mouse, Apoe"3-Leklen.Cclp 
-------
Appendix F. Benchmark Dose Modeling Results
F.l. Modeling of Noncancer Endpoints
As discussed in the body of the report under "Derivation of Oral Reference Doses," the
endpoints selected for benchmark dose (BMD) modeling were incidence of renal papillary
epithelial tubular/ductal hyperplasia in rats from Lieder et al. (2009a) and Lieder et al. (2009b);
thyroid hormones in pregnant mice and offspring at postnatal day (PND) 1, PND 30, and
PND 60 from Feng et al. (2017) and adult rats from NTP (2019); and developmental effects
(i.e., eye opening, first estrus, vaginal opening) from Feng et al. (2017). The animal doses in the
study, converted to human equivalent doses (HEDs), were used in the BMD modeling; the data
are available for download in Health Assessment Workspace Collaborative (HAWC). BMD
modeling was conducted by experts in quantitative Benchmark Dose Software (BMDS) analysis
and interpretation. Links to the data and modeling output are included in Table F-l. The selected
point of departure (POD) (HED) listed in Table F-l represents the best fitting model for each
endpoint; if the data were determined to not be amenable to BMD modeling, the no observed
adverse effect level (NOAEL) or lowest observed adverse effect level (LOAEL) is listed.
Figure F-l illustrates the doses examined and NOAEL, LOAEL, BMD, and benchmark dose
lower confidence limit (BMDL) values for the potential critical effects.
Table F-l. Candidate PODs for the derivation of the subchronic and chronic RfDs for
PFBS (CASRN 375-73-5) and the related compound K+PFBS (CASRN 29420-49-3)
Endpoint/reference
Species/life stage—sex
Selected POD (HED)a
(mg/kg-d)
Kidney effects
Kidney histopathology—papillary epithelial tubular/ductal
hvDcrolasia—Lieder et al. (2009a)
Rat/Male
BMDLm = 0.489
Rat/Female
BMDLio= 0.300
Kidney histopathology—papillary epithelial tubular/ductal
hvDcrolasia—Lieder et al. (2009b)
Rat/Po—Male
BMDLm = 0.351
Rat/Po—Female
BMDLm = 0.265
Kidney histopathology—papillary epithelial tubular/ductal
hvDcrolasia—Lieder etal. (2009b)
Rat/Fi—Male
BMDLm = 0.776
Rat/Fi—Female
BMDLm = 0.478
Thyroid effects
Total T4 - NTP (2019)
Rat—Male
LOAEL = 0.34
Rat—Female
BMDLi sd = 0.037
Free T4-NTP (2019)
Rat—Male
LOAEL = 0.34
Rat—Female
BMDLi sd = 0.027
Total T4—Fens et al. (2017)
Mouse/Po—Female
BMDLi sr> = 0.093
Free T4—Fens et al. (2017)
Mouse/Po—Female
NOAEL = 0.21
TSH—Fens et al. (2017)
Mouse/Po—Female
NOAEL = 0.21
Total T4 PND 1 (fetal n)h —Fens et al. (2017)
Mouse/Fi—Female
NOAEL = 0.21
Total T4 PND 1 (litter n)b —Fens et al. (2017)
Mouse/Fi—Female
BMDLo.ssd = 0.095
(BMDLisd = 0.25)
A-l
-------
Endpoint/reference
Species/life stage—sex
Selected POD (HED)a
(mg/kg-d)
Total T4 PND 30—Feng et al. (20171
Mouse/Fi—Female
NOAEL = 0.21
Total T4 PND 60—Feng et al. (20171
Mouse/Fi—Female
NOAEL = 0.21
TSH PND 30—Feng et al. (20171
Mouse/Fi—Female
NOAEL = 0.21
Developmental effects
Eves opening (fetal «1b—Feng et al. (20171
Mouse/Fi—Female
NOAEL = 0.21
Eves opening (litter «1b —Feng et al. (20171
Mouse/Fi—Female
BMDLn^sn = 0.073

(BMDLisd = 0.16)
Vaginal opening (fetal n)b —Feng et al. (20171
Mouse/Fi—Female
BMDLo ^sn = 0.15

(BMDLisd = 0.35)
Vaginal opening (litter n)b —Feng et al. (20171
Mouse/Fi—Female
BMDLo ssn = 0.094
(BMDLisd = 0.22)
First estrous (fetal n)b —Feng et al. (20171
Mouse/Fi—Female
NOAEL = 0.21
First estrous (litter n)b —Feng et al. (20171
Mouse/Fi—Female
NOAEL = 0.21
Notes: BW = body weight; RfD = reference dose; PFBS = perfluorobutane sulfonic acid; CASRN = Chemical
Abstracts Service Registry Number; K+PFBS = potassium perfluorobutane sulfonate; T3 = total triiodothyronine;
T4 = total thyroxine; TSH = thyroid-stimulating hormone.
aFollowing U.S. EPA (201 lb") guidance, animal doses from candidate principal studies were converted to HEDs
through the application of a dosimetric adjustment factor (DAF), where HED = dose x DAF. See Table 8 in
assessment for full details. Links are to the HAWC BMDS session containing full modeling results for that endpoint.
b Fetal endpoints from Feng et al. (20171 were modeled alternatively using dose group sizes based either on total
number of fetuses or dams. Given that it appears that Feng etal. (20171 did not use the litter as the statistical unit of
analysis, it is unclear if the study-reported standard errors pertain to litters or fetuses. Alternatively, modeling fetal
endpoints using litter n or fetal n provides two modeling results that bracket the "true" variance among all fetuses in
a dose group (i.e., using the fetal n will under-estimate the true variance while using the litter n will over-estimate
the true variance). Individual animal data were requested from study authors but were unable to be obtained.
A-2
-------
study name
experiment name
animal description
endpoinl name
observation time
Lieder, 2009, 1578546
90 Day Oral
Rat, Crl:Cd@(Sd)lgs Br Vaf/Plustm (cf)
Kidney, Hyperplasia, Papillary Tubular/Ductal Epithelium
90.0 days


Rat. CrI:Cd@(Sd)Igs Br Vaf/Plustm (9)
Kidney, Hyperplasia, Papillary Tubular/Ductal Epithelium
90.0 days
Lieder, 2009,1578545
2 Generation Oral
P0 Rat, Sprague-Dawley (cf)
Kidney, Hyperplasia. Papillary Tubular/Ductal Epilhelium
None not reported


PO Rat, Sprague-Dawley (9)
Kidney, Hyperplasia. Papillary Tubular/Ductal Epithelium
None not reported


PI Rat. Spraguc-Dawlcy (cP)
Kidney, Hyperplasia, Papillary Tubular/Ductal Epithelium
120.0 days


F1 Ral, Sprague-Dawley (9)
Kidney, Hyperplasia. Papillary Tubular/Ductal Epilhelium
120.0 days
NTP, 2018,4309741
28 Day Oral
Rat. Harlan Sprague-Dawley ((f )
Tetraiodothyronine (T4), Free
None not reported


Rat, Harlan Sprague-Dawley (9)
Tetraiodothyronine (T4), Free
None not reported


Rat. Harlan Sprague-Dawley (cf)
Tetraiodothyronine (T4), Total
None not reported


Rat. Harlan Sprague-Dawley (9)
Tetraiodothyronine (T4), Total
None not reported
Peng 2017,3856465
20 Day Oral Gestation
P0 Mouse, ICR (9)
Tetraiodothyronine (T4), Free
Tetraiodothyronine (T4), Total
20.0 gestational day (GD)
20.0 gestational day (GD)


Fl Mouse, ICR (9)
Tetraiodothyronine (T4). Total
1.0 post-natal day (PND)
30.0 post-natal day (PND)
60.0 post-natal day (PND)



Tetraiodothyronine (T4). Total - Litter N
1.0 post-natal day (PND)


PO Mouse, ICR (9)
Thyroid Stimulating Hormone (TSH)
20.0 gestational day (GD)


Fl Mouse. ICR (9)
Thyroid Stimulating Hormone (TSH)
1.0 post-natal day (PND)
30.0 post-natal day (PND)
60.0 post-natal day (PND)



Eye Opening - Fetal N
12.0 post-natal day (PND)



Eye Opening - Litter N
12.0 post-natal day (PND)



First Estrous - Fetal N
24.0 post-natal day (PND)



First Estrous - Litter N
24.0 post-natal day (PND)



Vaginal Opening - Fetal N
24.0 post-natal day (PND)



Vaginal Patency - Litter N
24.0 post-natal day (PND)
I'KBS Candidate PODs for RfDs
-0— *-
0-
-©—e-
O bmd
Q BMDL
O NOAEL
O LOAEL
• Doses
|—| Dose Range
o o
o ©
o
c -
o
-1—I I I 11 11
o •o
Dose (mg/kg/day)
Figure F-l. Candidate PODs for the derivation of the subchronic and chronic RfDs for PFBS
(click to see interactive data graphic).
A-3
-------
F.2. Modeling Procedure for Continuous Noncancer Data
BMD modeling of continuous data was conducted on the HAWC website using the U.S.
Environmental Protection Agency's (EPA's) BMDS (Version 2.7). All continuous models
available within the software were fit using a benchmark response (BMR) of 1 standard
deviation (SD). For continuous data of effects in developing offspring, including thyroid
hormone changes, a BMR of 0.5 SD change from the control mean is used for to account for
effects occurring in a sensitive life stage. A 1 SD BMR is also presented as the basis for model
comparison as directed in the EPA BMD Technical Guidance (U.S. EPA. 2012). An adequate fit
is judged based on the %2 goodness-of-fit p-value (p > 0.1), magnitude of the scaled residuals in
the vicinity of the BMR, and visual inspection of the model fit. In addition to these three criteria
forjudging adequacy of model fit, a determination is made as to whether the variance across
dose groups is homogeneous. If a homogeneous variance model is deemed appropriate based on
the statistical test provided by BMDS (i.e., Test 2), the final BMD results are estimated from a
homogeneous variance model. If the test for homogeneity of variance is rejected (p< 0.1), the
model is run again while modeling the variance as a power function of the mean to account for
this nonhomogeneous variance. If this nonhomogeneous variance model does not adequately fit
the data (i.e., Test 3;p<0. 1), the data set is considered unsuitable for BMD modeling. In cases
in which a model with # parameters = # dose-groups was fit to the data set and all parameters
were estimated and no p-value was calculated, that model was not considered for estimation of a
POD unless no other model provided adequate fit. Among all models providing adequate fit, the
BMDL from the model with the lowest Akaike's information criterion (AIC) was selected as a
potential POD when BMDL values were sufficiently close (within threefold). Otherwise, the
lowest BMDL was selected as a potential POD from which to derive the oral reference
dose/inhalation reference concentration (RfD/RfC).
Modeling Predictions for Serum Total T4 in PND 1 Female Offspring (litter n)
The modeling results for total T4 in PND 1 female offspring (litter n) exposed gestation days
(GDs) 1-20 are shown in Table F-2. The Exponential 4 model (Figure F-2) was selected given
appropriate fit to the data and that the BMDL values differed by greater than threefold. The
output for the EPA's BMDS model run is also provided below.
A-l
-------
Table F-2. Modeling results for total T4 in PND 1 female offspring (litter n) exposed
GDs 1-20 a
Model
Global p-
value
AIC
BMDo.ssd
(HED)
(mg/kg-d)
BMDLo.ssd
(HED)
(mg/kg-d)
BMDisd
(HED)
(mg/kg-d)
BMDLisd
(HED)
(mg/kg-d)
Residual
of interest
Linear
0.5652
-4.74898
0.7778
0.5120
1.5557
1.0241
0.348
Polynomial
0.5652
-4.74898
0.7778
0.5120
1.5557
1.0241
0.348
Power
0.5652
-4.74898
0.7778
0.5120
1.5557
1.0241
0.348
Hill
-999
-1.89
0.368
0.0704
0.8677
0.2294
-6.01e-7
Exponential-
M2
0.77
-5.3672
0.5546
0.3017
1.2555
0.6694
-0.5752
Exponential-
M3
0.77
-5.3672
0.5546
0.3017
1.2555
0.6694
-0.5752
Exponential-
M4b
0.8583
-3.8581
0.3346
0.0951
0.8708
0.2498
-0.08305
Exponential-
ly
-999
-1.89
0.3807
0.0958
0.8669
0.2517
-4.356e-7
Notes'. BMD = maximum likelihood estimate of the exposure concentration associated with the selected BMR; BMDL = 95%
lower confidence limit on the BMD (subscripts denote BMR: i.e., 0.5 SD = exposure concentration associated with 0.5 SD
change from the control mean).
aFeng et al. ('2017').
b Selected model. Exponential 4 model was selected given appropriate fit to the data and that the BMDL values differed by
greater than threefold. The Hill and Exponential 5 models were not selected because they did not return a p-value.
A-2
-------
Tetraiodothyronine (T4), Total - Litter N
Dose (mg/kg-day HED)
Figure F-2. Exponential (Model 4) for total T4 in PND 1 female offspring (litter n) exposed
GDs 1-20 (Feng et al. (2017).
Exponential Model. (Version: 1.11; Date: 03/14/2017)
Input Data File: C:\Windows\TEMP\bmds-dfile-k4vsthrz.(d)
Gnuplot Plotting File:
Mon Aug 17 15:16:06 2020
BMDS Model Run
A-3
-------
The form of the response function by Model:
Model 2: Y[dose] = a * expjsign * b * dose}
Model 3: Y[dose] = a * exp{sign * (b * dose)Ad}
Model 4: Y[dose] = a * [c-(c-l) * exp{-b * dose}]
Model 5: Y[dose] = a * [c-(c-l) * exp{-(b * dose)Ad}]
Note: Y[dose] is the median response for exposure = dose;
sign = +1 for increasing trend in data;
sign = -1 for decreasing trend.
Model 2 is nested within Models 3 and 4.
Model 3 is nested within Model 5.
Model 4 is nested within Model 5.
Dependent variable = Response
Independent variable = Dose
Data are assumed to be distributed: normally
Variance Model: exp(lnalpha +rho *ln(Y[dose]))
rho is set to 0.
A constant variance model is fit.
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 500
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
MLE solution provided: Exact
A-4
-------
Initial Parameter Values
Variable	Model 4
Inalpha	-1.29725
rho	0 Specified
a	1.512
b	1.50054
c	0.434618
d	1 Specified
Parameter Estimates
Variable
Inalpha
a
b
c
Model 4
-1.29645
1.45283
1.10398
0.417162
Std. Err.
0.0611565
0.148029
1.13864
0.225239
NC = No Convergence
Table of Stats From Input Data
A-5
-------
Dose	N	Obs Mean Obs Std Dev
0	10	1.44	0.329
0.21	10	1.3	0.657
0.86	10	0.92	0.493
2.14	10	0.69	0.657
Estimated Values of Interest
Dose	Est Mean	Est Std Scaled Residual
0	1.453	0.523	-0.07759
0.21	1.278	0.523	0.1354
0.86	0.9337	0.523	-0.08305
2.14	0.6858	0.523	0.02529
Other models for which likelihoods are calculated:
Model Al:	Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)A2
Model A3:	Yij = Mu(i) + e(ij)
Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)
A-6
-------
Model R:	Yij = Mu + e(i)
Var{e(ij)} = SigmaA2
Likelihoods of Interest
Model	Log(likelihood)	DF	AIC
A1	5.944999	5	-1.889998
A2	8.698072	8	-1.396144
A3	5.944999	5	-1.889998
R	0.3138778	2	3.372244
4	5.929054	4	-3.858109
Additive constant for all log-likelihoods = -36.76. This constant added to th
e
above values gives the log-likelihood including the term that does not
depend on the model parameters.
Explanation of Tests
Test 1: Does response and/or variances differ among Dose levels? (A2 vs. R)
Test 2: Are Variances Homogeneous? (A2 vs. Al)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 6a: Does Model 4 fit the data? (A3 vs 4)
A-7
-------
Tests of Interest
Test	-2*log(Likelihood Ratio)	D. F.	p-value
Test 1	16.77	6	0.01017
Test 2	5.506	3	0.1383
Test 3	5.506	3	0.1383
Test 6a	0.03189	1	0.8583
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose
levels, it seems appropriate to model the data.
The p-value for Test 2 is greater than .1. A homogeneous
variance model appears to be appropriate here.
The p-value for Test 3 is greater than .1. The modeled
variance appears to be appropriate here.
The p-value for Test 6a is greater than .1. Model 4 seems
to adequately describe the data.
Benchmark Dose Computations:
Specified Effect = 1.000000
A-8
-------
Risk Type = Estimated standard deviations from control
Confidence Level = 0.950000
BMD =	0.87078
BMDL = 0.249811
BMDU =
21400
Exponential Model. (Version: 1.11; Date: 03/14/2017)
Input Data File: C:\Windows\TEMP\bmds-dfile-171ffb4f.(d)
Gnuplot Plotting File:
Mon Aug 17 15:16:07 2020
BHDS Model Run
The form of the response function by Model:
Model 2:
Model 3:
Model 4:
Model 5:
* dose}
(b * dose)Ad}
Y[dose] = a * exp{sign
Y[dose] = a * expjsign *
Y[dose] = a * [c-(c-l) * exp{-b * dose}]
Y[dose] = a * [c-(c-l) * exp{-(b * dose)Ad}]
Note: Yfdose] is the median response for exposure = dose;
A-9
-------
sign = +1 for increasing trend in data;
sign = -1 for decreasing trend.
Model 2 is nested within Models 3 and 4.
Model 3 is nested within Model 5.
Model 4 is nested within Model 5.
Dependent variable = Response
Independent variable = Dose
Data are assumed to be distributed: normally
Variance Model: exp(lnalpha +rho *ln(Y[dose]))
rho is set to 0.
A constant variance model is fit.
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 500
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
MLE solution provided: Exact
Initial Parameter Values
Variable	Model 4
Inalpha	-1.29725
A-10
-------
rho	8 Specified
a	1.512
b	1.50054
c	0.434618
d	1 Specified
Parameter Estimates
Variable	Model 4	Std. Err.
lnalpha	-1.29645	0.0611565
a	1.45283	0.148029
b	1.10398	1.13864
c	0.417162	0.225239
NC = No Convergence
Table of Stats From Input Data
Dose	N	Obs Mean Obs Std Dev
0	10	1.44	0.329
0.21	10	1.3	0.657
0.86	10	0.92	0.493
2.14	10	0.69	0.657
A-ll
-------
Dose
Estimated Values of Interest
Est Mean	Est Std Scaled Residual
0	1.453
0.21	1.278
0.86	0.9337
2.14	0.6858
0.523	-0.07759
0.523	0.1354
0.523	-0.08305
0.523	0.02529
Other models for which likelihoods are calculated:
Model Al:	Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2:	Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)A2
Model A3:	Yij = Mu(i) + e(ij)
Var{e(ij)} = exp(lalpha + log(mean(i)) * rho)
Model R:	Yij = Mu + e(i)
Var{e(ij)} = SigmaA2
Likelihoods of Interest
A-12
-------
Model	Log(likelihood)	DF	AIC
A1	5.944999	5	-1.889998
A2	8.698072	8	-1.396144
A3	5.944999	5	-1.889998
R	0.3138778	2	3.372244
4	5.929054	4	-3.858109
Additive constant for all log-likelihoods = -36.76. This constant added to th
e
above values gives the log-likelihood including the term that does not
depend on the model parameters.
Explanation of Tests
Test
1:
Does response and/or variances differ among Dose levels? (A2 vs. R)
Test
2:
Are Variances Homogeneous? (A2 vs. Al)
Test
3:
Are variances adequately modeled? (A2 vs. A3)
Test
6a:
Does Model 4 fit the data? (A3 vs 4)
Tests of Interest
Test	-2*log(Likelihood Ratio)	D. F.	p-value
Test 1	16.77	6	0.01017
A-13
-------
Test 2
5.506
3
0.1383
Test 3
5.506
3
0.1383
Test 6a
0.03189
1
0.8583
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose
levels, it seems appropriate to model the data.
The p-value for Test 2 is greater than .1. A homogeneous
variance model appears to be appropriate here.
The p-value for Test 3 is greater than .1. The modeled
variance appears to be appropriate here.
The p-value for Test 6a is greater than .1. Model 4 seems
to adequately describe the data.
Benchmark Dose Computations:
Specified Effect = 0.500000
Risk Type = Estimated standard deviations from control
Confidence Level = 0.950000
BMD =
0.33455
A-14
-------
ESMDL = 0.0950923
BMDU =	1.22544
F.3. Modeling Procedure for Dichotomous Noncancer Data
BMD modeling of dichotomous noncancer data (see Figure F-l) was conducted on the HAWC
website using the EPA's BMDS Version 2.7. For these data, the Gamma, Logistic, Log-Logistic,
Log-Probit, Multistage, Probit, and Weibull dichotomous models available within the software
were fit using a BMR of 10% extra risk. The Multistage model is run for all polynomial degrees
up to n - 2, where n is the number of dose groups including control. Adequacy of model fit was
judged based on the %2 goodness-of-fitp-value (p > 0.1), scaled residuals at the data point
(except the control) closest to the predefined BMR (absolute value < 2.0), and visual inspection
of the model fit. In the cases where no best model was found to fit to the data, a reduced data set
without the high-dose group was further attempted for modeling and the result was presented
along with that of the full data set. In cases in which a model with # parameters = # dose-groups
was fit to the data set and all parameters were estimated and no p-v alue was calculated, that
model was not considered for estimation of a POD unless no other model provided adequate fit.
Among all models providing adequate fit, the BMDL from the model with the lowest AIC was
selected as a potential POD when BMDL values were sufficiently close (within threefold) (see
Table F-l). Otherwise, the lowest BMDL was selected as a potential POD.
A-15
-------
Appendix G. Quality Assurance
EPA has an agency-wide quality assurance (QA) policy, and that policy is outlined in the EPA
Quality Manual for Environmental Programs (see CIO 2105-P-01-0) and follows the
specifications outlined in EPA Order CIO 2105.0. The goal of the QA policy is to assure that
environmental data used to support Agency decisions are of adequate quality and usability for
their intended purpose.
As required by CIO 2105.0, ORD maintains a Quality Management Program, which is
documented in an internal Quality Management Plan (QMP). The latest version was developed
in 2013 and was developed using Guidance for Developing Quality Systems for Environmental
Programs (QA/G-1). An NCEA-specific QMP was also developed in 2013 as an appendix to the
ORD QMP. Quality Assurance for products developed within CPHEA is managed under the
ORD QMP and applicable appendices.
This assessment has been designated as High Profile and is classified as QA Category A.
Category A designations require reporting of all critical QA activities, including audits.
Another requirement of the Agency quality system includes the use of project-specific planning
documents referred to as Quality Assurance Project Plans (QAPPs) that describe how specific
data collection efforts will be planned, implemented, and assessed. Specific management of
quality assurance in this assessment is documented in an Umbrella Quality Assurance Project
Plan, which was developed using the EPA Guidance for Quality Assurance Project Plans
( * ^ . ). The latest approved version of the QAPP is dated September 2019. During
assessment development, additional QAPPs may be applied for quality assurance management.
They include:
Title
Document Number
Date
Program Quality Assurance Project Plan
(PQAPP) for the Provisional Peer-Reviewed
Toxicity Values (PPRTVs) and Related
Assessments/Documents
L-CPAD-0032718-QP
October 2015 (last
updated 2020)
Umbrella Quality Assurance Project Plan for
NCEA PFAS Toxicity Assessments
B-IO-0031652-QP-1 -2
July 2018 (last
updated September
2019)
Quality Assurance Project Plan (QAPP) for
Enhancements to Benchmark Dose Software
(BMDS)
B-003742-QP-1-0
July 2019
During assessment development, this project underwent quality audit:
Date
Type of audit
Major findings
Actions taken
September 18, 2020
Technical System
Audit
None
None
A-16
-------
During assessment development, the assessment was subjected to external reviews by individual
letters from expert peer reviewers and by other federal agency partners including the Executive
Offices of the President. Peer review reports during these review steps are available at
https://www.epa.gov/pfas/genx-and-pfbs-draft-toxicitv-assessments. In addition, the assessment
underwent public comment from November 21, 2018 to January 22, 2019. The public comments
are available in the Docket ID No. EPA-HQ-OW-2018-0614. Prior to release, the final draft
assessment was submitted to management and QA clearance. During this step the CPHEA QA
Director and QA Managers review the project QA documentation and ensure EPA QA
requirements have been met.
A-17
-------
Appendix H. References
3M (3M Company). (2000a). Acute dermal irritation study of T-7485 applied to New Zealand
white rabbits. (Study Number: 132-004). St. Paul, MN: 3M Corporate Toxicology.
3M (3M Company). (2000b). Acute dermal toxicity study of T-7485 applied to Sprague-Dawley
rats. (Study Number: 132-010). St. Paul, MN: 3M Corporate Toxicology.
3M (3M Company). (2000c). Acute ocular irritation study of T-7485 applied to New Zealand
white rabbits. (Study Number: 132-005). St. Paul, MN: 3M Corporate Toxicology.
3M (3M Company). (2000d). A repeated dose range-finding toxicity study of T-7485 in
Sprague-Dawley rats. (Study Number: 132-006). St. Paul, MN: 3M Pharmaceuticals.
3M (3M Company). (2001). A 28-day oral (gavage) toxicity study of T-7485 in Sprague-Dawley
rats. (Study Number: 132-007). St. Paul, MN: 3M Corporate Toxicology.
3M (3M Company). (2002a). Delayed contact hypersensitivity study of T-7485 in Hartley guinea
pigs (maximization test). (Study Number: 132-015). St. Paul, MN: 3M Corporate
Toxicology.
3M (3M Company). (2002b). Environmental, health, safety, and regulatory (EHSR) profile of
perfluorobutane sulfonate (PFBS). :
http://multimedia.3m.com/mws/media/172303O/ehsr-profile-of-perfluorobutane-
sulfonate-pfb s .pdf
3M (3M Company). (2010). TSCA 8(e) substantial risk notice: Sulfonate-based and carboxylic-
based fluorochemicals, docket 8EHQ-0598-373 - Results from a mechanistic
investigation of the effect of PFBS, PFHS, and PFOS on lipid and lipoprotein metabolism
in transgenic mice [TSCA Submission], (8EHQ-10-00373DH). St. Paul, MN.
Aerostar SES LLC. (2017). Final site inspection report of fire fighting foam usage at Dover Air
Force Base, Kent County, Delaware.
Alexander. EK; Pearce. EN: Brent GA; Brown. RS: Chen. Ft; Dosiou. C: Grobman. W:
Laurberg, P; Lazarus. JH: Mandel. SJ: 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.
http://dx.doi.org/10.1089/thy.2016.0457
Andersen, SL; Andersen, S; Liew, Z; Vestergaard, P; Olsen, J. (2018). Maternal Thyroid
Function in Early Pregnancy and Neuropsychological Performance of the Child at 5
Years of Age. J Clin Endocrinol Metab 103: 660-670. http://dx.doi.org/10.1210/ic.2Q17-
02171
A-18
-------
Anderson, RH; Long, GC; Porter, RC; Anderson, JK. (2016). Occurrence of select perfluoroalkyl
substances at U.S. Air Force aqueous film-forming foam release sites other than fire-
training areas: Field-validation of critical fate and transport properties. Chemosphere 150:
678-685. http://dx.doi.Org/10.1016/i.chemosphere.2016.01.014
Anonymous. (2017). Final technical memorandum proposed groundwater monitoring well
locations perfluorinated compound delineation, areas of concern perfluorinated
compound-1. Ellsworth Air Force Base, South Dakota. Anonymous.
ASTSWMO (Association of State and Territorial Solid Waste Management Officials). (2015).
Perfluorinated chemicals (PFCS): Perfluorooctanoic acid (PFOA) & perfluorooctane
sulfonate (PFOS) information paper.
AT SDR (Agency for Toxic Substances and Disease Registry). (2009). Draft toxicological profile
for perfluoroalkyls. Agency for Toxic Substances and Disease Regristry.
http://www.atsdr.cdc.gov/toxprofiles/tp200.pdf
AT SDR (Agency for Toxic Substances and Disease Registry). (2015). Draft toxicological profile
for perfluoroalkyls. https://www.atsdr.cdc.gov/toxprofiles/tp200.pdf
Bao, J: Lee, YL; Chen, PC: Jin, YH; Dong, GH. (2014). Perfluoroalkyl acids in blood serum
samples from children in Taiwan. Environ Sci Pollut Res Int 21: 7650-7655.
http://dx.doi.org/10.1007/sll356-014-2594-4
Bao, WW: Qian, ZM; Geiger, SD; Liu, E; Liu, Y; Wang, SO: Lawrence, WR; Yang, BY; Hu,
LW: Zeng, XW: Dong, GH. (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. or g/10.1016/i. scitotenv .2017.07.124
Berk, M; Williams, LJ; Andreazza, A: Pasco, JA; Dodd, S: Jacka, FN: Moylan, S: Reiner, EJ;
Magalhaes, PVS. (2014). Pop, heavy metal and the blues: secondary analysis of persistent
organic pollutants (POP), heavy metals and depressive symptoms in the NHANES
National Epidemiological Survey. BMJ Open 4: e005142.
http ://dx. doi. org/10.113 6/bmi open-2014-005142
Bijland, S; Rensen, PCN; Pieterman, EJ: Maas, ACE; van der Hoorn, JW; van Erk, MJ; Havekes,
LM; Willems van Dijk, K; Chang, SC: Ehresman, DJ; Butenhoff, JL; Princen, HMG.
(2011). Perfluoroalkyl sulfonates cause alkyl chain length-dependent hepatic steatosis
and hypolipidemia mainly by impairing lipoprotein production in APOE*3-Leiden CETP
mice. Toxicol Sci 123: 290-303. http://dx.doi.org/10.1093/toxsci/kfrl42
Bogdanska, J: Sundstrom, M; Bergstrom, U; Borg, D; Abedi-Valugerdi, M; Bergman, A:
Depierre, J: Nobel, S. (2014). Tissue distribution of S-35-labelled
perfluorobutanesulfonic acid in adult mice following dietary exposure for 1-5 days.
Chemosphere 98: 28-36. http://dx.doi.org/10.1016/j.chemosphere.2013.09.062
A-19
-------
Bomhard, E; Loser, E. (1996). Acute toxicologic evaluation of perfluorobutanesulfonic acid. J
Am Coll Toxicol B Acute Toxic Data 15: SI05.
http://dx.doi.org/10.1177/1091581896015QS175
Calvo, R; Obregon, MJ; 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.
Casey. BM; Thom. EA; Peaceman. AM: Varner. MW; Sorokin. Y; Hirtz. DG: Reddv. UM;
Wapner. RJ; Thorp. JM; Saade. G: Tita. AT: Rouse. DJ: Sibai. B; lams. JD; Mercer. BM;
Tolosa. J: Caritis. SN: Vandorsten. JP. (2017). Treatment of subclinical hypothyroidism
or hypothyroxinemia in pregnancy. N Engl J Med 376: 815-825.
http://dx.doi.org/10.1056/NEJMoal6062Q5
CDC (Centers for Disease Control and Prevention). (2009). Fourth national report on human
exposure to environmental chemicals. Atlanta, GA: U.S. Department of Health and
Human Services, Centers for Disease Control and Prevention.
http://www.cdc.gov/exposurereport/
CDC (Centers for Disease Control and Prevention). (2018). National report on human exposure
to environmental chemicals. Updated tables, March 2018. Atlanta, GA: U.S. Department
of Health and Human Services, https://www.cdc.gov/exposurereport/
Chen. Q: Huang. R; Hua. L; Guo. Y; Huang. L; Zhao. Y; Wang. X: Zhang. J. (2018). Prenatal
exposure to perfluoroalkyl and polyfluoroalkyl substances and childhood atopic
dermatitis: A prospective birth cohort study. Environ Health 17: 1-12.
http://dx.doi.org/10.1186/sl2940-018-0352-7
Chen. Q: Zhang. X: Zhao. Y; Lu. W: Wu. J: Zhao. S: Zhang. J: Huang. L. (2019). Prenatal
exposure to perfluorobutanesulfonic acid and childhood adiposity: A prospective birth
cohort study in Shanghai, China. Chemosphere 226: 17-23.
http://dx.doi. org/10.1016/i. chemosphere.2019.03.095
Chengelis, CP: Kirkpatrick, JB; Myers, NR; Shinohara, M; Stetson, PL; Sved, DW. (2009).
Comparison of the toxicokinetic behavior of perfluorohexanoic acid (PFHxA) and
nonafluorobutane-1-sulfonic acid (PFBS) in cynomolgus monkeys and rats. Reprod
Toxicol 27: 400-406. http://dx.doi.Org/10.1016/i.reprotox.2009.01.013
Choksi, NY: Jahnke, GD; St Hilaire, C: Shelby, M. (2003). Role of thyroid hormones in human
and laboratory animal reproductive health [Review], Birth Defects Res B Dev Reprod
Toxicol 68: 479491. http://dx.doi.org/10.1002/bdrb. 10045
Corsini, E; Sangiovanni, E; Avogadro, A: Galbiati, V: Viviani, B; Marinovich, M; Galli, CL;
Dell'Agli, M; Germolec, DR. (2012). In vitro characterization of the immunotoxic
potential of several perfluorinated compounds (PFCs). Toxicol Appl Pharmacol 258:
248-255. http://dx.doi.Org/10.1016/i.taap.2011.ll.004
A-20
-------
Crissman, JW; Goodman, DG; Hildebrandt PK; Maronpot, RR: Prater, DA; Riley, JH; Seaman,
WJ; Thake, DC. (2004). Best practices guideline: Toxicologic histopathology. Toxicol
Pathol 32: 126-131. http://dx.doi.org/10.1080/0192623049Q268756
Dewitt, JC: Peden-Adams, MM: Keller, JM; Germolec, DR. (2012). Immunotoxicity of
perfluorinated compounds: recent developments [Review], Toxicol Pathol 40: 300-311.
http://dx.doi.org/10.1177/0192623311428473
Dong. GH: Tung. KY: Tsai. CH: Liu. MM: Wang. D: Liu. W: Jin. YH; Hsieh. WS: Lee. YL;
Chen. PC. (2013a). Serum polyfluoroalkyl concentrations, asthma outcomes, and
immunological markers in a case-control study of Taiwanese children. Environ Health
Perspect 121: 507-513, 513e501-508. http://dx.doi.org/10.1289/ehp.1205351
Dong. GH: Tung. KY: Tsai. CH: Liu. MM: Wang. D: Liu. W: Jin. YH: Hsieh. WS: Lee. YL:
Chen. PC. (2013b). Serum polyfluoroalkyl concentrations, asthma outcomes, and
immunological markers in a case-control study of Taiwanese children: Supplemental
materials [Supplemental Data], Environ Health Perspect 121: 1-8.
EFSA (European Food Safety Authority). (2012). Perfluoroalkylated substances in food:
Occurrence and dietary exposure. EFSA J 10: 2743.
Eriksen, KT; Raaschou-Nielsen, O; S0rensen, M; Roursgaard, M; Loft, S; M0ller, P. (2010).
Genotoxic potential of the perfluorinated chemicals PFOA, PFOS, PFBS, PFNA and
PFHxA in human HepG2 cells. Mutat Res 700: 39-43.
http: //dx. doi. or g/10.1016/i. mr gentox .2010.04.024
ESTCP (Environmental Security Technology Certification Program). (2017). FAQs regarding
PFASs associated with AFFF use at U.S. military sites.
http://www.dtic.mil/dtic/tr/fulltext/u2/1044126.pdf
Fabrega, F; Kumar, V: Benfenati, E; Schuhmacher, M; Domingo, JL; Nadal, M. (2015).
Physiologically based pharmacokinetic modeling of perfluoroalkyl substances in the
human body. Toxicol Environ Chem 97: 814-827.
http ://dx. doi .org/10.1080/02772248.2015.1060976
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, MJ; van Eiisden, M; Loomans, EM; Vriikotte, TG: Rotteveel, J. (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
Fisher, DA. (1997). Fetal thyroid function: Diagnosis and management of fetal thyroid disorders.
Clin Obstet Gynecol 40: 16-31.
A-21
-------
Fisher, DA; Klein, AH. (1981). Thyroid development and disorders of thyroid function in the
newborn [Review], N Engl J Med 304: 702-712.
http://dx.doi.org/10.1056/NEJM1981031930412Q5
Forhead, AJ; Fowden, AL. (2014). Thyroid hormones in fetal growth and prepartum maturation
[Review], J Endocrinol 221: R87-R103. http://dx.doi.org/10.1530/JQE-14-0025
Foster, P; Gray, LE, Jr. (2013). Toxic responses of the reproductive system. In CD Klaassen
(Ed.), Casarett & Doull's toxicology: The basic science of poisons (8th ed., pp. 861-906).
New York, NY: McGraw-Hill Education.
https://www.mheducation.com/highered/product/casarett-doull-s-toxicology-basic-
science-poisons-9th-edition-klaassen/9781259863 745.html
Fraser, AJ: Webster, TF; Watkins, DJ: Strynar, MJ; Kato, K; Calafat, AM: Vieira, VM; Mcclean,
MP. (2013). Polyfluorinated compounds in dust from homes, offices, and vehicles as
predictors of concentrations in office workers' serum. Environ Int 60: 128-136.
http://dx.doi.Org/10.1016/i.envint.2013.08.012
Ghassabian, A; Trasande, L. (2018). Disruption in thyroid signaling pathway: A mechanism for
the effect of endocrine-disrupting chemicals on child neurodevelopment. Front
Endocrinol (Lausanne) 9: 209. http://dx.doi.org/doi: 10.3389/fendo.2018.00204.
eCollection 2018
Gilbert, ME. (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, ME: Rovet, J: Chen, Z; Koibuchi, N. (2012). Developmental thyroid hormone
disruption: prevalence, environmental contaminants and neurodevelopmental
consequences. Neurotoxicology 33: 842-852.
http://dx.doi.Org/10.1016/i.neuro.2011.11.005
Gilbert, ME: Sanchez-Huerta, K; Wood, C. (2016). Mild thyroid hormone insufficiency during
development compromises activity-dependent neuroplasticity in the hippocampus of
adult male rats. Endocrinology 157: 774-787. http://dx.doi.org/10.1210/en.2015-1643
Gilbert, ME; Zoeller, RT. (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.
Glynn, A; Berger, U; Bignert, A; Ullah, S; Aune, M; Lignell, S; Darnerud, PO. (2012).
Perfluorinated alkyl acids in blood serum from primiparous women in Sweden: serial
sampling during pregnancy and nursing, and temporal trends 1996-2010. Environ Sci
Technol 46: 9071-9079. http://dx.doi.org/10.1021/es301168c
Grandiean, P. (2018). Delayed discovery, dissemination, and decisions on intervention in
environmental health: a case study on immunotoxicity of perfluorinated alkylate
A-22
-------
substances [Editorial], Environ Health 17: 62. http://dx.doi.org/10.1186/sl2940-Q18-
0405-v
Gyllenhammar, I; Diderholm, B; Gustafsson, J; Berger, U; Ridefelt, P; Benskin, JP; Lignell, S;
Lampa, E: Glynn. A. (2018). Perfluoroalkyl acid levels in first-time mothers in relation to
offspring weight gain and growth. Environ Int 111: 191-199.
http://dx.doi.Org/10.1016/i.envint.2017.12.002
Haddow. JE; Palomaki. GE: Allan. WC: Williams. JR; Knight GJ: Gagnon. J: O'Heir. CE:
Mitchell. ML: Hermos. RJ: Waisbren. SE; Faix. JD; Klein. RZ. (1999). Maternal thyroid
deficiency during pregnancy and subsequent neuropsychological development of the
child. N Engl J Med 341: 549-555. http://dx.doi.org/10.1056/NEJM19990819341Q801
Hales. C: Taylor. PN: Channon. S: Paradice. R; Mcewan. K; Zhang. L; Gvedu. M; Bakhsh. A:
Okosieme. O: Muller. I; Draman. MS: Gregory. JW: Davan. C: Lazarus. JH; Rees. DA:
Ludgate, M. (2018). Controlled Antenatal Thyroid Screening II: Effect of Treating
Maternal Suboptimal Thyroid Function on Child Cognition. J Clin Endocrinol Metab
103: 1583-1591. http://dx.doi.org/10.1210/ic.2017-02378
Henrichs. J: Bongers-Schokking. JJ; Schenk. JJ; Ghassabian. A: Schmidt. HG: Visser. TJ;
Hooiikaas. H; de Muinck Keizer-Schrama. SM; Hofman. A: Jaddoe. VV: Visser. W:
Steegers. EA; Verhulst. FC: de Riike. YB; 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/10.1210/ic.2010-Q415
Hernandez. A. Fiering. S.; Martinez. E; Galton. VA; St Germain. D. (2002). The gene locus
encoding iodothyronine deiodinase type 3 (Dio3) is imprinted in the fetus and expresses
antisense transcripts. Endocrinology 134: 4483-4486.
Hill, AB. (1965). The environment and disease: Association or causation? Proc R Soc Med 58:
295-300.
Hillman, NH; Kallapur, SG; Jobe, AH. (2012). Physiology of transition from intrauterine to
extrauterine life. Clin Perinatol 39: 769-783.
Houtz, EF; Sutton, R; Park, JS; Sedlak, M. (2016). Poly- and perfluoroalkyl substances in
wastewater: Significance of unknown precursors, manufacturing shifts, and likely AFFF
impacts. Water Res 95: 142-149. http://dx.doi.Org/10.1016/i.watres.2016.02.055
Hu, XC: Andrews, DO: Lindstrom, AB: Bruton, TA; Schaider, LA: Grandiean, P; Lohmann, R;
Carignan, CC: Blum, A: Balan, SA; Higgins, CP: Sunderland, EM. (2016). Detection of
poly- and perfluoroalkyl substances (PFASs) in U.S. drinking water linked to industrial
sites, military fire training areas, and wastewater treatment plants. Environ Sci Technol
Lett 3: 344-350. http://dx.doi.org/10.1021/acs.estlett.6b00260
A-23
-------
Huang, M; Jiao, J: Zhuang, P; Chen, X; Wang, J; Zhang, Y. (2018). Serum polyfluoroalkyl
chemicals are associated with risk of cardiovascular diseases in national US population.
Environ Int 119: 37-46. http://dx.doi.Org/10.1016/i.envint.2018.05.051
Huang, MC: Dzierlenga, AL; Robinson, VG: Waidvanatha, S: Devito, MJ; Eifrid, MA:
Granville, CA; Gibbs, ST: Blystone, CR. (2019a). Toxicokinetics of perfluorobutane
sulfonate (PFBS), perfluorohexane-l-sulphonic acid (PFHxS), and perfluorooctane
sulfonic acid (PFOS) in male and female Hsd:Sprague Dawley SD rats after intravenous
and gavage administration. Toxicology Reports 6: 645-655.
http://dx.doi.Org/10.1016/i.toxrep.2019.06.016
Huang, R: Chen, O: Zhang, L: Luo, K: Chen, L: Zhao, S: Feng, L: Zhang, J. (2019b). Prenatal
exposure to perfluoroalkyl and polyfluoroalkyl substances and the risk of hypertensive
disorders of pregnancy. Environ Health 18:5. http://dx.doi.Org/10.l 186/sl2940-018-
0445-3
Hulbert, AJ. (2000). Thyroid hormones and their effects: A new perspective [Review], Biol Rev
Camb Philos Soc 75: 519-631.
Julvez, J: Alvarez-Pedrerol, M; Rebagliato, M; Murcia, M; Forns, J: Garcia-Esteban, R;
Lertxundi, N: Espada, M; Tardon, A: Riano Galan, I; Sunyer, J. (2013). Thyroxine levels
during pregnancy in healthy women and early child neurodevelopment. Epidemiology
24: 150-157. http://dx.doi.org/10.1097/EDE.0b013e318276ccd3
Kato, K; Calafat, AM: Needham, LL. (2009). Polyfluoroalkyl chemicals in house dust. Environ
Res 109: 518-523. http://dx.doi.Org/10.1016/i.envres.2009.01.005
Kim, D; Kim, U, nJ; Kim, H; Choi, SD: Oh, J. (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.Org/10.1016/i.envres.2016.02.037
Kim, SJ: Choi, EJ; Choi, GW: Lee, YB; Cho, HY. (2019). Exploring sex differences in human
health risk assessment for PFNA and PFDA using a PBPK model. Arch Toxicol 93: 311-
330. http://dx.doi.org/10.1007/sQ0204-018-2365-v
Klein, AH; Oddie, TH; Parslow, M; Foley, TP; Fisher, DA. (1982). Developmental changes in
pituitary-thyroid function in the human fetus and newborn. Early Hum Dev 6: 321-330.
http://dx.doi.org/10.1016/0378-3782(82N)90070-6
Kooistra, L; Crawford, S: van Baar, AL: Brouwers, EP; Pop, VJ. (2006). Neonatal effects of
maternal hypothyroxinemia during early pregnancy. Pediatrics 117: 161-167.
http://dx.doi.org/10.1542/peds.2005-0227
Koopdonk-Kool, JM; de Vijlder, JJ; Veenboer, GJ: Ris-Stalpers, C: Kok, JH; Vulsma, T; Boer,
K; Visser, TJ. (1996). Type II and type III deiodinase activity in human placenta as a
function of gestational age. J Clin Endocrinol Metab 81: 2154-2158.
A-24
-------
Korevaar, TIM; Muetzel, R; Medici, M; Chaker, L; Jaddoe, VWV; de Rijke, YB; Steegers, EAP;
Visser. TJ; White. T; Tiemeier. H; Peeters. RP. (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/S2213-8587d5X)0327-7
Lasier. PJ; Washington. JW: Hassan. SM; Jenkins. TM. (2011). Perfluorinated chemicals in
surface waters and sediments from northwest Georgia, USA, and their bioaccumulation
in Lumbriculus variegatus. Environ Toxicol Chem 30: 2194-2201.
http: //dx. doi. or g/10.1002/etc. 622
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; Rumpler, J; Das, KP; Wood, CR; Schmid, JE; Strynar, MJ; Wambaugh, JF. (2020).
Pharmacokinetic profile of Perfluorobutane Sulfonate and activation of hepatic nuclear
receptor target genes in mice. Toxicology 441: 152522.
http://dx.doi.Org/10.1016/i.tox.2020.152522
Lavado-Autric, R; Auso, E; Garcia-Velasco, JV: 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/JCI16262
Lazarus, JH; Bestwick, JP; Channon, S: Paradice, R; Maina, A: Rees, R; Chiusano, E; John, R;
Guaraldo, V: George, LM; Perona, M; DaH'Amico, D; Parkes, AB; Joomun, M; Wald,
NJ. (2012). Antenatal thyroid screening and childhood cognitive function. N Engl J Med
366: 493-501. http://dx.doi.org/10.1056/NEJMoallQ6104
Levie, D; Korevaar, TIM; Bath, SC; Dalmau-Bueno, A; Murcia, M; Espada, M; Dineva, M;
Ibarluzea, JM; Sunyer, J; Tiemeier, H; Rebagliato, M; Ravman, MP; Peeters, RP;
Guxens, M. (2018). Thyroid Function in Early Pregnancy, Child IQ, and Autistic Traits:
A Meta-Analysis of Individual Participant Data. J Clin Endocrinol Metab 103: 2967-
2979. http://dx.doi.org/10.1210/ic.2018-0Q224
Li, Y; Fletcher, T; Mucs, D; Scott, K; Lindh, CH; Tallying, P; Jakobsson, K. (2018). Half-lives
of PFOS, PFHxS and PFOA after end of exposure to contaminated drinking water. Occup
Environ Med 75: 46-51. http://dx.doi.org/10.1136/oemed-2017-104651
Lieder, PH; Chang, SC; York, RG; Butenhoff, JL. (2009a). Toxicological evaluation of
potassium perfluorobutanesulfonate in a 90-day oral gavage study with Sprague-Dawley
rats. Toxicology 255: 45-52. http://dx.doi.Org/10.1016/i.tox.2008.10.002
Lieder, PH; York, RG; Hakes, DC; Chang, SC; Butenhoff, JL. (2009b). A two-generation oral
gavage reproduction study with potassium perfluorobutanesulfonate (K+PFBS) in
Sprague Dawley rats. Toxicology 259: 33-45. http://dx.doi.Org/10.1016/i.tox.2009.01.027
A-25
-------
Liew, Z; Goudarzi, H; Oulhote, Y. (2018). Developmental exposures to perfluoroalkyl
substances (PFASs): An update of associated health outcomes [Review], Curr Environ
Health Rep 5: 1-19. http://dx.doi.org/10.1007/s40572-018-Q173-4
Liu. X; Guo. Z; Krebs. KA; Pope. RH; Roache. NF. (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
Mangelsdorf. I; Buschmann. J: Orthen. B. (2003). Some aspects relating to the evaluation of the
effects of chemicals on male fertility [Review], Regul Toxicol Pharmacol 37: 356-369.
http://dx.doi.org/10.1016/S0273-2300(03N)00026-6
Manzano-Salgado. CB; Casas. M; Lopez-Espinosa. MJ; Ballester. F; Basterrechea. M; Grimalt.
JO: Jimenez. AM: Kraus. T; Schettgen. T; Sunyer. J: Vrijheid. M. (2015). Transfer of
perfluoroalkyl substances from mother to fetus in a Spanish birth cohort. Environ Res
142: 471-478. http://dx.doi.Org/10.1016/i.envres.2015.07.020
MDCH (Michigan Department of Community Health). (2015). Letter health consultation:
Evaluation of drinking water near Wurtsmith Air Force Base, Oscoda, Iosco County,
Michigan [Personal Communication], Atlanta, GA: U.S. Department of Health and
Human Services, Agency for Toxic Substances and Disease.
https://www.atsdr.cdc.gov/HAC/pha/WurtsmithAFB/WurtsmithAFB H2Q %20LHC 09
-08-2015 508.pdf
MDH (Minnesota Department of Health). (2017). Toxicological summary for: Perfluorobutane
sulfonate. Minnesota Department of Health, Health Based Guidance for Water Health
Risk Assessment Unit, Environmental Health Division.
http://www.health.state.mn.us/divs/eh/risk/guidance/gw/pfbssummary.pdf
Min. H: Dong. J: Wang. Y: Wang. Y: Teng. W: Xi. O: Chen. J. (2016). Maternal
hypothyroxinemia-induced neurodevelopmental impairments in the progeny [Review],
Mol Neurobiol 53: 1613-1624. http://dx.doi.org/10.1007/sl2035-Q15-9101-x
Mirabella, G; Feig, D; Astzalos, E; Perlman, K; Rovet, JF. (2000). The effect of abnormal
intrauterine thyroid hormone economies on infant cognitive abilities. J Pediatr Endocrinol
Metab 13: 191-194.
Nakavama. S: Strynar. MJ: Helfant. L; Egeghy. P; Ye. X: Lindstrom. AB. (2007). Perfluorinated
compounds in the Cape Fear Drainage Basin in North Carolina. Environ Sci Technol 41:
5271-5276. http://dx.doi.org/10.1021/es070792v
Nakavama. SF; Strynar. MJ: Reiner. JL; Delinsky. AD: Lindstrom. AB. (2010). Determination
of perfluorinated compounds in the upper Mississippi river basin. Environ Sci Technol
44: 4103-4109. http://dx.doi.org/10.1021/esl00382z
NASEM (National Academies of Sciences, Engineering, and Medicine). (2017). Application of
systematic review methods in an overall strategy for evaluating low-dose toxicity from
A-26
-------
endocrine active chemicals. Washington, D.C.: The National Academies Press.
http://dx.doi.org/10.17226/24758
Negro, R; Soldin, OP; Obregon, MJ; Stagnaro-Green, A. (2011). Hypothyroxinemia and
pregnancy [Review], EndocrPract 17: 422-429. http://dx.doi.org/10.4158/EP10309.RA
NICNAS (National Industrial Chemicals Notification and Assessment Scheme). (2005). Existing
chemical hazard assessment report. Potassium perfluorbutane sulfonate. Canberra,
Australia: Commonwealth of Australia, Australian Government Publishing.
http://www.nicnas.gov.au/ data/assets/pdf file/0004/4927/Potassium Perfluorobutane
Sulfonate PDF.pdf
NRC (National Research Council). (1983). Risk assessment in the federal government:
Managing the process. Washington, DC: National Academy Press.
http://dx.doi.org/10.17226/366
NTP (National Toxicology Program). (2005). Perfluorobutane sulfonate. Genetic toxicology -
bacterial mutagenicity. Chemical effects in biological systems (CEBS). (Study ID
A32037). Research Triangle Park, NC: National Institutes of Environmental Sciences,
National Institutes of Health.
http://tools.niehs.nih.gov/cebs3/ntpViews/?activeTab=detail&studvNumber=A32037
NTP (National Toxicology Program). (2011). NTP Protocol Outline of the Class Study of
Perfluorinated Chemicals [perfluorobutane sulfonate (PFBS), perfluorohexane sulfonate
potassium salt (PFHSKslt), perfluorooctane sulfonate (PFOS), perfluorohexanoic acid
(PFHxA), perfluorooctanoic acid (PFOA), perfluorononanoic acid (PFNA),
perfluorodecanoic acid (PFDA) and WY 14643 (WY)] in Harlan Sprague Dawley rats
administered daily by gavage for 28-Days [NTP],
NTP (National Toxicology Program). (2012). Perfluorobutane sulfonate. Genetic toxicology -
micronucleus [NTP], Research Triangle Park, NC: U.S. Department of Health and
Human Services, National Institutes of Environmental Sciences, National Institutes of
Health, https://tools.niehs.nih.gov/cebs3/ntpViews/?studvNumber=002-01134-0002-
0000-3
NTP (National Toxicology Program). (2015). Protocol to evaluate the evidence for an
association between perfluorooctanoic acid (PFOA) or perfluorooctane sulfonate (PFOS)
exposure and immunotoxicity. Department of Health and Human Services, Office of
Health Assessment and Translation.
https://ntp.niehs.nih.gov/ntp/ohat/pfoa pfos/protocol 201506 508.pdf
NTP (National Toxicology Program). (2016). NTP Monograph: Immunotoxicity associated with
exposure to perfluorooctanoic acid (PFOA) or perfluorooctane sulfonate (PFOS) [NTP],
Research Triangle Park, NC: U.S. Department of Health and Human Services, Office of
Health Assessment and Translation.
https://ntp.niehs.nih.gov/ntp/ohat/pfoa pfos/pfoa pfosmonograph 508.pdf
A-27
-------
NTP (National Toxicology Program). (2019). NTP technical report on the toxicity studies of
perfluoroalkyl sulfonates (perfluorobutane sulfonic acid, perfluorohexane sulfonate
potassium salt, and perfluorooctane sulfonic acid) administered by gavage to Sprague
Dawley (Hsd: Sprague Dawley SD) rats. (Toxicity Report 96). Research Triangle Park,
NC. https://ntp.niehs.nih.gov/ntp/htdocs/st rpts/tox096 508.pdf
01 sen. GW; Chang. SC; Noker. PE; Gorman. GS; Ehresman. DJ; Lieder. PH; Butenhoff JL.
(2009). A comparison of the pharmacokinetics of perfluorobutanesulfonate (PFBS) in
rats, monkeys, and humans. Toxicology 256: 65-74.
http://dx.doi.Org/10.1016/i.tox.2008.l 1.008
Olsen. GW: Mair. DC: Lange. CC: Harrington. LM; Church. TR; Goldberg. CL; Herron. RM;
Hanna, H; Nobiletti, JB; Rios, JA; Reagen, WK; Ley, CA. (2017). Per- and
polyfluoroalkyl substances (PFAS) in American Red Cross adult blood donors, 2000-
2015. Environ Res 157: 87-95. http://dx.doi.Org/10.1016/i.envres.2017.05.013
Pant. KJ. (2001). Evaluation of a test article in the Salmonella typhimurium/Escherichia coli
plate incorporation/preincubation mutation assay in the presence and absence of induced
rat liver S-9. (Sponsor's study no. 132-008). Redfield, AR: Primedica Redfield.
Patel. J: Landers. K; Li. H; Mortimer. RH; Richard. K. (2011). Thyroid hormones and fetal
neurological development. J Endocrinol 209: 1-8. http://dx.doi.org/10.153O/JOE-10-0444
Pop. VJ: Brouwers. EP; Vader. HL; Vulsma. T; van Baar. AL; de Viilder. JJ. (2003). Maternal
hypothyroxinaemia during early pregnancy and subsequent child development: a 3-year
follow-up study. Clin Endocrinol 59: 282-288. http://dx.doi.org/10.1046/i. 1365-
2265.2003.01822.x
Post. GB; Louis. JB: Lippincott. RL; Procopio. NA. (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
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.
Oin. XD; Qian. Z; Vaughn. MG: Huang. J: Ward. P; Zeng. XW; Zhou. Y; Zhu. Y; Yuan. P; Li.
M: Bai. Z; Paul. G: Hao. YT; Chen. W: Chen. PC: Dong. GH; Lee. YL. (2016). Positive
associations of serum perfluoroalkyl substances with uric acid and hyperuricemia in
children from Taiwan. Environ Pollut 212: 519-524.
http://dx.doi.Org/10.1016/i.envpol.2016.02.050
Oin. XD: Qian. ZM; Dharmage. SC: Perret. J: Geiger. SD: Rigdon. SE; Howard. S: Zeng. XW:
Hu. LW: Yang. BY: Zhou. Y: Li. M: Xu. SL: Bao. WW: Zhang. YZ: Yuan. P: Wang. J:
Zhang. C: Tian. YP: Nian. M: Xiao. X: Chen. W: Lee. YL: Dong. GH. (2017).
Association of perfluoroalkyl substances exposure with impaired lung function in
children. Environ Res 155: 15-21. http://dx.doi.Org/10.1016/i.envres.2017.01.025
A-28
-------
Rabah, SA; Gowan, IL; Pagnin, M; Osman, N; Richardson, SJ. (2019). Thyroid Hormone
Distributor Proteins During Development in Vertebrates [Review], Front Endocrinol
(Lausanne) 10: 506. http://dx.doi.org/10.3389/fendo.2019.005Q6
Regulatory Science Associates. (2018). A literature review of the current state of the science
regarding species differences in the control of, and response to, thyroid hormone
perturbations. Part 1: A human health perspective. (RSA/ECPAOOI THYROID).
Brussels, Belgium: European Crop Protection Association, http://cefic-lri.org/wp-
content/uploads/2018/06/RSA-Draft-Thvroid-State-of-Science-Review-v4.pdf
Regulatory Science Associates. (2019). A literature review of the current state of the science
regarding species differences in the control of, and response to, thyroid hormone
perturbations. Part 1: A human health perspective. (RSA/ ECPA005 4111 001).
Brussels, Belgium: European Crop Protection Association.
Roman, GC; Ghassabian, A; Bongers-Schokking, JJ; Jaddoe, VW; Hofrnan, A; de Rijke, YB;
Verhulst, FC: Tiemeier, H. (2013). Association of gestational maternal hypothyroxinemia
and increased autism risk. Ann Neurol 74: 733-742. http://dx.doi.org/10.10Q2/ana.23976
Rovet, JF. (2014). The role of thyroid hormones for brain development and cognitive function
[Review], EndocrDev 26: 26-43. http://dx.doi.Org/10.l 159/000363153
Rudmann, DG: Foley, GL. (2018). Chapter 18. Female Reproductive System. In MA Wallig;
WM Haschek; CG Rousseaux; B Bolon (Eds.), Fundamentals of toxicologic pathology
(3rd ed., pp. 517-545). Cambridge, MA: Academic Press.
http://dx.doi.org/10.1016/B978-0-12-809841-7.00Q18-6
Savin, S: Cveiic, D; Nedic, O: Radosavlievic, R. (2003). Thyroid hormone synthesis and storage
in the thyroid gland of human neonates. J Pediatr Endocrinol Metab 16: 521-528.
Savu, L; Vranckx, R; Maya, M; Gripois, D; Blouquit, MF; Nunez, EA. (1989). Thyroxine-
binding globulin and thyroxine-binding prealbumin in hypothyroid and hyperthyroid
developing rats. Biochim Biophys Acta 992: 379-384.
Savu, L; Vranckx, R; Rouaze-Romet, M; Maya, M; Nunez, EA: Treton, J: Flink, IL. (1991). A
senescence up-regulated protein: the rat thyroxine-binding globulin (TBG). Biochim
Biophys Acta 1097: 19-22.
Schaider, LA: Balan, SA: Blum, A: Andrews, DO: Strynar, MJ; Dickinson, ME: Lunderberg,
DM; Lang, JR; Peaslee, GF. (2017). Fluorinated compounds in US fast food packaging.
Environ Sci Technol Lett 4: 105-111. http://dx.doi.org/10.1021/acs.estlett.6b00435
Seo, SH; Son, MH; Choi, SD; Lee, DH; Chang, YS. (2018). Influence of exposure to
perfluoroalkyl substances (PFASs) on the Korean general population: 10-year trend and
health effects. Environ Int 113: 149-161. http://dx.doi.Org/10.1016/i.envint.2018.01.025
A-29
-------
Sepulvado, JG; Blaine, AC; Hundal, LS; Higgins, CP. (2011). Occurrence and fate of
perfluorochemicals in soil following the land application of municipal biosolids. Environ
Sci Technol 45: 8106-8112. http://dx.doi.org/10.1021/eslQ3903d
Sferruzzi-Perri. AN: Vaughan. OR: Forhead. AJ: Fowden. AL. (2013). Hormonal and nutritional
drivers of intrauterine growth [Review], Curr Opin Clin Nutr Metab Care 16: 298-309.
http://dx.doi.org/10.1097/MCQ.0b013e32835e3643
Shiue, I. (2016). Urinary arsenic, pesticides, heavy metals, phthalates, polyaromatic
hydrocarbons, and polyfluoroalkyl compounds are associated with sleep troubles in
adults: USA NHANES, 2005-2006. Environ Sci Pollut Res Int 24: 3108-3116.
http://dx.doi.org/10.1007/sll356-016-8Q54-6
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/i.envint.2018.01.010
Stagnaro-Green, A; Abalovich, M; Alexander, E; Azizi, F; Mestman, J: Negro, R; Nixon, A;
Pearce, EN: Soldin, OP: Sullivan, S: Wiersinga, W. (2011). Guidelines of the American
Thyroid Association for the diagnosis and management of thyroid disease during
pregnancy and postpartum. Thyroid 21: 1081-1125.
http://dx.doi.org/10.1089/thy.2011.0087
Stenzel, D; Huttner, WB. (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
Sundstrom, M; Bogdanska, J: Pham, HV; Athanasios, V: Nobel, S: Mcalees, A: Eriksson, J:
Depierre, JW; Bergman, A. (2012). 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
Thompson, W; Russell, G; Baragwanath, G; Matthews, J: Vaidya, B; Thompson-Coon, J. (2018).
Maternal thyroid hormone insufficiency during pregnancy and risk of
neurodevelopmental disorders in offspring: A systematic review and meta-analysis. Clin
Endocrinol 88: 575-584. http://dx.doi.org/10.Ill 1/cen. 13550
Tsai, C: Huang, J: Hwang, BF; Lee, YL. (2010). Household environmental tobacco smoke and
risks of asthma, wheeze and bronchitic symptoms among children in Taiwan. Respir Res
11: 11. http://dx.doi.org/10.1186/1465-9921-11-11
U.S. EPA (U.S. Environmental Protection Agency). (1991a). Guidelines for developmental
toxicity risk assessment. Fed Reg 56: 63798-63826.
U.S. EPA (U.S. Environmental Protection Agency). (1991b). Guidelines for developmental
toxicity risk assessment (pp. 1-71). (EPA/600/FR-91/001). Washington, DC: U.S.
A-30
-------
Environmental Protection Agency, Risk Assessment Forum.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm7deicN23162
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/recordisplav.cfm?deid=71993&CFID=51174829&CFTO
KEN=25006317
U.S. EPA (U.S. Environmental Protection Agency). (1996a). Guidelines for reproductive toxicity
risk assessment. Fed Reg 61: 56274-56322.
U.S. EPA (U.S. Environmental Protection Agency). (1996b). Guidelines for reproductive
toxicity risk assessment (pp. 1-143). (EPA/630/R-96/009). Washington, DC: U.S.
Environmental Protection Agency, Risk Assessment Forum.
https://www.epa.gov/sites/production/files/2014-
11/documents/guidelines repro toxicitv.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/2Q13-
09/documents/cancer guidelines final 3-25-05.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2006). A framework for assessing health
risk of environmental exposures to children (pp. 1-145). (EPA/600/R-05/093F).
Washington, DC: U.S. Environmental Protection Agency, Office of Research and
Development, National Center for Environmental Assessment.
http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid=l 58363
U.S. EPA (U.S. Environmental Protection Agency). (201 la). Exposure factors handbook: 2011
edition [EPA Report], (EPA/600/R-090/052F). Washington, DC.
http://cfpub.epa. gov/ncea/cfm/recordisplay.cfm?deid=236252
U.S. EPA (U.S. Environmental Protection Agency). (201 lb). 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/2Q13-
09/documents/recommended-use-of-b w3 4 .pdf
A-31
-------
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). (2014a). Child-specific exposure scenarios
examples [EPA Report], (EPA/600/R-14/217F). Washington, DC.
http://cfpub.epa.gov/ncea/risk/recordisplav.cfm?deid=262211
U.S. EPA (U.S. Environmental Protection Agency). (2014b). Draft: Health effects document for
perfluorooctane sulfonate (PFOS) [EPA Report], (EPA/822/R-14/002). Washington, DC:
U.S. Environmental Protection Agency, Office of Water, Health and Ecological Criteria
Division. https://peerreview.versar.com/epa/pfoa/pdf/Health-Effects-Document-for-
Perfluorooctane- Sulfonate-(PF O S) .pdf
U.S. EPA (U.S. Environmental Protection Agency). (2014c). Framework for human health risk
assessment to inform decision making. Final [EPA Report], (EPA/100/R-14/001).
Washington, DC: U.S. Environmental Protection, Risk Assessment Forum.
https://www.epa.gov/risk/framework-human-health-risk-assessment-inform-decision-
making
U.S. EPA (U.S. Environmental Protection Agency). (2014d). Guidance for applying quantitative
data to develop data-derived extrapolation factors for interspecies and intraspecies
extrapolation [EPA Report], (EPA/100/R-14/002F). Washington, DC: Risk Assessment
Forum, Office of the Science Advisor. https://www.epa.gov/sites/production/files/2Q15-
01/ documents/ddef-final .pdf
U.S. EPA (U.S. Environmental Protection Agency). (2014e). Health effects document for
perfluorooctanoic acid (PFOA). (EPA/822/R-14/001). Washington, DC: Office of Water,
Health and Ecological Criteria Division.
https://peerreview.versar.com/epa/pfoa/pdf/Health-Effects-Document-for-
Perfluorooctanoic-Acid-(PFOA).pdf
U.S. EPA (U.S. Environmental Protection Agency). (2014f). Provisional peer-reviewed toxicity
values for perfluorobutane sulfonate (CASRN 375-73-5) and related compound
potassium perfluorobutane sulfonate (CASRN 29420-49-3) [EPA Report], (EPA/690/R-
14/012F). https://hhpprtv.ornl.gov/issue papers/PerfluorobutanesulfonicacidPFBS.pdf
U.S. EPA (U.S. Environmental Protection Agency). (2017). Unregulated Contaminant
Monitoring Rule 3 (2013-2015) occurrence data. Retrieved from
https://www.epa.gov/dwucmr/occurrence-data-unregulated-contaminant-monitoring-rule
U.S. EPA (U.S. Environmental Protection Agency). (2018). Safe drinking water information
system [Database], Retrieved from https://www3.epa.gov/enviro/facts/sdwis/search.html
Van Den Hove. MF; Beckers. C: Devlieger. H; De Zegher. F; De Naver. P. (1999). Hormone
synthesis and storage in the thyroid of human preterm and term newborns: effect of
thyroxine treatment [Review], Biochimie 81: 563-570.
A-32
-------
Vulsma, T; Gons, MH; de Vijlder, JJ. (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/NEJM1989070632101Q3
Wang. B: Zhang. R: Jin. F: Lou. H: Mao. Y: Zhu. W: Zhou. W: Zhang. P: Zhang. J. (2017).
Perfluoroalkyl substances and endometriosis-related infertility in Chinese women.
Environ Int 102: 207-212. http://dx.doi.Org/10.1016/i.envint.2017.03.003
Wang. Y; Rogan. WJ: Chen. PC: Lien. GW: Chen. HY; Tseng. YC: Longnecker. MP: Wang. SL.
(2014). Association between maternal serum perfluoroalkyl substances during pregnancy
and maternal and cord thyroid hormones: Taiwan maternal and infant cohort study.
Environ Health Perspect 122: 529-534. http://dx.doi.org/10.1289/ehp.1306925
Wasco. EC: Martinez. E; Grant. KS: St Germain. EA; St Germain. PL. (2003). Determinants of
iodothyronine deiodinase activities in rodent uterus. Endocrinology 144: 4253-4261.
http://dx.doi.org/10.1210/en.2003-049Q
Weaver. Y. iM; Ehresman. DJ; Butenhoff. JL; Hagenbuch. B. (2010). Roles of rat renal organic
anion transporters in transporting perfluorinated carboxylates with different chain
lengths. Toxicol Sci 113: 305-314. http://dx.doi.org/10.1093/toxsci/kfp275
White, SS; Calafat, AM; Kuklenyik, Z; Villanueva, L; Zehr, RD; Helfant, L; Strynar, MJ;
Lindstrom. AB; Thibodeaux. JR; Wood. C: Fenton. SE. (2007). Gestational PFOA
exposure of mice is associated with altered mammary gland development in dams and
female offspring. Toxicol Sci 96: 133-144. http://dx.doi.org/10.1093/toxsci/kfll77
Woldemeskel. M. (2017). Chapter 64 Toxicologic pathology of the reproductive system. In RC
Gupta (Ed.), Reproductive and Developmental Toxicology (Second Edition) (2nd ed., pp.
745-763). San Diego, CA: Academic Press. http://dx.doi.org/10.1016/B978-0-12-8Q4239-
7.00064-0
Wong. F; Macleod. M; Mueller. JF; Cousins. IT. (2014). Enhanced elimination of
perfluorooctane sulfonic acid by menstruating women: evidence from population-based
pharmacokinetic modeling. Environ Sci Technol 48: 8807-8814.
http://dx.doi.org/10.1021/es500796y
Xu, J. (2001). Test for chemical induction of chromosome aberration in cultured Chinese
hamster ovary (CHO) cells with and without metabolic activation. (Sponsor's study no.
132-009). Redfield, AR: Primedica Redfield.
Xu. Y; Fletcher. T; Pineda. D; Lindh. CH; Nilsson. C: Glynn. A: Yogs. C: Norstrom. K; Lilia. K;
Jakobsson. K; Li. Y. (2020). Serum Half-Lives for Short- and Long-Chain Perfluoroalkyl
Acids after Ceasing Exposure from Drinking Water Contaminated by Firefighting Foam.
Environ Health Perspect 128: 77004. http://dx.doi.org/10.1289/EHP6785
Yang. L; Li. J: Lai. J: Luan. H; Cai. Z; Wang. Y; Zhao. Y; Wu. Y. (2016). Placental transfer of
perfluoroalkyl substances and associations with thyroid hormones: Beijing prenatal
exposure study. Sci Rep 6: 21699. http://dx.doi.org/10.1038/srep21699
A-33
-------
Yao, 0: Shi, R; Wang, C; Han, W; Gao, Y, u; Zhang, Y, an; Zhou, Y; Ding, G; Tian, Y. (2019).
Cord blood Per- and polyfluoroalkyl substances, placental steroidogenic enzyme, and
cord blood reproductive hormone. Environ Int 129: 573-582.
http://dx.doi.Org/10.1016/i.envint.2019.03.047
York, RG. (2002). Oral (gavage) developmental toxicity study of potassium perfluorobutane
sulfonate (PFBS) in rats. (Argus Research Protocol Number 418023). St. Paul, MN: 3M
Corporate Toxicology.
York, RG. (2003 a). Oral (galvage) dosage-range developmental toxicity study of potassium
perfluorobutane sulfonate (PFBS) in rats. (Sponsor's study no. T-7485.11). Horsham, PA:
Argus Research.
York, RG. (2003b). Oral (gavage) repeated dose 90-day toxicity study of potassium
perfluorobutane sulfonate (PFBS) in rats. (Argus Research Protocol Number 418026).
Washington, D.C.: U.S. Environmental Protection Agency.
York, RG. (2003c). Oral (gavage) two-generation (one litter per generation) reproduction study
of perfluorobutane sulfonate (PFBS) in rats. (Argus Research Protocol Number 418021).
Washington, D.C.: U.S. Environmental Protection Agency.
York, RG. (2003d). Oral (gavage) two-generation (one litter per generation) reproduction study
of perfluorobutane sulfonate (PFBS) in rats. Appendix D. Report tables - F1 generation
male rats. (Argus Research Protocol Number 418021). Washington, D.C.: U.S.
Environmental Protection Agency.
York, RG. (2003e). Oral (gavage) two-generation (one litter per generation) reproduction study
of perfluorobutane sulfonate (PFBS) in rats. Appendix K. Hisopathology report. (Argus
Research Protocol Number 418021). Washington, D.C.: U.S. Environmental Protection
Agency.
Zeng, XW: Qian, Z; Emo, B; Vaughn, M; Bao, J: Oin, XD; Zhu, Y; Li, J: Lee, YL; Dong, GH.
(2015).	Association of polyfluoroalkyl chemical exposure with serum lipids in children.
Sci Total Environ 512-513: 364-370. http://dx.doi.org/10.1016/j.scitotenv.2015.01.042
Zhang, S: Tan, R; Pan, R; Xiong, J: Tian, Y; Wu, J: Chen, L. (2018). Association of
perfluoroalkyl and polyfluoroalkyl substances with premature ovarian insufficiency in
Chinese women. J Clin Endocrinol Metab 103: 2543-2551.
http://dx.doi.org/10.1210/ic.2017-02783
Zhang, X: Lohmann, R; Dassuncao, C: Hu, XC: Weber, AK; Vecitis, CD: Sunderland, EM.
(2016).	Source attribution of poly- and perfluoroalkyl substances (PFASs) in surface
waters from Rhode Island and the New York metropolitan area. Environ Sci Technol Lett
3: 316-321. http://dx.doi.org/10.1021/acs.estlett.6b00255
Zhang, Y; Beesoon, S: Zhu, L; Martin, JW. (2013). Biomonitoring of perfluoroalkyl acids in
human urine and estimates of biological half-life. Environ Sci Technol 47: 10619-10627.
http://dx.doi.org/10.1021/es401905e
A-34
-------
Zhao, Z; Xie, Z; Moeller, A; Sturm, R; Tang, J; Zhang, G, an; Ebinghaus, R. (2012). Distribution
and long-range transport of polyfluoroalkyl substances in the Arctic, Atlantic Ocean and
Antarctic coast. Environ Pollut 170: 71-77.
http://dx.doi.Org/10.1016/i.envpol.2012.06.004
Zhou, W: Zhang, L; Tong, C: Fang, F; Zhao, S: Tian, Y; Tao, Y; Zhang, J: Study, SBC. (2017a).
Plasma perfluoroalkyl and polyfluoroalkyl substances concentration and menstrual cycle
characteristics in preconception women. Environ Health Perspect 125: 067012.
http://dx.doi.org/10.1289/EHP1203
Zhou, Y; Bao, WW: Qian, ZM; Dee Geiger, S: Parrish, KL; Yang, BY: Lee, YL; Dong, GH.
(2017b). Perfluoroalkyl substance exposure and urine CC16 levels among asthmatics: A
case-control study of children. Environ Res 159: 158-163.
http://dx.doi.Org/10.1016/i.envres.2017.08.005
Zhou, Y; Hu, LW; Qian, ZM; Chang, JJ; King, C; Paul, G; Lin, S; Chen, PC: Lee, YL; Dong,
GH. (2016). Association of perfluoroalkyl substances exposure with reproductive
hormone levels in adolescents: By sex status. Environ Int 94: 189-195.
http://dx.doi.Org/10.1016/i.envint.2016.05.018
Zhu. Y: Qin. XD: Zeng. XW: Paul. G: Morawska. L: Su. MW: Tsai. CH: Wang. SO: Lee. YL:
Dong. GH. (2016). Associations of serum perfluoroalkyl acid levels with T-helper cell-
specific cytokines in children: By gender and asthma status. Sci Total Environ 559: 166-
173. http://dx.doi.org/10.1016/i.scitotenv.2016.03.187
Zoeller. RT; Tan. SW: Tyl. RW. (2007). General background on the hypothalamic-pituitary-
thyroid (HPT) axis [Review], CritRev Toxicol 37: 11-53.
http://dx.doi.org/10.1080/104084406Q1123446
A-35
-------