^Cm
EPA/635/R-23/027Fb
www.ftpa.nnv/iris
IRIS Toxicological Review of Perfluorohexanoic Acid
[PFHxA, CASRN 307-24-4] and Related Salts
Supplemental Information
April 2023
Integrated Risk Information System
Center for Public Health and Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
-------
Supplemental Information—PFHxA and Related Salts
DISCLAIMER
This document has been reviewed by the U.S. Environmental Protection Agency, Office of
Research and Development and approved for publication. Any mention of trade names, products, or
services does not imply an endorsement by the U.S. government or the U.S. Environmental
Protection Agency. EPA does not endorse any commercial products, services, or enterprises.
ii
-------
Supplemental Information—PFHxA and Related Salts
TABLE OF CONTENTS
TABLE OF CONTENTS iii
APPENDIX A. SYSTEMATIC REVIEW PROTOCOL FOR THE PFAS IRIS ASSESSMENTS A-l
APPENDIX B. BENCHMARK DOSE MODELING RESULTS B-l
B.l. MODELING PROCEDURE FOR CONTINUOUS NONCANCER DATA B-l
B.2. MODELING PROCEDURE FOR DICHOTOMOUS NONCANCER DATA B-2
B.3. MODELING PROCEDURE FOR DICHOTOMOUS NONCANCER DEVELOPMENTAL
TOXICITY DATA B-2
B.4. HEMOGLOBIN—FEMALE RATS (KLAUNIG ET AL., 2015) B-3
B.5. HEMOGLOBIN—MALE RATS (CHENGELIS ET AL., 2009B) B-4
B.6. HEMOGLOBIN—FEMALE RATS (CHENGELIS ET AL., 2009B) B-5
B.7. HEMOGLOBIN—MALE RATS (LOVELESS ET AL., 2009) B-6
B.8. HEMOGLOBIN—FEMALE RATS (LOVELESS ET AL., 2009) B-8
B.9. RED BLOOD CELLS—MALE RATS (KLAUNIG ET AL., 2015) B-9
B.10. RED BLOOD CELLS—FEMALE RATS (KLAUNIG ET AL., 2015) B-10
B.ll. RED BLOOD CELLS—MALE RATS (CHENGELIS ET AL., 2009B) B-12
B.12. RED BLOOD CELLS—FEMALE RATS (CHENGELIS ET AL., 2009B) B-13
B.13. RED BLOOD CELLS—MALE RATS (LOVELESS ET AL., 2009) B-14
B.14. RED BLOOD CELLS—FEMALE RATS (LOVELESS ET AL., 2009) B-16
B.15. HEPATOCELLULAR NECROSIS—FEMALE RATS (KLAUNIG ET AL., 2015) B-17
B.16. HEPATOCELLULAR HYPERTROPHY—FEMALE RATS (LOVELESS ET AL., 2009) B-18
B.17. HEPATOCELLULAR HYPERTROPHY—MALE RATS (LOVELESS ETAL., 2009) B-19
B.18. POSTNATAL (Fi) COMBINED RAT BODY WEIGHT ON PND 0 (LOVELESS ET AL., 2009) B-21
B.19. POSTNATAL (Fi) COMBINED MOUSE BODY WEIGHT (PHASE 2) ON PND 0 (IWAI AND
HOBERMAN, 2014) B-22
B.20. POSTNATAL (Fi) COMBINED MOUSE BODY WEIGHT (PHASE 1) ON PND 0 (IWAI AND
HOBERMAN, 2014) B-24
B.21. POSTNATAL (Fi) COMBINED MOUSE BODY WEIGHT (PHASES 1 AND 2) ON PND 0 (IWAI
AND HOBERMAN, 2014) B-25
B.22. POSTNATAL (Fi) COMBINED MOUSE BODY WEIGHT (PHASE 2) ON PND 4 (IWAI AND
HOBERMAN, 2014) B-27
iii
-------
Supplemental Information—PFHxA and Related Salts
B.23. POSTNATAL (Fi) COMBINED MOUSE BODY WEIGHT (PHASE 1) ON PND 4 (IWAI AND
HOBERMAN, 2014) B-28
B.24. POSTNATAL (Fi) COMBINED MOUSE BODY WEIGHT (PHASES 1 AND 2) ON PND 4 (IWAI
AND HOBERMAN, 2014) B-29
B.25. PERINATAL MORTALITY (PHASE 2) ON PND 0-21 (IWAI AND HOBERMAN, 2014) B-31
B.26. PERINATAL MORTALITY (PHASE 1) ON PND 0-21 (IWAI AND HOBERMAN, 2014) B-32
B.27. PERINATAL MORTALITY (PHASES 1 AND 2) ON PND 0-21 (IWAI AND HOBERMAN,
2014) B-33
B.28. TOTAL THYROXINE (T4) IN MALE RATS - (NTP, 2018) B-34
APPENDIX C. EVALUATION OF PFHxA ELIMINATION C-l
C.l. EVALUATION OF PFHXA ELIMINATION IN RATS AND MICE C-l
C.l.l. Mice C-2
C.l.2. Rats C-4
C.2. EVALUATION OF PFHXA ELIMINATION IN HUMANS C-4
APPENDIX D. QUALITY ASSURANCE FOR THE IRIS TOXICOLOGICAL REVIEW OF PFHxA D-l
APPENDIX E. SUMMARY OF PUBLIC AND EXTERNAL PEER REVIEW COMMENTS AND EPA'S
DISPOSITION E-l
E.l. CHARGE QUESTIONS 1 AND 2 - SYSTEMATIC REVIEW AND DOCUMENTATION E-2
E.l.l. External Peer Reviewer Comments on Systematic Review and Documentation E-3
E.l.2. Public Comments on Systematic Review and Documentation E-7
E.2. CHARGE QUESTION 3: NONCANCER HAZARD IDENTIFICATION E-7
E.2.1. External Peer Review Comments on Hepatic Effects E-8
E.2.2. Public Comments on Hepatic Effects E-10
E.2.3. External Peer Review Comments on Developmental Effects E-10
E.2.4. Public Comments on Developmental Effects E-ll
E.2.5. External Peer Review Comments on Hematopoietic Effects E-ll
E.2.6. Public Comments on Hematopoietic Effects E-13
E.2.7. External Peer Review Comments on Endocrine Effects E-13
E.2.8. Public Comments on Endocrine Effects E-14
E.2.9. External Peer Review Comments on All Other Potential Health Effects E-14
E.2.10. Public Comments on All Other Potential Health Effects E-16
E.3. CHARGE QUESTIONS 4 AND 5: NONCANCER TOXICITY VALUES DATA SELECTION E-16
E.3.1. External Peer Review Comments on Noncancer Toxicity Values Data Selection E-17
E.3.2. Public Comments on Noncancer Toxicity Values Data Selection E-19
iv
-------
Supplemental Information—PFHxA and Related Salts
E.4. CHARGE QUESTIONS 6,7, AND 8: NONCANCER TOXICITY VALUE DERIVATION E-19
E.4.1. External Peer Review Comments on Noncancer Toxicity Value Derivation E-20
E.4.2. Public Comments on Noncancer Toxicity Value Derivation E-24
E.5. CHARGE QUESTION 9 AND 10: CARCINOGENICITY HAZARD IDENTIFICATION AND
TOXICITY VALUE DERIVATION E-26
E.5.1. External Peer Review Comments on Carcinogenicity Hazard Identification and
Toxicity Value Derivation E-27
E.5.2. Public Comments on Carcinogenicity Hazard Identification and Toxicity Value
Derivation E-27
E.6. ADDITIONAL COMMENTS E-27
E.6.1. Additional External Peer Review Comments E-27
E.6.2. Additional Public Comments E-28
REFERENCES R-l
v
-------
Supplemental Information—PFHxA and Related Salts
TABLES
Table B-l. Dose response data for hemoglobin in female rats (Klaunig et al., 2015) B-3
Table B-2. Benchmark dose results for hemoglobin in female rats—constant variance, BMR = 1
standard deviation (Klaunig et al., 2015) B-3
Table B-3. Dose response data for hemoglobin in male rats (Chengelis et al., 2009b) B-4
Table B-4. Dose response data for hemoglobin in female rats (Chengelis et al., 2009b) B-5
Table B-5. Benchmark dose results for hemoglobin in female rats—non-constant variance,
BMR = 1 standard deviation (Chengelis et al., 2009b) B-5
Table B-6. Dose response data for hemoglobin in male rats (Loveless et al., 2009) B-6
Table B-7. Benchmark dose results for hemoglobin in male rats—non-constant variance,
BMR = 1 standard deviation (Loveless et al., 2009) B-7
Table B-8. Dose response data for hemoglobin in female rats (Loveless et al., 2009) B-8
Table B-9. Benchmark dose results for hemoglobin in female rats—constant variance, BMR = 1
standard deviation (Loveless etal., 2009) B-8
Table B-10. Dose response data for red blood cells in male rats (Klaunig et al., 2015) B-9
Table B-ll. Benchmark dose results for red blood cells in male rats—non-constant variance,
BMR = 1 standard deviation (Klaunig et al., 2015) B-9
Table B-12. Dose response data for red blood cells in female rats (Klaunig et al., 2015) B-10
Table B-13. Benchmark dose results for red blood cells in female rats—constant variance,
BMR = 1 standard deviation (Klaunig et al., 2015) B-ll
Table B-14. Dose response data for red blood cells in male rats (Chengelis et al., 2009b) B-12
Table B-15. Benchmark dose results for red blood cells in male rats—non-constant variance,
BMR = 1 standard deviation (Chengelis et al., 2009b) B-12
Table B-16. Dose response data for red blood cells in female rats (Chengelis et al., 2009b) B-13
Table B-17. Benchmark dose results for red blood cells in female rats—constant variance,
BMR = 1 standard deviation (Chengelis et al., 2009b) B-13
Table B-18. Dose response data for red blood cells in male rats (Loveless et al., 2009) B-14
Table B-19. Benchmark dose results for red blood cells in male rats—non-constant variance,
BMR = 1 standard deviation (Loveless et al., 2009) B-15
Table B-20. Dose response data for red blood cells in female rats (Loveless et al., 2009) B-16
Table B-21. Benchmark dose results for red blood cells in female rats—constant variance,
BMR = 1 standard deviation (Loveless et al., 2009) B-16
Table B-22. Dose response data for hepatocellular necrosis in female rats (Klaunig et al., 2015) B-17
Table B-23. Dose response data for hepatocellular hypertrophy in female rats (Loveless et al.,
2009) B-18
Table B-24. Dose response data for hepatocellular hypertrophy in male rats (Loveless et al.,
2009) B-19
Table B-25. Benchmark dose results for hepatocellular hypertrophy in male rats—nested model
BMR = 10% extra risk (Loveless et al., 2009) B-20
Table B-26. Dose response data for postnatal (Fi) combined rat body weight on PND 0 (Loveless
etal., 2009) B-21
Table B-27. Benchmark dose results for postnatal (Fi) combined rat body weight on PND
0—non-constant variance, BMR = 5% relative deviation (Loveless et al., 2009) B-21
Table B-28. Dose response data for postnatal (Fi) combined mouse body weight (phase 2) on
PND 0 (Iwai and Hoberman, 2014) B-22
vi
-------
Supplemental Information—PFHxA and Related Salts
Table B-29. Benchmark dose results for postnatal (Fi) combined mouse body weight (phase 2)
on PND 0—constant variance, BMR = 5% relative deviation (Iwai and Hoberman,
2014) B-22
Table B-30. Dose response data for postnatal (Fi) combined mouse body weight (phase 1) on
PND 0 (Iwai and Hoberman, 2014) B-24
Table B-31. Benchmark dose results for postnatal (Fi) combined mouse body weight (phase 1)
on PND 0—non-constant variance, BMR = 5% relative deviation (Iwai and
Hoberman, 2014) B-24
Table B-32. Dose response data for postnatal (Fi) combined mouse body weight (phases 1 and 2)
on PND 0 (Iwai and Hoberman, 2014) B-25
Table B-33. Benchmark dose results for postnatal (Fi) combined mouse body weight (phases 1
and 2) on PND 0—non-constant variance, BMR = 5% relative deviation (Iwai and
Hoberman, 2014) B-26
Table B-34. Dose response data for postnatal (Fi) combined mouse body weight (phase 2) on
PND 4 (Iwai and Hoberman, 2014) B-27
Table B-35. Benchmark dose results for postnatal (Fi) combined mouse body weight (phase 2)
on PND 4—constant variance, BMR = 5% relative deviation (Iwai and Hoberman,
2014) B-27
Table B-36. Dose response data for postnatal (Fi) combined mouse body weight (phase 1) on
PND 4 (Iwai and Hoberman, 2014) B-28
Table B-37. Benchmark dose results for postnatal (Fi) combined mouse body weight (phase 1)
on PND 4—non-constant variance, BMR = 5% relative deviation (Iwai and
Hoberman, 2014) B-28
Table B-38. Dose response data for postnatal (Fi) combined mouse body weight (phase 1) on
PND 4 (Iwai and Hoberman, 2014) B-29
Table B-39. Benchmark dose results for postnatal (Fi) combined mouse body weight (phases 1
and 2) on PND 4—non-constant variance, BMR = 5% relative deviation (Iwai and
Hoberman, 2014) B-30
Table B-40. Nested model summary for perinatal mortality (phase 2) on PND 0-21, BMR = 1%
extra risk (Iwai and Hoberman, 2014) B-31
Table B-41. Nested model summary for perinatal mortality (phase 1) on PND 0-21, BMR = 1%
extra risk (Iwai and Hoberman, 2014) B-32
Table B-42. Nested model summary for perinatal mortality (phases 1 and 2) on PND 0-21,
BMR = 1% extra risk (Iwai and Hoberman, 2014) B-33
Table B-43. Dose response data for thyroxine (T4) in male rats (NTP, 2018) B-34
Table B-44. Benchmark dose results for total thyroxine (T4) in male rats—constant variance,
BMR = 1 standard deviation (NTP, 2018) B-34
FIGURES
Figure B-l. Dose response curve for the Linear model fit to hemoglobin in female rats (Klaunig et
al., 2015) B-4
Figure B-2. Dose response data for hemoglobin in male rats (Chengelis et al., 2009b) B-5
Figure B-3. Dose response data for hemoglobin in female rats (Chengelis et al., 2009b) B-6
Figure B-4. Dose response data for hemoglobin in male rats (Loveless et al., 2009) B-7
vii
-------
Supplemental Information—PFHxA and Related Salts
Figure B-5. Dose response curve for the Polynomial Degree 3 model fit to hemoglobin in female
rats (Loveless et al., 2009) B-9
Figure B-6. Dose response data red blood cells in male rats (Klaunig et al., 2015) B-10
Figure B-7. Dose response curve for the Linear model fit to red blood cells in female rats (Klaunig
et al., 2015) B-ll
Figure B-8. Dose response data for red blood cells in male rats (Chengelis et al., 2009b) B-13
Figure B-9. Dose response curve for the Exponential 5 model fit to red blood cells in female rats
(Chengelis et al., 2009b) B-14
Figure B-10. Dose response curve for the Linear model fit to red blood cells in male rats
(Loveless et al., 2009) B-16
Figure B-ll. Dose response curve for the Linear model fit to red blood cells in female rats
(Loveless et al., 2009) B-17
Figure B-12. Dose response data for hepatocellular necrosis in female rats (Klaunig et al., 2015) B-18
Figure B-13. Dose response data for hepatocellular hypertrophy in female rats (Loveless et al.,
2009) B-19
Figure B-14. Dose response curve for the Multistage Degree 1 model fit to hepatocellular
hypertrophy in male rats (Loveless et al., 2009) B-20
Figure B-15. Dose response curve for the Hill model fit to postnatal (Fi) combined rat body
weight on PND 0 (Loveless et al., 2009) B-22
Figure B-16. Dose response curve for the Polynomial Degree 3 model fit to postnatal (Fi)
combined rat body weight (phase 2) on PND 0 (Iwai and Hoberman, 2014) B-23
Figure B-17. Dose response data for postnatal (FI) combined rat body weight (phase 1) on PND
0 (Iwai and Hoberman, 2014) B-25
Figure B-18. Dose response data for postnatal (FI) combined rat body weight (phases 1 and 2)
on PND 0 (Iwai and Hoberman, 2014) B-26
Figure B-19. Dose response curve for the Polynomial model (poly 3) fit to postnatal (FI)
combined rat body weight (phase 2) on PND 4 (Iwai and Hoberman, 2014) B-28
Figure B-20. Dose response data for postnatal (FI) combined rat body weight (phase 1) on PND
4 (Iwai and Hoberman, 2014) B-29
Figure B-21. Dose response curve for the Exponential 5 model fit to postnatal (FI) combined
mouse body weight (phases 1 and 2) on PND 4 (Iwai and Hoberman, 2014) B-31
Figure B-22. Dose response curve for the Nested National Center for Toxicological Research
model fit to perinatal mortality (phase 2) on PND 0-21(lwai and Hoberman,
2014) B-32
Figure B-23. Dose response data perinatal mortality (phase 1) on PND 4(lwai and Hoberman,
2014) B-33
Figure B-24. Dose response data perinatal mortality (phases 1 and 2) on PND 0-21 (Iwai and
Hoberman, 2014) B-34
Figure B-25. Dose response curve for the Hill model fit to thyroxine (T4) in male rats (NTP, 2018) B-35
Figure C-l. Fits of population pharmacokinetic model to data for male (top row) and female
(remaining rows) mice following 2-350 mg/kg oral exposure PFHxA C-3
Figure C-2. Fits of human PFHxA data from ski-wax technician blood samples C-5
viii
-------
Supplemental Information—PFHxA and Related Salts
ABBREVIATIONS AND ACRONYMS
ADME
absorption, distribution, metabolism,
i.v.
intravenous
and excretion
LDH
lactate dehydrogenase
AFFF
aqueous film-forming foam
LLOQ
lower limit of quantitation
A:G
albumin:globulin ratio
LOQ
limit of quantitation
AIC
Akaike's information criterion
LOAEL
lowest-observed-adverse-effect level
ALP
alkaline phosphatase
LOD
limit of detection
ALT
alanine aminotransferase
LOEC
lowest observed effect concentration
APTT
activated partial thromboplastin time
MCH
mean cell hemoglobin
AST
aspartate aminotransferase
MCHC
mean cell hemoglobin concentration
atm
atmosphere
MCV
mean corpuscular volume
ATSDR
Agency for Toxic Substances and
MOA
mode of action
Disease Registry
MW
molecular weight
AUC
area under the curve
NCTR
National Center for Toxicological
BMD
benchmark dose
Research
BMDL
benchmark dose lower confidence limit
NOAEL
no-observed-adverse-effect level
BMDS
Benchmark Dose Software
NPL
National Priorities List
BMR
benchmark response
NTP
National Toxicology Program
BUN
blood urea nitrogen
ORD
Office of Research and Development
BW
body weight
OECD
Organisation for Economic
Cmax
maximum concentration
Co-operation and Development
CAR
constitutive androstane receptor
OSF
oral slope factor
CASRN
Chemical Abstracts Service registry
osRfD
organ/system-specific oral reference
number
dose
CBC
complete blood count
PBPK
physiologically based pharmacokinetic
CHO
Chinese hamster ovary (cell line cells)
PC
partition coefficient
CI
confidence interval
PECO
populations, exposures, comparators,
CL
clearance
and outcomes
CLa
clearance in animals
PFAA
perfluoroalkyl acids
CLh
clearance in humans
PFAS
per- and polyfluoroalkyl substances
CPHEA
Center for Public Health and
PFBA
perfluorobutanoic acid
Environmental Assessment
PFBS
perfluorobutane sulfonate
CPN
chronic progressive nephropathy
PFCA
perfluorinated carboxylic acid
DAF
dosimetric adjustment factor
PFDA
perfluorodecanoic acid
DNA
deoxyribonucleic acid
PFHxA
perfluorohexanoic acid
DTXSID
DSSTox substance identifier
PFHxS
perfluorohexane sulfonate
EPA
Environmental Protection Agency
PFNA
perfluorononanoic acid
FTOH
fluorotelomer alcohol
PFOA
perfluorooctanoic acid
GD
gestation day
PFOS
perfluorooctane sulfonate
GGT
y-glutamyl transferase
PK
pharmacokinetic
HAWC
Health Assessment Workplace
PND
postnatal day
Collaborative
POD
point of departure
HCT
hematocrit
PODhed
human equivalent dose POD
HED
human equivalent dose
PPAR
peroxisome proliferated activated
HERO
Health and Environmental Research
receptor
Online
PQAPP
programmatic quality assurance
HGB
hemoglobin
project plan
HSA
human serum albumin
PT
prothrombin time
IQR
interquartile range
QA
quality assurance
IRIS
Integrated Risk Information System
QAPP
quality assurance project plan
ISI
Influential Scientific Information
QMP
quality management plan
IUR
inhalation unit risk
RBC
red blood cells
ix
-------
RfC
RfD
RNA
ROS
RXR
SD
TP
TRI
TSCATS
TSH
UF
UFa
UFc
UFd
UFh
UFl
UFs
Vd
Supplemental Information—PFHxA and Related Salts
reference concentration
oral reference dose
ribonucleic acid
reactive oxygen species
retinoid X receptor
standard deviation
total protein
Toxics Release Inventory
Toxic Substances Control Act Test
Submissions
thyroid stimulating hormone
uncertainty factor
interspecies uncertainty factor
composite uncertainty factor
evidence base deficiencies uncertainty
factor
human variation uncertainty factor
LOAEL to NOAEL uncertainty factor
subchronic to chronic uncertainty
factor
volume of distribution
x
-------
Supplemental Information—PFHxA and Related Salts
APPENDIX A. SYSTEMATIC REVIEW PROTOCOL FOR
THE PFAS IRIS ASSESSMENTS
A single systematic review protocol was used to guide the development of five separate IRIS
PFAS [per- and polyfluoroalkyl substances] assessments (i.e., perfluorobutanoic acid [PFBA],
perfluorohexanoic acid [PFHxA], perfluorohexane sulfonate [PFHxS], perfluorononanoic acid
[PFNA], and perfluorodecanoic acid [PFDA]). This "Systematic Review Protocol for the PFAS IRIS
Assessments" was released for public comment and subsequently updated. The updated protocol
and prior versions can be found at the following location:
http://cfpub.epa.gov/ncea/iris drafts/recordisplay.cfm?deid=345065
A-l
-------
Supplemental Information—PFHxA and Related Salts
APPENDIX B. BENCHMARK DOSE MODELING
RESULTS
As discussed in the body of the report (see Section 5), the endpoints selected for benchmark
dose (BMD) modeling were hepatocellular hypertrophy from Chengelis etal. (2009a) and Loveless
etal. (20091: hemoglobin and red blood cells from Chengelis etal. (2009a): Loveless etal. (20091.
and Klaunigetal. f 20151: postnatal body weight decreases from Loveless etal. f20091 and Iwai and
Hoberman f20141: and perinatal mortality from Iwai and Hoberman f20141. The animal doses in
the studies were used in the BMD modeling and then converted to human equivalent doses (HEDs)
using the ratio of animal-to-human serum half-lives.
B.l. MODELING PROCEDURE FOR CONTINUOUS NONCANCER DATA
BMD modeling of continuous noncancer data was conducted using EPA's Benchmark Dose
Software (BMDS, Version 3.2). For these data, the Exponential, Hill, Polynomial, and Power models
available within the software are fit using a benchmark response (BMR) of 1 standard deviation
(SD) when no toxicological information was available to determine an adverse level of response.
When toxicological information was available, the BMR was based on relative deviation, as outlined
in the Benchmark Dose Technical Guidance (U.S. EPA. 20121. An adequate fit is judged on the basis
of ax2 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
addition to these three criteria for judging adequacy of model fit, a determination is made as to
whether the variance across dose groups is homogeneous. If a homogeneous variance model, also
referred to as a "constant variance" (CV) model, is deemed appropriate based on the statistical test
provided by BMDS (Test 2 for homogeneity of variance), the final BMD results are presented for the
CV model. If the Test 2 p-value is significant (p < 0.05), the model is run again while modeling the
variance as a power function of the mean to account for this nonhomogeneous variance, also
referred to as "non-constant variance" (NCV). If the NCV model provides adequate fit to the
variance of the data (i.e., Test 3 p-value > 0.05), the final BMD results are presented for the NCV
model. If this nonhomogeneous variance model does not adequately fit the data (i.e., Test 3;
p < 0.05), the data set is considered unsuitable for BMD modeling. In some cases, the data may be
remodeled after removing one or more of the highest dose groups; if the reduced data can be
modeled and results in a better fit in the low-dose region, these results will be presented with
information to indicate that one or more dose groups were removed in the results table. In cases
where 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 point of
B-l
-------
Supplemental Information—PFHxA and Related Salts
departure (POD). Among all models providing adequate fit, the benchmark dose lower confidence
limit (BMDL) from the model with the lowest Akaike's information criterion (AIC) was selected as a
potential POD when BMDL estimates differed by less than threefold. When BMDL estimates differed
by greater than threefold, the model with the lowest BMDL was selected to account for model
uncertainty.
For body weight and relative liver weight, a BMR equal to 10% relative deviation (increase
or decrease) from the control mean was used based on a biological consideration. For continuous
developmental toxicity data, a BMR equal to 0.5 SD was used. The use of 1 SD for the BMR for
continuous endpoints is based the observation that shifting the distribution of the control group by
1 SD results in ~10% of animals falling beyond an adversity cutoff defined at the ~1.5 percentile in
the control group fCrump. 19951. This roughly approximates the 10% extra risk commonly used as
the BMR for dichotomous endpoints. Thus, the use of 0.5 SD for continuous developmental toxicity
endpoints roughly approximates the extra risk of 5% commonly used for dichotomous
developmental toxicity endpoints; similarly, the BMR for perinatal body weight is half of that for
adults (5% vs. 10% relative deviation) based on biological consideration.
B.2. MODELING PROCEDURE FOR DICHOTOMOUS NONCANCER DATA
BMD modeling of dichotomous noncancer data was conducted using EPA's Benchmark Dose
Software (BMDS, version 3.2). For these data, the Gamma, Logistic, LogLogistic, LogProbit,
Multistage, Probit, Weibull, and Dichotomous Hill models available within the software were fit
using a benchmark response (BMR) of 10% extra risk (5% extra risk for developmental endpoints
and 1% for mortality). 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 on the basis of
X2 goodness of fit p value (p > 0.1), scaled residuals at the data point (except the control) closest to
the predefined benchmark response (absolute value <2.0), and visual inspection of the model fit.
Among all models providing adequate fit, the benchmark dose lower confidence limit (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.
B.3. MODELING PROCEDURE FOR DICHOTOMOUS NONCANCER
DEVELOPMENTAL TOXICITY DATA
For dichotomous developmental toxicity data, data for individual animals were requested
from the study authors when possible. This allowed the use of the nested logistic model, which
statistically accounts for intralitter similarity (the propensity of littermates to respond more like
one another than pups from another litter) by estimating intralitter correlation and using litter-
specific covariates. Judging model fit for this model is identical to the procedure used for regular
dichotomous models. If individual animal data is available, the nested logistic model is used instead
B-2
-------
Supplemental Information—PFHxA and Related Salts
of other models; this will be indicated in the results table and individual animal data will not be
reported.
For all data types discussed in Sections B.1-B.4, the NOAEL/LOAEL approach may be used
in lieu of BMD modeling to derive reference values, for example, when BMD modeling fails. The
NOAEL/LOAEL approach may also be taken when a response is only observed in the highest dose
group unless the response for that group is sufficiently close to the BMR, in which case BMD
modeling results are used to derive values.
B.4. HEMOGLOBIN-FEMALE RATS fKLAUNIG ET AL.. 20151
Table B-l. Dose response data for hemoglobin in female rats fKlaunig et al..
20151
Dose (mg/kg-d)
Number of
animals
Mean (g/dL)
Standard deviation
0
10
15.5
0.97
5
10
15.7
0.73
30
9
15.5
0.79
200
20
14.7
0.91
Table B-2. Benchmark dose results for hemoglobin in female rats—constant
variance, BMR = 1 standard deviation fKlaunig et al.. 20151
Models
Goodness
of fit
(p-value)
Test 2
(p-value)
AIC
1SD
Scaled residual
for dose group
near BMD
BMD
BMDL
Exponential 2
0.83237671
0.7551
127.5023
182.1091
120.47632
-0.016328458
Exponential 3
0.57454386
0.7551
129.4505
189.9502
120.87504
0.002065941
Exponential 4
0.83237684
0.7551
127.5023
182.0793
120.47647
-0.015845878
Exponential 5
0.57331833
0.7551
129.4525
188.501
120.85884
-0.000961421
Hill
NA
0.7551
131.4228
42.56095
31.718073
0.000755846
Polynomial (Poly 3)
0.56681655
0.7551
129.4634
191.4936
123.01116
0.00113966
Polynomial (Poly 2)
0.56681165
0.7551
129.4634
191.4757
123.0163
0.001201719
Power
0.57425021
0.7551
129.451
190.1218
123.04966
0.002120702
Linear
0.83402366
0.7551
127.4983
182.7286
122.7699
-0.014122961
Bold row indicates the selected model and values.
B-3
-------
Supplemental Information—PFHxA and Related Salts
Frequentist Linear Model with BMR of 1 Std. Dev. for the BMD
and 0.95 Lower Confidence Limit for the BMDL
18
14
-------
Supplemental Information—PFHxA and Related Salts
Hemoglobin (HGB)
16.5
13
0 20 40 60 80 100 120 140 160 180 200
Dose
Figure B-2. Dose response data for hemoglobin in male rats fChengelis et al..
2009b)
X-axis is dose (mg/kg-day), and y-axis is mean level of hemoglobin (g/dL).
B.6. HEMOGLOBIN-FEMALE RATS (CHENGELIS ET AL.. 2009B1
Table B-4. Dose response data for hemoglobin in female rats (Chengelis etal..
2009b)
Dose (mg/kg-d)
Number of
animals
Mean (g/dL)
Standard deviation
0
10
15.6
0.46
10
10
15.8
1.4
50
10
15.2
0.85
200
10
14.6
0.83
Table B-5. Benchmark dose results for hemoglobin in female rats—non-
constant variance, BMR = 1 standard deviation fChengelis et al.. 2009bl
Models
Goodness
of fit
(p-value)
Test 3
(p-value)
AIC
1SD
Scaled residual
for dose group
near BMD
BMD
BMDL
Exponential 2
0.2289302
0.0118
113.344
177.8625
106.4881
0.107636338
Exponential 3
0.2289313
0.0118
113.344
177.8107
106.4883
0.107765953
Exponential 4
0.0996002
0.0118
115.1073
145.8206
37.82113
0.036336773
Exponential 5
0.165197
0.0118
114.3214
53.45159
25.38329
0.084452285
B-5
-------
Supplemental Information—PFHxA and Related Salts
Models
Goodness
of fit
(p-value)
Test 3
(p-value)
AIC
1SD
Scaled residual
for dose group
near BMD
BMD
BMDL
Hill
NA
0.0118
116.3216
68.93813
40.08308
0.084913389
Polynomial (Poly 3)
0.2265515
0.0118
113.3649
179.1794
109.5823
0.105758473
Polynomial (Poly 2)
0.2265515
0.0118
113.3649
179.1817
110.5758
0.105711892
Power
0.2265515
0.0118
113.3649
179.174
109.6196
0.105830655
Linear
0.2265515
0.0118
113.3649
179.1809
110.1348
0.10575497
Both constant and nonconstant variance models failed to model the variance of the data.
Hemoglobin (HGB)
13.5
0 20 40 60 80 100 120 140 160 180 200
Dose
Figure B-3. Dose response data for hemoglobin in female rats fChengelis etal..
2009b).
X-axis is dose (mg/kg-day), and y-axis is mean level of hemoglobin (g/dL).
B.7. HEMOGLOBIN-MALE RATS fLOVELESS ET AL.. 20091
Table B-6. Dose response data for hemoglobin in male rats (Loveless et al..
20091
Dose (mg/kg-d)
Number of
animals
Mean (g/dL)
Standard deviation
0
10
15.4
0.5
20
10
15.5
0.41
100
10
4.5
0.7
500
10
9.9
2.8
B-6
-------
Supplemental Information—PFHxA and Related Salts
Table B-7. Benchmark dose results for hemoglobin in male rats—non-constant
variance, BMR = 1 standard deviation fLoveless et al.. 20091
Model
Goodness
of fit
(p-value)
Test 3
(p-value)
AIC
1SD
Scaled residual
for dose group
near BMD
BMD
BMDL
Exponential 2
<0.0001
<0.0001
199.3758
9.377807
7.193218
-2.377990291
Exponential 3
<0.0001
<0.0001
239.5418
855.7401
0
0.802984088
Exponential 4
<0.0001
<0.0001
190.0784
6.631732
3.474481
-0.818335738
Exponential 5
0.071088
<0.0001
138.3961
70.68336
21.21839
-2.570261966
Hill
0.07109
<0.0001
138.3961
38.98926
21.0774
0.362134233
Polynomial (Poly 3)
<0.0001
<0.0001
239.7107
891.7542
383.4838
0.626839397
Polynomial (Poly 2)
<0.0001
<0.0001
239.7107
891.7546
383.3929
0.626839015
Power
<0.0001
<0.0001
239.7107
891.7544
383.3928
0.626839886
Both constant and nonconstant variance models failed to model the variance of the data.
Hemoglobin (HGB)
18
o
0 50 100 150 200 250 300 350 400 450 500
Dose
Figure B-4. Dose response data for hemoglobin in male rats (Loveless et al..
20091.
X-axis is dose (mg/kg-d), and y-axis is mean level of hemoglobin (g/dL).
B-7
-------
Supplemental Information—PFHxA and Related Salts
B.8. HEMOGLOBIN-FEMALE RATS fLOVELESS ET AL.. 20091
Table B-8. Dose response data for hemoglobin in female rats fLoveless et al..
20091
Dose (mg/kg-d)
Number of
animals
Mean (g/dL)
Standard deviation
0
10
15.6
0.7
20
10
15.8
0.8
100
10
15.6
0.4
500
9
13.3
0.9
Table B-9. Benchmark dose results for hemoglobin in female rats—constant
variance, BMR = 1 standard deviation (Loveless et al.. 2009)
Model
Goodness
of fit
(p-value)
Test 2
(p-value)
AIC
1SD
Scaled residual
for dose group
near BMD
BMD
BMDL
Exponential 2
0.214488
0.107799
89.79631
134.0618
104.1801
1.219441036
Exponential 3
0.50507
0.107799
89.16158
264.9174
126.3561
-0.026708969
Exponential 4
0.214489
0.107799
89.7963
134.0137
104.1803
1.219907658
Exponential 5
0.505122
0.107799
89.16147
266.3836
126.3586
-0.034100913
Hill
NA
0.107799
91.14613
115.7416
102.1374
-1.06893 x 10"06
Polynomial (Poly 3)
0.800193
0.107799
87.16312
268.4412
127.6129
-0.023130227
Polynomial (Poly 2)
0.800179
0.107799
87.16315
267.1194
127.8182
-0.018215642
Power
0.452941
0.107799
89.2806
372.7895
126.1226
0.000358346
Linear
0.259194
0.107799
89.41767
141.5272
111.6505
1.119316772
Bold row indicates the selected model and values.
B-8
-------
Supplemental Information—PFHxA and Related Salts
Frequentist Polynomial Degree 3 Model with BMR of 1 Std. Dev.
for the BMD and 0.95 Lower Confidence Limit for the BMDL
18
14
-------
Supplemental Information—PFHxA and Related Salts
Model
Goodness
of fit
(p-value)
Test 3
(p-value)
AIC
1SD
Scaled residual
for dose group
near BMD
BMD
BMDL
Hill
NA
0.863381
121.2797
-9999
0
-9999
Polynomial (Poly 3)
<0.0001
0.863381
143.4552
774.4171
120.0894
0.076434
Polynomial (Poly 2)
<0.0001
0.863381
143.4552
774.9496
132.5465
0.076318
Power
<0.0001
0.863381
143.4552
775.0219
106.6368
0.076271
Linear
<0.0001
0.863381
143.4552
775.3403
153.4792
0.076151
Both constant and nonconstant variance models failed to model the data.
Red Blood Ce
12
T 1
8 j*
6
¦——Data
4
2
0
0 10 20 30 40 50 60 70 80 30 100
Dose
Figure B-6. Dose response data red blood cells in male rats (Klaunig et al..
2015).
X-axis is dose (mg/kg-d), and y-axis is mean level of red blood cells (million/nL).
B.10. RED BLOOD CELLS-FEMALE RATS fKLAUNIG ET AL.. 20151
Table B-12. Dose response data for red blood cells in female rats (Klaunig et
al.. 20151
Dose (mg/kg-d)
Number of
animals
Mean
(million/pL)
Standard deviation
0
10
8.14
0.52
5
10
8.23
0.58
30
9
8.12
0.37
200
20
7.48
0.68
B-10
C
o
CL
w
cu
cc
-------
Supplemental Information—PFHxA and Related Salts
Table B-13. Benchmark dose results for red blood cells in female
rats—constant variance, BMR = 1 standard deviation fKlaunig et al.. 20151
Model
Goodness
of fit
(p-value)
Test 2
(p-value)
AIC
1SD
Scaled residual
for dose group
near BMD
BMD
BMDL
Exponential 2
0.896578
0.204474
88.37423
153.772
105.8225
-0.0212
Exponential 3
0.702156
0.204474
90.30213
168.4422
106.2748
0.001301
Exponential 4
0.896578
0.204474
88.37423
153.7727
105.8231
-0.02121
Exponential 5
NA
0.204474
92.30211
168.1679
30.68698
0.001229
Hill
NA
0.204474
92.28508
40.32119
31.54188
0.000759
Polynomial (Poly 3)
0.690261
0.204474
90.3147
175.8228
109.4569
0.000354
Polynomial (Poly 2)
0.69227
0.204474
90.31254
173.1861
109.4699
0.000719
Power
0.701613
0.204474
90.3027
169.0362
109.5006
0.000962
Linear
0.900552
0.204474
88.36539
155.595
109.1493
-0.01798
Bold row indicates the selected model and values.
Frequentist Linear Model with BMR of 1 Std. Dev. for the BMD
and 0.95 Lower Confidence Limit for the BMDL
10
8
c#
-------
Supplemental Information—PFHxA and Related Salts
B.ll. RED BLOOD CELLS-MALE RATS fCHENGELIS ET AL.. 2009B1
Table B-14. Dose response data for red blood cells in male rats fChengelis et
al..2009hl
Dose (mg/kg-d)
Number of
animals
Mean
(million/pL)
Standard deviation
0
10
8.89
0.32
10
10
8.84
0.281
50
10
8.88
0.69
200
10
8.17
0.593
Table B-15. Benchmark dose results for red blood cells in male rats—non-
constant variance, BMR = 1 standard deviation (Chengelis et al.. 2009b)
Model
Goodness
of fit
(p-value)
Test 3
(p-value)
AIC
1SD
Scaled
residual
for dose
group
near BMD
BMD
BMDL
Exponential 2
0.211487
0.046614
60.49036
113.155
66.64235
0.93824
Exponential 3
0.211488
0.046614
60.49034
113.324
66.64082
0.936908
Exponential 4
0.112482
0.046614
61.90217
63.88893
16.43649
1.512764
Exponential 5
0.123707
0.046614
61.75292
51.86424
17.01699
1.668664
Hill
NA
0.046614
61.38555
49.50692
16.01355
1.522475
Polynomial (Poly 3)
0.208929
0.046614
60.51469
115.2832
69.56043
0.914633
Polynomial (Poly 2)
0.208929
0.046614
60.51469
115.2939
69.56068
0.914397
Power
0.208929
0.046614
60.51469
115.2866
69.56292
0.914574
Linear
0.208929
0.046614
60.51469
115.2954
69.55948
0.914492
Both constant and nonconstant variance models failed to model the variance data.
B-12
-------
Supplemental Information—PFHxA and Related Salts
Red Blood Cell (RBC)
3
2
1
0
0 20 40 60 80 100 120 140 160 180 200
Dose
Figure B-8. Dose response data for red blood cells in male rats (Chengelis et
al..2009hl.
X-axis is dose (mg/kg-d), and y-axis is mean level of red blood cells (million/nL).
B.12. RED BLOOD CELLS-FEMALE RATS fCHENGELIS ET AL.. 2009B1
Table B-16. Dose response data for red blood cells in female rats (Chengelis et
al..2009bl
Dose (mg/kg-d)
Number of
animals
Mean
(million/iiL)
Standard deviation
0
10
8.62
0.338
10
10
8.53
0.696
50
10
8.32
0.491
200
10
7.93
0.43
Table B-17. Benchmark dose results for red blood cells in female
rats—constant variance, BMR = 1 standard deviation fChengelis et al.. 2009bl
Model
Goodness
of fit
(p-value)
Test 2
(p-value)
AIC
1SD
Scaled
residual
for dose
group
near BMD
BMD
BMDL
Exponential 2
0.819031
0.13452
61.22185
145.9541
94.47522
0.13169
Exponential 3
0.819031
0.13452
61.22185
145.9541
94.47455
0.13169
B-13
-------
Supplemental Information—PFHxA and Related Salts
Model
Goodness
of fit
(p-value)
Test 2
(p-value)
AIC
1SD
Scaled
residual
for dose
group
near BMD
BMD
BMDL
Exponential 4
0.828537
0.13452
62.86949
112.0384
27.37312
-0.13653
Exponential 5
0.527493
0.13452
63.2218
145.9511
16.32358
0.131694
Hill
NA
0.13452
64.90674
95.16729
22.04822
0.034663
Polynomial Degree 3
0.805881
0.13452
61.25422
148.2376
97.83829
0.128637
Polynomial Degree 2
0.805881
0.13452
61.25422
148.2285
97.83846
0.128826
Power
0.805881
0.13452
61.25422
148.2268
97.80444
0.128858
Linear
0.805881
0.13452
61.25422
148.2178
97.81736
0.129033
Bold row indicates the selected model and values.
10
8
50
100
Dose
150
Estimated Probability
Response at BMD
O Data
BMD
BMDL
200
Figure B-9. Dose response curve for the Exponential 5 model fit to red blood
cells in female rats fChengelis et al.. 2009bl.
X-axis is dose (mg/kg-d), and y-axis is mean level of red blood cells (million/nL).
B.13. RED BLOOD CELLS-MALE RATS fLOVELESS ET AL.. 20091
Table B-18. Dose response data for red blood cells in male rats fLoveless et al..
20091
Dose (mg/kg-d)
Number of
animals
Mean
(million/iiL)
Standard deviation
0
10
8.89
0.36
B-14
-------
Supplemental Information—PFHxA and Related Salts
Dose (mg/kg-d)
Number of
animals
Mean
(million/iiL)
Standard deviation
20
10
8.95
0.34
100
10
8.46
0.41
500
10
6.09
1.27
Table B-19. Benchmark dose results for red blood cells in male rats—non-
constant variance, BMR = 1 standard deviation fLoveless et al.. 20091
Model
Goodness of
fit
(p-value)
Test 3
(p-value)
AIC
1SD
Scaled
residual
for dose
group
near BMD
BMD
BMDL
Exponential 2
0.281567
0.991476
64.79171
52.64163
38.9282
0.746143
Exponential 3
0.376218
0.991476
65.03997
78.0673
43.76706
-0.23387
Exponential 4
0.281572
0.991476
64.79167
52.63495
38.92813
0.744847
Exponential 5
NA
0.991476
66.45382
91.34257
46.77432
-0.01779
Hill
NA
0.991476
66.44302
97.70618
94.33382
-0.01642
Polynomial (Poly 3)
0.291705
0.991476
65.36868
73.55976
45.76816
-0.24307
Polynomial (Poly 2)
0.291695
0.991476
65.36872
73.60792
45.76059
-0.24304
Power
0.341547
0.991476
65.16156
77.54244
46.28623
-0.27696
Linear
0.445951
0.991476
63.87203
59.08585
44.57007
0.743535
Bold row indicates the selected model and values.
B-15
-------
Supplemental Information—PFHxA and Related Salts
Frequentist Linear Model with BMR of 1 Std. Dev. for the BMD
and 0.95 Lower Confidence Limit for the BMDL
10
^ Estimated Probability
Response at BMD
O Data
BMD
BMDL
500
Figure B-10. Dose response curve for the Linear model fit to red blood cells in
male rats (Loveless etal.. 2009).
X-axis is dose (mg/kg-d), and y-axis is mean level of red blood cells (million/nL).
B.14. RED BLOOD CELLS-FEMALE RATS fLOVELESS ET AL.. 20091
Table B-20. Dose response data for red blood cells in female rats (Loveless et
al.. 20091
Dose (mg/kg-d)
Number of
animals
Mean
(million/iiL)
Standard deviation
0
10
8.34
0.43
20
10
8.53
0.52
100
10
8.32
0.27
500
9
6.85
0.63
Table B-21. Benchmark dose results for red blood cells in female
rats—constant variance, BMR = 1 standard deviation fLoveless et al.. 20091
Model
Goodness of
fit
(p-value)
Test 2
(p-value)
AIC
1SD
Scaled
residual
for dose
group
near BMD
BMD
BMDL
Exponential 2
0.21884
0.087567
57.58768
133.0328
102.9324
1.002642
Exponential 3
0.331861
0.087567
57.49047
238.0109
116.9504
-0.08346
Exponential 4
0.21884
0.087567
57.58768
133.0037
102.9322
1.002848
B-16
-------
Supplemental Information—PFHxA and Related Salts
Model
Goodness of
fit
(p-value)
Test 2
(p-value)
AIC
1SD
Scaled
residual
for dose
group
near BMD
BMD
BMDL
Exponential 5
0.331861
0.087567
57.49047
238.0095
116.9523
-0.08345
Hill
NA
0.087567
59.42671
113.2878
101.18
-2.4 x 10"06
Polynomial Degree 3
0.320732
0.087567
57.53481
261.7164
122.0763
-0.18275
Polynomial Degree 2
0.320735
0.087567
57.5348
261.8718
122.0761
-0.1845
Power
0.330478
0.087567
57.49587
243.0686
122.3971
-0.09028
Linear
0.268591
0.087567
57.17798
142.5548
112.3638
0.87655
Bold row indicates the selected model and values.
Frequentist Linear Model with BMR of 1 Std. Dev. for the BMD
and 0.95 Lower Confidence Limit for the BMDL
10
8
-------
Supplemental Information—PFHxA and Related Salts
Dose (mg/kg-d)
Number of
animals
Incidence
Percentage of incidence
10
60
0
0
50
60
3
5
200
70
12
17
The NOAEL was selected over the modeled data. This was based on dose spacing (5,10, and 50 mg/kg-d) that was
closer to 0 than the dose at which an effect was observed (LOAEL = 200 mg/kg-d). The NOAEL of 50 mg/kg-day is
more health protective than the BMDL of ~100 mg//kg/d and therefore chosen over modeled data.
Hepatocellular Necrosis - Female Rats
i
0.9
0.8
0.7
^ 0.6
tt 0.5
3
Figure B-12. Dose response data for hepatocellular necrosis in female rats
(Klaunig et al.. 2015).
X-axis is dose (mg/kg-d), and y-axis is percent incidence.
B.16. HEPATOCELLULAR HYPERTROPHY-FEMALE RATS fLOVELESS ET
AL.. 20091
Table B-23. Dose response data for hepatocellular hypertrophy in female rats
fLoveless et al.. 20091
Dose (mg/kg-d)
Number of
animals
Incidence
Percentage of incidence
0
10
0
0
20
10
0
0
100
n
0
0
500
10
5
50
B-18
-------
Supplemental Information—PFHxA and Related Salts
This data set is not considered appropriate for BMD modeling. The response in the high dose group (50%) is much
larger than the BMR and there was no response in all other dose groups.
Hepatocellular Hypertrophy, Female Rats
i
0.9
0.8
0.7
0 50 100 150 200 250 300 350 400 450 500
Dose
Figure B-13. Dose response data for hepatocellular hypertrophy in female rats
(Loveless et al.. 2009).
X-axis is dose (mg/kg-d), and y-axis is percent incidence.
B.17. HEPATOCELLULAR HYPERTROPHY-MALE RATS fLOVELESS ET AL..
20091
Table B-24. Dose response data for hepatocellular hypertrophy in male rats
fLoveless et al.. 20091
Dose (mg/kg-d)
Number of
animals
Incidence
Percentage of incidence
0
10
0
0
20
10
0
0
100
10
4
40
500
10
10
100
B-19
-------
Supplemental Information—PFHxA and Related Salts
Table B-25. Benchmark dose results for hepatocellular hypertrophy in male
rats—nested model BMR = 10% extra risk fLoveless et al.. 20091
Model
Goodness of fit
(p-value)
AIC
10% Extra risk
Scaled
residual
for dose
group
near BMD
BMD
BMDL
Dichotomous Hill
l
17.46023
85.47371
28.3855
-1.1 X 10"06
Gamma
0.999944
17.46046
70.57884
20.71965
0.00025
Log-Logistic
1
15.46023
85.49796
28.38513
2.91 x 10"07
Multistage Degree 3
0.997823
15.54154
59.28867
16.83509
-0.20131
Multistage Degree 2
0.902071
17.8587
46.58058
16.60448
-0.44287
Multistage Degree 1
0.391117
20.9779
18.16542
10.6581
-1.10904
Weibull
0.987025
17.51174
62.10697
19.73504
0.025832
Logistic
0.999997
17.46024
89.81641
41.88635
4.78 x 10"05
Log-Probit
1
17.46023
78.71963
26.71976
-7.5 x 10 12
Probit
0.999765
15.47853
71.58692
37.71366
0.012573
Bold row indicates the selected model and values.
Frequentist Multistage Degree 1 Model with BMR of 10% Extra
Risk for the BMD and 0.95 Lower Confidence Limit for the BMDL
i €>
Estimated Probability
Response at BMD
— — — Linear Extrapolation
O Data
BMD
BMDL
500
Figure B-14. Dose response curve for the Multistage Degree 1 model fit to
hepatocellular hypertrophy in male rats fLoveless et al.. 20091.
X-axis is dose (mg/kg-d), and y-axis is percent incidence.
B-20
-------
Supplemental Information—PFHxA and Related Salts
B.18. POSTNATAL (Fi) COMBINED RAT BODY WEIGHT ON PND 0
fLOVELESS ET AL.. 20091
Table B-26. Dose response data for postnatal (Fi) combined rat body weight
on PND 0 (Loveless et al.. 2009)
Dose (mg/kg-d)
Number of
animals
Mean (g)
Standard deviation
0
20
7.1
0.9
20
20
6.8
0.6
100
20
6.3
0.4
500
20
5.8
0.4
Table B-27. Benchmark dose results for postnatal (Fi) combined rat body
weight on PND 0—non-constant variance, BMR = 5% relative deviation
(Loveless et al.. 2009)
Model
Goodness
of fit
(p-value)
Test 3
(p-value)
AIC
5% relative deviation
Scaled
residual
for dose
group
near BMD
BMD
BMDL
Exponential 2
0.000613
0.257697
150.4828
154.17
126.6598
-2.10808
Exponential 3
0.000613
0.257697
150.4828
154.2311
126.6606
-2.10618
Exponential 4
0.417875
0.257697
138.3442
28.86879
18.04413
-0.30628
Exponential 5
0.417869
0.257697
138.3442
28.89287
18.02549
-0.3069
Hill
0.721731
0.257697
137.8148
20.37779
10.61916
0.013748
Polynomial (Poly 3)
0.000461
0.257697
151.0527
164.7639
137.82
-2.14368
Polynomial (Poly 2)
0.000461
0.257697
151.0527
164.763
137.8213
-2.144
Power
0.000461
0.257697
151.0527
164.7277
137.8381
-2.14471
Linear
0.000461
0.257697
151.0527
164.7482
137.8256
-2.14516
Bold row indicates the selected model and values.
B-21
-------
Supplemental Information—PFHxA and Related Salts
Frequentist Hill Model with BMR of 0.05 Rel. Dev. for the BMD
and 0.95 Lower Confidence Limit for the BMDL
8
=©
Estimated Probability
Response at BMD
O Data
BMD
BMDL
500
Figure B-15. Dose response curve for the Hill model fit to postnatal (Fi)
combined rat body weight on PND 0 (Loveless et al.. 2009).
X-axis is dose (mg/kg-d), and y-axis is mean body weight (g).
B.19. POSTNATAL (Fi) COMBINED MOUSE BODY WEIGHT (PHASE 2) ON
PND 0 flWAI AND HOBERMAN. 20141
Table B-28. Dose response data for postnatal (Fi) combined mouse body
weight (phase 2) on PND 0 flwai and Hoberman. 20141
Dose (mg/kg-d)
Number of litters
Mean (g)
Standard deviation
0
20
1.562
0.120
7
17
1.561
0.119
35
19
1.579
0.115
175
20
1.447
0.180
Table B-29. Benchmark dose results for postnatal (Fi) combined mouse body
weight (phase 2) on PND 0—constant variance, BMR = 5% relative deviation
(Iwai and Hoberman. 2014)
Model
Goodness
of fit
(p-value)
Test 2
(p-value)
AIC
5% Relative deviation
Scaled
residual
for dose group
near BMD
BMD
BMDL
Exponential 2
0.5476652
0.11356
-83.22986065
110.1988
72.6152
-0.18098
Exponential 3
0.6427413
0.11356
-82.21886799
162.9802
78.154
-0.00108
B-22
-------
Supplemental Information—PFHxA and Related Salts
Model
Goodness
of fit
(p-value)
Test 2
(p-value)
AIC
5% Relative deviation
Scaled
residual
for dose group
near BMD
BMD
BMDL
Exponential 4
0.5476662
0.11356
-83.22986423
110.2315
72.6152
-0.18210
Exponential 5
0.6427936
0.11356
-82.21893566
163.6378
78.15859
-0.00024
Hill
NA
0.11356
-80.21900716
80.26504
36.86639
0.37613
Polynomial (Poly 3)
0.971011
0.11356
-86.19475647
151.5619
80.06441
-0.00309
Polynomial (Poly 2)
0.8402282
0.11356
-84.08587922
140.6661
79.398
-0.018554
Power
0.6428503
0.11356
-82.21900901
172.0405
121.5756
2.26045 x 10"05
Linear
0.5601161
0.11356
-83.27482038
111.6004
75.16344
-0.16700
Bold row indicates the selected model and values.
Frequentist Polynomial Degree 3 Model with BMRof0.05 Rel.
Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
1.8
1.6
1.4
1.2
-------
Supplemental Information—PFHxA and Related Salts
B.20. POSTNATAL (Fi) COMBINED MOUSE BODY WEIGHT (PHASE 1) ON
PND 0 flWAI AND HOBERMAN. 20141
Table B-30. Dose response data for postnatal (Fi) combined mouse body
weight (phase 1) on PND 0 (Iwai and Hoberman. 2014)
Dose (mg/kg-d)
Number of litters
Mean (g)
Standard deviation
0
19
1.597
0.166
100
19
1.484
0.100
350
19
1.365
0.237
500
13
1.396
0.187
Table B-31. Benchmark dose results for postnatal (Fi) combined mouse body
weight (phase 1) on PND 0—non-constant variance, BMR = 5% relative
deviation (Iwai and Hoberman. 2014)
Model
Goodness of
fit
(p-value)
Test 3
(p-value)
AIC
5% relative deviation
Scaled
residual
for dose
group
near BMD
BMD
BMDL
Exponential 2
0.1454254
0.01314
-38.99028302
153.0166
106.9649
-0.9649
Exponential 3
0.1454276
0.01314
-38.99031401
152.9732
106.9641
-0.9649
Exponential 4
0.0502847
0.01314
-37.01452559
152.2536
22.05564
-0.9633
Exponential 5
NA
0.01314
-38.08812965
101.2731
78.25327
-0.6831
Hill
NA
0.01314
-38.08803429
100.2818
93.46723
-0.6814
Polynomial (Poly 3)
0.1237777
0.01314
-38.66793064
163.1923
116.9646
-0.9923
Polynomial (Poly 2)
0.1237777
0.01314
-38.66793064
163.1927
116.9612
-0.9923
Power
0.1237777
0.01314
-38.66793064
163.1924
116.9832
-0.9923
Linear
0.1237777
0.01314
-38.66793064
163.1923
117.1098
-0.9923
Both constant and nonconstant models failed to model the variance data.
B-24
-------
Supplemental Information—PFHxA and Related Salts
Combined Body Weight, Phase 1 on PND 0
£0.8
^ 0.6
0.4
0.2
0
0 50 100 150 200 250 BOO 350 400 450 500
Dose
Figure B-17. Dose response data for postnatal (FI) combined rat body weight
(phase 1) on PND 0 (Iwai and Hoberman. 2014).
X-axis is dose (mg/kg-d), and y-axis is mean body weight (g).
B.21. POSTNATAL (Fi) COMBINED MOUSE BODY WEIGHT (PHASES 1
AND 2) ON PND 0 flWAI AND HOBERMAN. 20141
Table B-32. Dose response data for postnatal (Fi) combined mouse body
weight (phases 1 and 2) on PND 0 flwai and Hoberman. 20141
Dose (mg/kg-d)
Number of litters
Mean (g)
Standard deviation
0
27
1.577
0.154
7
17
1.561
0.119
35
19
1.579
0.115
100
19
1.484
0.1
175
20
1.447
0.18
350
19
1.365
0.237
500
13
1.396
0.187
B-25
-------
Supplemental Information—PFHxA and Related Salts
Table B-33. Benchmark dose results for postnatal (Fi) combined mouse body
weight (phases 1 and 2) on PND 0—non-constant variance, BMR = 5% relative
deviation flwai and Hoberman. 20141
Model
Goodness
of fit
(p-value)
Test 3
(p-value)
AIC
5% relative deviation
Scaled
residual
for dose
group
near BMD
BMD
BMDL
Exponential
<0.0001
<0.0001
-113.7079045
142.9071
106.0466
-0.7543
Exponential
<0.0001
<0.0001
-113.7083052
143.1479
105.978
-0.7565
Exponential
<0.0001
<0.0001
-113.841077
96.60292
54.95459
-0.4165
Exponential
<0.0001
<0.0001
-114.9537952
124.9714
78.92795
-1.0491
Hill
<0.0001
<0.0001
-114.828486
120.2128
87.33994
-0.9886
Polynomial (Poly 6)
<0.0001
<0.0001
-113.1738885
151.8416
120.7674
-0.8470
Polynomial (Poly 5)
<0.0001
<0.0001
-113.1738881
151.8497
118.5011
-0.8469
Polynomial (Poly 4)
<0.0001
<0.0001
-113.1738881
151.8707
114.3476
-0.8473
Polynomial (Poly 3)
<0.0001
<0.0001
-113.1738834
124.918
89.12952
-0.8474
Polynomial (Poly 2)
<0.0001
<0.0001
-113.1738827
124.9218
89.12984
-0.8475
Power
<0.0001
<0.0001
-113.1738835
124.9172
89.12881
-0.8474
Linear
<0.0001
<0.0001
-113.1738818
124.9413
89.12913
-0.8478
Both constant and nonconstant models failed to model the variance data.
Combined Body Weight, Phase 1 and 2 on PND 0
1.8
—
0.6
0.4
0.2
0
0 50 100 150 200 250 300 350 400 450 500
Dose
Figure B-18. Dose response data for postnatal (Fl) combined rat body weight
(phases 1 and 2) on PND 0 flwai and Hoberman. 20141.
B-26
-------
Supplemental Information—PFHxA and Related Salts
X-axis is dose (mg/kg-d), and y-axis is mean body weight (g).
B.22. POSTNATAL (Fi) COMBINED MOUSE BODY WEIGHT (PHASE 2) ON
PND 4 flWAI AND HOBERMAN. 20141
Table B-34. Dose response data for postnatal (Fi) combined mouse body
weight (phase 2) on PND 4 flwai and Hoberman. 20141
Dose (mg/kg-d)
Number of litters
Mean (g)
Standard deviation
0
20
2.844
0.307
7
16
2.850
0.320
35
19
2.976
0.335
175
20
2.726
0.442
Table B-35. Benchmark dose results for postnatal (Fi) combined mouse body
weight (phase 2) on PND 4—constant variance, BMR = 5% relative deviation
(Iwai and Hoberman. 2014)
Model
Goodness of
fit
(p-value)
Test 2
(p-value)
AIC
5% relative deviation
Scaled
residual
for dose
group
near BMD
BMD
BMDL
Exponential
0.2719642
0.3216
62.91273812
169.116
79.86226
-0.259556406
Exponential
0.1915765
0.3216
64.01402032
171.7121
88.6223
0.000775748
Exponential
0.2719642
0.3216
62.91273812
169.1154
79.86194
-0.259556232
Exponential
0.1915772
0.3216
64.01401491
171.7578
88.62246
0.000965695
Hill
0.191581
0.3216
64.01398562
90.59589
37.60928
1.049750693
Polynomial (poly 3)
0.6280973
0.3216
60.04847961
167.4876
89.7897
-0.008262428
Polynomial (poly 2)
0.3908438
0.3216
62.1874628
164.7988
88.29103
-0.042810195
Power
0.1915812
0.3216
64.01398451
174.0668
106.2633
4.3382 x 10"06
Linear
0.2746896
0.3216
62.89279543
168.0092
81.96061
-0.247433913
Bold row indicates the selected model and values.
B-27
-------
Supplemental Information—PFHxA and Related Salts
Frequentist Polynomial Degree 3 Model with BMR of 0.05 Rel.
Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
3.5
3
2.5
QJ
00
Estimated Probability
O
Q_
1/1
z
1.5
1
0.5
0
Response at BMD
-------
Supplemental Information—PFHxA and Related Salts
Model
Goodness of
fit
(p-value)
Test 3
(p-value)
AIC
5% relative deviation
Scaled
residual
for dose
group
near BMD
BMD
BMDL
Exponential
0.0805258
0.01040
90.85553976
84.40685
60.76206
-0.0404
Exponential
0.0358662
0.01040
92.22063699
59.93938
28.42237
0.3237
Exponential
NA
0.01040
90.99580479
113.1456
55.45824
-0.6168
Hill
NA
0.01040
90.99580314
102.7811
95.12445
-0.6171
Polynomial (poly 3)
0.064938
0.01040
91.28582701
94.9049
70.88712
-0.1219
Polynomial (poly 2)
0.064938
0.01040
91.28582701
94.90514
70.88731
-0.1219
Power
0.064938
0.01040
91.28582692
94.90368
70.88898
-0.1219
Linear
0.064938
0.01040
91.28582698
94.90395
70.88785
-0.1220
Combined Body Weight, Phase 1 on PND 4
1
0.5
o
0 50 100 150 200 250 300 350 400 450 500
Dose
Figure B-20. Dose response data for postnatal (FI) combined rat body weight
(phase 1) on PND 4 flwai and Hoberman. 20141.
X-axis is dose (mg/kg-d), and y-axis is mean body weight (g).
B.24. POSTNATAL (Fi) COMBINED MOUSE BODY WEIGHT (PHASES 1 AND
2) ON PND 4 flWAI AND HOBERMAN. 20141
Table B-38. Dose response data for postnatal (Fi) combined mouse body
weight (phase 1) on PND 4 (Iwai and Hoberman. 2014)
Dose (mg/kg-d)
Number of litters
Mean (g)
Standard deviation
0
38
2.902
0.387
B-29
-------
Supplemental Information—PFHxA and Related Salts
Dose (mg/kg-d)
Number of litters
Mean (g)
Standard deviation
7
16
2.85
0.320
35
19
2.976
0.335
100
19
2.771
0.248
175
20
2.726
0.442
350
17
2.256
0.650
500
11
2.382
0.482
Table B-39. Benchmark dose results for postnatal (Fi) combined mouse body
weight (phases 1 and 2) on PND 4—non-constant variance, BMR = 5% relative
deviation (Iwai and Hoberman. 2014)
Model
Goodness of
fit
(p-value)
Test 3
(p-value)
AIC
5% relative deviation
Scaled
residual
for dose
group
near BMD
BMD
BMDL
Exponential
0.1128908
0.2000
147.7836654
96.26436
71.43879
-0.0186
Exponential
0.0802807
0.2000
149.2060003
120.9626
73.66277
-0.3003
Exponential
0.1128924
0.2000
147.7836603
96.26532
71.43823
-0.0189
Exponential
0.4079141
0.2000
145.7744096
155.2176
102.9449
0.4071
Hill
0.3365496
0.2000
146.2553381
167.5058
144.5742
0.2848
Polynomial (poly 6)
0.110707
0.2000
147.8372526
103.2872
78.92902
-0.0985
Polynomial (poly 5)
0.110707
0.2000
147.8372543
103.3039
78.96729
-0.0986
Polynomial (poly 4)
0.110707
0.2000
147.8372526
103.2829
78.85834
-0.0984
Polynomial (poly 3)
0.110707
0.2000
147.8372498
103.2662
78.84649
-0.0983
Polynomial (poly 2)
0.110707
0.2000
147.8372526
103.2867
78.85151
-0.0985
Power
0.0672452
0.2000
149.6433133
118.4167
79.55287
-0.2561
Linear
0.110707
0.2000
147.8372526
103.2844
78.84044
-0.0985
Bold row indicates the selected model and values.
B-30
-------
Supplemental Information—PFHxA and Related Salts
Frequentist Exponential Degree 5 Model with BMRof 0.05 Rel.
Dev. for the BMD and 0.95 Lower Confidence Limit for the BMDL
3.5
3H jj
2.5
ID
E 2
o
CL
3 1.5
C
-------
Supplemental Information—PFHxA and Related Salts
Frequentist Nested Logistic Model with BMR of 0.01 Std. Dev. for
the BMD and 0.95 Lower Confidence Limit for the BMDL
0.14
Dose
Figure B-22. Dose response curve for the Nested National Center for
Toxicological Research model fit to perinatal mortality (phase 2) on PND 0-
21 flwai and Hoberman. 20141.
X-axis is dose (mg/kg-d), and y-axis is mortality.
B.26. PERINATAL MORTALITY (PHASE 1) ON PND 0-21 flWAI AND
HOBERMAN. 20141
Table B-41. Nested model summary for perinatal mortality (phase 1) on PND
0-21, BMR = 1% extra risk (Iwai and Hoberman. 2014)
Model type
Litter-specific
covariate
Intralitter
correlation
Goodness
of fit
(p-value)
AIC
BMD
BMDL
Nested Logistic
Yes
Yes
0.053
356.23
206.1
105.8
Yes
No
<0.0001
478.37
238.9
177.2
No
Yes
0.0593
353.33
201.7
98.61
No
No
<0.0001
477.04
233.1
162.7
The means of the data cannot be modeled (all goodness of fit p-value > 0.1) for phase 1 data; the data is not
amenable to BMD modeling.
B-32
-------
Supplemental Information—PFHxA and Related Salts
Perinatal Mortality Phase 1 on PND 4
0.6
Figure B-23. Dose response data perinatal mortality (phase 1) on PND 4flwai
and Hoberman. 2014).
X-axis is dose (mg/kg-d), and y-axis is percent incidence.
B.27. PERINATAL MORTALITY (PHASES 1 AND 2) ON PND 0-21 flWAI
AND HOBERMAN. 20141
Table B-42. Nested model summary for perinatal mortality (phases 1 and 2)
on PND 0-21, BMR = 1% extra risk flwai and Hoberman. 20141
Model type
Litter-specific
covariate
Intralitter
correlation
Goodness
of fit
(p-value)
AIC
BMD
BMDL
Nested Logistic
Yes
Yes
0.024
495.44
150.9
85.15
Yes
No
<0.0001
632.14
199.7
138.2
No
Yes
0.029
491.80
147.7
83.59
No
No
<0.0001
629.52
195.2
134.0
The means of the data cannot be modeled (all goodness of fit p-value > 0.1) for phases land 2 data; the data is not
amenable to BMD modeling.
B-33
-------
Supplemental Information—PFHxA and Related Salts
Perinatal Mortality Data Phases 1 and 2 on PND 0-21
0.6
Dose
Figure B-24. Dose response data perinatal mortality (phases 1 and 2) on PND
0-21 (Iwai and Hoberman. 2014).
X-axis is dose (mg/kg-d), and y-axis is percent incidence.
B.28. TOTAL THYROXINE (T4) IN MALE RATS - fNTP. 20181
Table B-43. Dose response data for thyroxine (T4) in male rats fNTP. 20181
Dose (mg/kg-d)
Number of
animals
Mean (g)
Standard deviation
0
10
4.26
0.461692538
62.5
10
3.4
0.730486139
125
9
2.933
0.483
250
10
2.9
0.521775814
500
10
2.37
0.322552321
1,000
10
1.77
0.547074035
Table B-44. Benchmark dose results for total thyroxine (T4) in male
rats—constant variance, BMR = 1 standard deviation fNTP. 20181
Model
Goodness of
fit
(p-value)
Test 2
(p-value)
AIC
1SD
Scaled
residual
for dose
group
near BMD
BMD
BMDL
Exponential
0.001553
0.2402
17.8629446
145.2153
266.23938
-0.59715
B-34
-------
Supplemental Information—PFHxA and Related Salts
Model
Goodness of
fit
(p-value)
Test 2
(p-value)
AIC
1SD
Scaled
residual
for dose
group
near BMD
BMD
BMDL
Exponential
0.001553
0.2402
19.8639886
145.2153
266.23938
-0.59715
Exponential
0.028228
0.2402
17.8629446
43.1832
131.51922
-1.09556
Exponential
0.028228
0.2402
15.9850469
43.1832
131.51922
-1.09556
Hill
0.121780
0.2402
16.4752017
25.96669
92.61432
-0.73657
Polynomial (poly 6)
0.000100
0.2402
18.9269281
261.7194
390.80563
-1.01211
Polynomial (poly 5)
0.000100
0.2402
17.8629446
240.1021
390.82343
-1.01211
Polynomial (poly 4)
0.000100
0.2402
15.8629568
240.1021
390.79722
-1.01211
Polynomial (poly 3)
0.000100
0.2402
17.8629446
240.1022
390.61877
-1.01211
Polynomial (poly 2)
0.000100
0.2402
17.8629436
240.1086
390.56516
-1.01211
Power
0.000100
0.2402
17.8629446
240.1066
390.5741
-1.01211
Linear
0.001553
0.2402
19.8639886
145.2153
266.23938
-0.59715
Bold row indicates the selected model and values.
5
4.5
4
3.5
to 3
1 2.5
CO
2 2
1.5
1
0.5
0
Figure B-25. Dose response curve for the Hill model fit to thyroxine (T4) in
male rats fNTP. 20181.
X-axis is dose (mg/kg-d), and y-axis is mean level of thyroxine (T4) (ng/dL).
Frequentist Hill Model with BMR of 1 Std. Dev. for the BMD and
0.95 Lower Confidence Limit for the BMDL
Estimated Probability
^—Response at BMD
O Data
BMD
BMDL
200
400
600
800
1000
B-35
-------
Supplemental Information—PFHxA and Related Salts
APPENDIX C. EVALUATION OF PFHXA
ELIMINATION
C.l. EVALUATION OF PFHXA ELIMINATION IN RATS AND MICE
Pharmacokinetic parameters were estimated separately for male and female rats and mice
using a hierarchical, Bayesian framework to allow for the partial pooling of time-course
concentration data across multiple studies. Data extracted from the studies described above were
fit to the following model formulation, which describes the absorption (when necessary),
distribution, and elimination phase of PFHxA through a two-compartment pharmacokinetic model:
Ct = absflag i(—Ai - Bi)e~kabs-lt + A#-"* + 8^ (C-l)
Here, / represents the ith compartment for PFHxA measurement (e.g., plasma, liver, kidney).
Ai and 6, represent the ratio of chemical mass going to each empirical compartment, normalized by
the central compartment volume, resulting in units of PFHxA concentration. For PFHxA
concentrations measured in the plasma (i.e., central compartment) following intravenous (i.v.)
exposure, absfiag,i\s set to zero to remove the absorption term.
Conventionally, each compartment with pharmacokinetic data is fit independently to
equation C-l and tissue-specific half-lives for each species and sex are derived from the estimated p,
i.e., ti/2,1 = ln(2)//?;. However, when a compound is in the elimination phase, (3 should be constant
across all tissues. To determine this PFHxA-specific /3 and use the time-course concentration data
from every study across multiple compartments, a partial pooling of data in a hierarchical Bayesian
framework assumes that, although /?, differs for each tissue, they are all sampled from a common
group distribution. Following completion of the Markov-chain Monte Carlo analysis, the top-level
posterior distribution of /3 is used to determine the median PFHxA half-life, with uncertainty, for
each species/sex. The remaining study-level coefficients are used to estimate the additional
pharmacokinetic values, for example, area under the curve (AUCmf), clearance (CL), volume of
distribution (Vdp).
Along with the half-life analysis, a separate distribution of CL = dose/AUCmf and Vdp = CL/(3
is generated for each experiment (study/route/dose/sex), where AUCmf is obtained by integrating
equation (C-l) from time = 0 to infinity, to yield
AUClm = £ + f - (C-2)
ai Pi kabs,i
Median and 5th and 95th percentiles of the distributions for ti/2,1, CL, and Vdp are then
pooled across each study/route/dose to calculate the species- and sex-dependent values.
C-l
-------
Supplemental Information—PFHxA and Related Salts
C.l.l. Mice
Data for male and female mice were obtained from Gannon etal. f20111 who evaluated the
pharmacokinetics after single oral doses of 2 and 100 mg/kg. Original data files were provided by e-
mail from Shawn Gannon, The Chemours Company, Wilmington, Delaware to Paul Schlosser, U.S.
EPA, Durham, North Carolina on January 23, 2020. Although the data for the 2 mg/kg dose
appeared appropriately censored below the dose-specific limit of quantification (LOQ), the 100
mg/kg data appeared to reach a plateau just above the corresponding LOQ (~0.25 |ig/g plasma), in
a concentration range for which clearance after the 2 mg/kg dose was quite rapid. EPA interpreted
this result as indicating an interfering background signal. For this reason, only data with measured
concentration >0.5 |ig/g plasma were used for the 100 mg/kg dose. The resulting statistics for the
elimination half-lives (90% confidence interval) are 2.8 hours (1.0-7.0 hours) and 6.7 hours
(2.2-16 hours) for females and males, respectively.
Female mouse data were from Daikin Industries (2010). who exposed groups of mice to 35,
175, or 350 mg/kg PFHxA by oral gavage and measured serum concentration attime-points up to
24 hours. Because three separate mice were analyzed at each time point, means and standard
deviations were calculated and used for statistical modeling. Data at the first time points with
concentrations below the lower limit of quantification (LLOQ) were assigned a value of LLOQ/V2
for the purpose of computing the means. Specifically, for the 24-hour time point, two of three
animals in the 175 and 350 mg/kg dose groups had results
-------
Supplemental Information—PFHxA and Related Salts
Gannon Male 1
2 mg/kg Gavage dose
Gannon Male 2
100 mg/kg Gavage dose
Gannon Female 1
2 mg/kg Gavage dose
Gannon Female 2
100 mg/kg Gavage dose
5 10 15 20 25
time [hrs]
Daikin Female 1
35 mg/kg Gavage dose
0 5 10 15 20 25
time [hrs]
Daikin Female 2
175 mg/kg Gavage dose
iO2!
[\
£
101-
" l\S
\\\
E
CT)
E
u
10°]
1 /
>1,
7/
E
u
c
o
10"1 ]
c
o
u
v s
u
rn
03
E
10-2 i
£
(0
jC
CL
10"31
Q.
o
i—i
i i r i
5 10 15 20 25
time [hrs]
Daikin Female 3
350 mg/kg Gavage dose
5 10 15 20 25
time [hrs]
—- 90% C.I.
— median
+ in vivo data
5 10 15 20 25
time [hrs]
Figure C-l. Fits of population pharmacokinetic model to data for male (top
row) and female (remaining rows) mice following 2-350 mg/kg oral exposure
PFHxA.
Source: Data from Gannon et al. (2011) and Daikin Industries (2010).
C-3
-------
Supplemental Information—PFHxA and Related Salts
C.1.2. Rats
PFHxA the following PK data for male and female rats were evaluated:
• Chengelis etal. (2009a): male and female Sprague-Dawley rats exposed once by intravenous
injection (i.v.; 10 mg/kg) or by single-day or Day 25 of repeated gavage (50,150, or
300 mg/kg). (i.v. data for males and females and oral data for males are provided in
published tables. Oral data for females were obtained by digitizing the plot of single-day
exposure data. The 25-day female rat data, however, were not digitized or used because the
digitization process has some uncertainty; reported dose-specific half-lives for females
were quite similar for the single- and 25-day studies, and results for males were similar
with and without the 25-day data.)
• Dzierlenga etal. (2019): male and female Sprague-Dawley rats exposed by i.v. (40 mg/kg)
or by gavage (40, 80, or 160 mg/kg; data from National Toxicology Program website).
• Gannon etal. (2011): male and female Sprague-Dawley rats exposed by gavage (2 or
100 mg/kg; data from study authors).
• Iwabuchi etal. (2017): male Wistar rats exposed by gavage (0.1 mg/kg; data from published
tables or digitized from figures).
The resulting statistics for the elimination half-lives, clearance values, and volumes of
distribution (with 90% confidence intervals) are listed in Table 5-3 (see Section 5.2.1, Approach for
Animal-Human Extrapolation of PFHxA Dosimetry).
Data for human PFHxA analysis were extracted from Nilsson etal. (2013) where PFHxA
concentrations were measured in the blood of ski wax technicians exposed to PFAS compounds
over the course of multiple ski seasons. Because timing of the initial PFHxA exposure and the
resulting absorption kinetics are unknown for this population, EPA fit a one-compartment infusion
pharmacokinetic model to the reported time-course data:
Here, / represents the ith ski wax technician and represents the time at which exposure
to PFHxA ends. All other model parameters are the same as described above for the rat and mouse
fits. Briefly, this model assumes a constant exposure to PFHxA throughout the ski season when time
is less than Once t,„/is reached, PFHxA is eliminated under a first order elimination assumption.
Similar to the methods described for the rat and mouse, /?, for each ski wax technician is
sampled hierarchically from a population distribution while all other parameters in the model are
fit only to the individual technician. Finally, to use limit of detection (LOD) data reported in this
C.2. EVALUATION OF PFHXA ELIMINATION IN HUMANS
(C-3)
C-4
-------
Supplemental Information—PFHxA and Related Salts
study, we implemented a left-censored likelihood function in the Bayesian inference model for
samples reported below the LOD (<0.05 ng/mL). This ensured that the likelihood function for these
data were sampled only from a probability distribution with an upper bound at the LOD.
Results for each ski wax technician are shown below following sampling of the technician-
specific posterior distributions. Technician half-lives (90% credible interval) are presented in the
panel for each technician with the population half-life determined to be 11.45 (6.06 - 21.21) days.
Technicians 1-8 represent data from the 2007-2008 ski season, when samples were taken late
enough in the spring to allow quantification of post-exposure clearance.
Population half-life (days): 11.45 (6.06 - 21.21)
half-life (days): 8.75 (3.21 - 17.06)
• Tec hi
A "fechl-cens
2 4 6 8 10
Months since Sept
half-life (days): 13.67 (6.87 - 25.77)
• "fech2
A "fech2-cens
2 4 6 8 10
Months since Sept
half-life (days): 8.63 (2.98 - 16.74)
10"1
d 10"4
!_
o
^ 10-7
>-
£ 10-io
10"13
:
• "fech3
A "fech3-cens
2 4 6 8 10
Months since Sept
101
10"
half-life (days): 9.88 (3.61 - 19.69)
• Tech4
A "fech4-cens
2 4 6 8 10
Months since Sept
half-life (days): 14.04 (5.10 - 59.02)
ioM
10"1
half-life (days): 12.84 (5.10 - 31.40)
half-life (days): 12.60 (4.36 - 42.52)
half-life (days): 13.56 (5.49 - 33.06)
S 10"3
3E 10"3
10"7
• tch5
~ "fech5-cens
2 4 6 8 10
Months since Sept
S lO"3
• -fech6
~ 'feche-cens
2 4 6 8 10
Months since Sept
10"1
10"3
10"5
10"7
10"9
• -fech7
~ 'fech7-cens
2 4 6 8 10
Months since Sept
»
• -fech8
~ Tech8-cens
2 4 6 8 10
Months since Sept
Figure C-2. Fits of human PFHxA data from ski-wax technician blood samples.
Blue circles represent data above LOD while black triangles are data samples reported at the LOD (<0.05 ng/mL).
90% credible intervals are illustrated with the light blue bands and dashed lines.
C-5
-------
Supplemental Information—PFHxA and Related Salts
APPENDIX D. QUALITY ASSURANCE FOR THE IRIS
TOXICOLOGICAL REVIEW OF PFHXA
This assessment is prepared under the auspices of the U.S. Environmental Protection
Agency's (EPA's) Integrated Risk Information System (IRIS) Program. The IRIS Program is housed
within the Office of Research and Development (ORD) in the Center for Public Health and
Environmental Assessment (CPHEA). EPA has an agency-wide quality assurance (QA) policy
outlined in the EPA Quality Manual for Environmental Programs (see CIO 2105-P-01.11 and follows
the specifications outlined in EPA Order CIO 2105.1.
As required by CIO 2105.1, ORD maintains a Quality Management Program, which is
documented in an internal Quality Management Plan (QMP). The latest version was developed in
2013 using Guidance for Developing Quality Systems for Environmental Programs (OA/G-1). A
National Center for Environmental Assessment (NCEA)/CPHEA-specific QMP also was 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.
The IRIS Toxicological Review of PFHxA is designated as Influential Scientific Information
(ISI) and is classified as QA Category A. Category A designations require reporting of all critical QA
activities, including audits. The development of IRIS assessments is done through a seven-step
process. Documentation of this process is available on the IRIS website:
https://www.epa.gOv/iris/basic-information-about-integrated-risk-information-system#process.
Specific management of PFAS assessments is documented in a Programmatic Quality
Assurance Project Plan (PQAPP). A PQAPP is developed using the EPA Guidance for Quality
Assurance Project Plans fOA/G-51 and the latest approved version is dated October 2021. All PFAS
assessments follow the PFAS PQAPP, and all assessment leads and team members are required to
receive QA training on the PFAS PQAPP. 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 PFAS Assessments
L-CPAD-0031652-QP-1-5
February 2023
Program Quality Assurance Project
Plan (PQAPP) for the Integrated Risk
Information System (IRIS) Program
L-CPAD-0030729-QP-1-5
June 2022
An Umbrella Quality Assurance
Project Plan (QAPP) for Dosimetry
L-CPAD-0032188-QP-1-2
December 2020
D-l
-------
Supplemental Information—PFHxA and Related Salts
Title
Document number
Date
and Mechanism-Based Models
(PBPK)
Quality Assurance Project Plan
(QAPP) for Enhancements to
Benchmark Dose Software (BMDS)
L-HEEAD-0032189-QP-1-2
October 2020
ICF-General Support of CPHEA
Human Health Assessment Activities
QAPP
L-CPAD-0031961-QP-1-2
April 2021
During assessment development, this project undergoes quality audits during assessment
development including:
Date
Type of audit
Major findings
Actions taken
August 2019
Technical System Audit
None
None
August 2020
Technical System Audit
None
None
July 2021
Technical System Audit
None
None
August 2022
Technical System Audit
None
None
During Step 3 and Step 6 of the IRIS process, the IRIS toxicological review is subjected to
external reviews by other federal agency partners, including the Executive Offices of the White
House. Comments during these IRIS process steps are available in the docketEPA-HQ-ORD-2021-
0561 on http://www.regulations.gov.
During Step 4 assessment development, the IRIS Toxicological Review of Perfluorohexanoic
Acid and Related Salts underwent public commentfrom February 2, 2022, to April 4, 2022.
Following this comment period, the toxicological review underwent external peer review by a
contractor-led panel performed by ERG from April 5,2022 to August 25,2022. The peer-review
report is available on the peer review website. All public and peer-review comments are available
in the docket EPA-HQ-ORD-2021-0561.
Prior to release (Step 7 of the IRIS process), the final toxicological review is submitted to
management and QA clearance. During this step the CPHEA QA Director and QA Managers review
the project QA documentation and ensure that EPA QA requirements are met
D-2
-------
Supplemental Information—PFHxA and Related Salts
APPENDIX E. SUMMARY OF PUBLIC AND
EXTERNAL PEER REVIEW COMMENTS AND EPA'S
DISPOSITION
The Toxicological Review of Perfluorohexanoic Acid and Related Salts was released for
public comment in February 2022. Public comments on the assessment were submitted to the U.S.
Environmental Protection Agency (EPA) by April 4, 2022. The Toxicological Review has also
undergone a formal external peer review in accordance with U.S. Environmental Protection Agency
(EPA) guidance on peer review fU.S. EPA. 20151. A public, external peer-review meeting was held
May 16 and 17, 2022, which included another opportunity for public comment. The external peer
reviewers were tasked with providing written answers to general questions on the overall
assessment approach, key conclusions, and areas of scientific controversy or uncertainty. A
summary of comments made by the external peer reviewers and public commenters, as well as
EPA's responses to these comments, are arranged by charge question. In many cases, the comments
of the individual reviewers have been synthesized and paraphrased for brevity (please consult the
final peer review report for the full text of the panel's comments: Peer Review Report! External
Peer Reviewers were asked to prioritize their comments to indicate their relative importance as
follows. The prioritization instructions are duplicated below from the IRIS PFHxA charge questions
to the peer reviewers, which can be found in the public EPA docket (EPA-HO-ORD-2Q21-0561):
• Tier 1: Necessary Revisions - Use this category for any revisions you believe are necessary
to adequately support and substantiate the analyses or scientific basis for the assessment
conclusions, or to improve the clarity of the presentation in the PFHxA Toxicological
Review.
• Tier 2: Suggested Revisions - Use this category for any revisions you encourage EPA to
implement to strengthen the analyses or scientific basis for the assessment conclusions, or
to improve the clarity of the presentation in the PFHxA Toxicological Review.
• Tier 3: Future Considerations - Use this category for any advice you have for scientific
exploration that might inform future work. While these recommendations are generally
outside the immediate scope or needs of the PFHxA Toxicological Review, they could inform
future reviews or research efforts.
Appendix E lists all Tier 1 recommendations and Tier 2 Suggestions from the external peer
reviewers organized by charge question. For Tier 3 Considerations, please refer to the external peer
review report linked above. Where public comments were made on topics raised by the external
peer reviewers, they are noted along with the external peer review comments. All Tier 1
E-l
-------
Supplemental Information—PFHxA and Related Salts
recommendations were implemented in this revised assessment, either through revision or
addition to the peer reviewed analyses or text. Tier 2 suggestions were considered in light of the
extent to which those suggestions would impact the conclusions or quantitative analyses of the
assessment, consistency across committee in raising the suggestion, and the level of effort to
implement For this assessment, all Tier 2 suggestions deemed to be impactful to the toxicity value
conclusions were implemented in this revised assessment. Additional public comments not raised
by the peer reviewers are included in a separate section at the end of each charge question section.
In many cases, the public comments have been synthesized and paraphrased for brevity, both in
this Appendix and in the summary, document provided as a courtesy to the external peer review
panel. Please see docket fEPA-H0-0RD-2021-05611 for both this summary document and the full
text of the submitted public comments.
External peer reviewer and public comments regarding requests for additions of clarifying
text or editorial or grammatical corrections have been made throughout the assessment as
appropriate; these comments and responses have not been tracked in this Appendix.
E.l. CHARGE QUESTIONS 1 AND 2 - SYSTEMATIC REVIEW AND
DOCUMENTATION
1) The Toxicological Review for PFHxA describes and applies a systematic review protocol for
identifying and screening pertinent studies. The protocol is described in brief detail in
Section 1.2.1 (Literature Searching and Screening) and in full detail in Appendix A
(Systematic Review Protocol for the PFAS IRIS Assessments). Please comment on whether
the search strategy and screening criteria for PFHxA literature are clearly described. If
applicable, please identify additional peer-reviewed studies of PFHxA that the assessment
should incorporate1.
2) The Toxicological Review provides an overview of individual study evaluations and the
results of those evaluations are made available in the Health Assessment Workplace
Collaborative linked HAWC. Note that a "HAWC FAQ for assessment readers" document is
available (scroll to the bottom of the page, and the document is available for download
under "attachments") and is intended to help the reviewer navigate this on-line resource.
Data from studies considered informative to the assessment are synthesized in the relevant
health effect-specific sections, and study data are available in HAWC.
a. Please comment on whether the study confidence conclusions for the PFHxA studies are
scientifically justified and clearly described, considering the important methodological
features of the assessed outcomes. Please indicate any study confidence conclusions
that are not justified and explain any alternative study evaluation decisions.
'Newly identified studies (i.e., studies identified by EPA or the public that meet PECO criteria but were not
addressed in the external review draft, for example due to recent publication) will be characterized by EPA
and presented to the peer review panel. This characterization will focus on EPA's judgment of whether the
studies would have a material impact on the conclusions (i.e., identified hazards or toxicity values) in the
external review draft. The peer review panel is asked to review EPA's characterization and provide tiered
recommendations to EPA regarding which studies, if any, to incorporate into the assessment before finalizing.
E-2
-------
Supplemental Information—PFHxA and Related Salts
b. Results from individual PFHxA studies are presented and synthesized in the health
system specific sections. Please comment on whether the presentation and analysis of
study results are clear, appropriate, and effective to allow for scientifically supported
syntheses of the findings across sets of studies.
E.l.l. External Peer Reviewer Comments on Systematic Review and Documentation
For charge question 1, "all reviewers agreed that the search strategy and criteria were
appropriate and clearly described. One reviewer noted how inherently challenging it is to identify
pertinent studies with the increasing interest in PFAS, which has led to an increasing rate of new
publications. Several reviewers provided references to additional studies for EPA's consideration."
For charge question 2, "six of the seven reviewers agreed that the confidence conclusions for the
PFHxA studies were scientifically justified and clearly described. For example, one reviewer noted
that the visual presentation of the evaluation results for the animal studies was very effective and
found the use of interactive graphics to be very convenient." The seventh reviewer provided a Tier
1 Recommendation to improve the presentation. The report also noted that "reviewers generally
found the presentation and analysis of the study results as they appear in the health system-specific
sections to be clear but recommended several Tier 1 and Tier 2 revisions to improve the clarity and
accuracy of the presentation." These comments are described below.
Tier 1 Recommendations
Comment: EPA should add text describing the major reasons for excluding the 194 articles
during the screening process, as shown in Figure 2-1.
EPA Response: Studies are excluded if they do not meet all PECO criteria. During screening,
most studies are excluded because they do not meet any or only meet a few of the PECO criteria.
Thus, a single screened out study typically has multiple reasons for exclusion which is unwieldy to
document, especially at the title and abstract level when screening may be needed for thousands of
studies. Some of the studies that did not meet all PECO criteria were considered to have potentially
relevant supplemental information. In these instances, tags were added (if not already present) to
indicate the type(s) of potentially relevant supplemental information and can be visualized using
the interactive HAWC literature tag tree available by clicking the following link:
https://hawc.epa.gov/lit/assessment/100500070/references/visualization/. A sentence was
added to Section 2.1 to clarify that excluded studies "did notmeetthe PECO and did not contain
potentially relevant supplemental information."
Comment: EPA should add several sentences to Section 1 that describe the in-press paper
EHP (DOI 10.1289/EHP 10343) shown in EPA's slides during the May 16, 2022, peer review. In
particular, the reviewer noted that the evidence maps illustrating how EPA is going to synthesize
evidence across the PFAS compounds would be a good addition to the text
EPA Response: A brief description of the EHP paper was added to Section 4.1, before Table
4.1 that provides an overview of health effects that have been described for several other recent
EPA PFAS assessments.
E-3
-------
Supplemental Information—PFHxA and Related Salts
Comment: EPA should update HAWC for PFHxA to include assessments/evaluations of
recent studies that will be considered in finalizing this Toxicological Review.
EPA Response: The date of the last literature search used for the Toxicological Review
(April 2022) was added to Section 2.1 and in HAWC. Updates to the literature incorporated into the
assessment are reflected in a separate document posted to the docket f"EPA-HO-ORD-2Q21-0561-
0019"! and provided to the peer reviewers. This document describes the consideration of the
studies deemed relevant based on the methods laid out in the protocol and documents the
justification for the subset of those incorporated into the revised assessment.
Comment: EPA should expand the discussion in Section 1.2.4 (or an additional section) on
the use of low confidence studies to support mechanistic evidence when the mechanistic evidence
is used across health effects.
EPA Response: Additional text was added to Section 1.2.4 on the use of low confidence
studies to support coherence of mechanistic findings.
Comment: In Table 3-28, EPA should include the results of two high confidence studies that
did not report significant changes to histopathology as the inclusion of only the one study with
significant effects is being highlighted, paints an incomplete picture.
EPA Response: The thyroid histopathology data from NTP f20181 and Klaunig etal. f20151
were added to the relevant table in Section 3.2.5.
Comment: To clarify how decisions were made for each health endpoint, EPA should add a
brief section on the considerations used in evaluating study quality and summarize the basis for
assignments. Inclusion of this information solely within the HAWC template does not enable the
reader to readily identify the basis for judgments about individual studies or the rationale behind
the assignments.
EPA Response: Additional text was added to Section 1.2.2 that describes the study
evaluation for the epidemiology and animal toxicology studies. Readers are referred to the Protocol
for a detailed description of the study evaluation approach for both human epidemiology and
animal toxicology studies (Appendix A, Sections 6.2 and 6.3, respectively).
Comment: EPA should enumerate the adaptations made to the structured evaluation
considerations first introduced by Hill (1965).
EPA Response: Additional text was added to Section 1.2.4 that describes the specific
modified Hill considerations that are applied for IRIS Assessments. Readers are referred to the
Protocol (Appendix A, Section 9) for detailed descriptions of the considerations and the application
during evidence synthesis and integration.
Tier 2 Suggestions
Comment: EPA should summarize key points for other EPA PFAS reviews so a user of the
IRIS materials could see similarities and differences in this family of related chemicals. The
reviewer noted that users of the IRIS documents will usually be addressing mixtures of these
E-4
-------
Supplemental Information—PFHxA and Related Salts
compounds in the field, therefore, a common summary in one place would help the user community
coordinate the information.
EPA Response: Table 4-1 has been added to the assessment (see Section 4.1) to facilitate
comparisons of toxicity hazard conclusions across EPA PFAS assessments.
Comment: In the systematic review protocol in Appendix A (Table 5-2), EPA should clarify
why dam health (e.g., weight gain, food consumption) was only considered in "Developmental" and
not in "Reproductive" or tied to the specific effect on dam health observed (e.g., weight gain as an
endpoint).
EPA Response: Although effects on dam body weight were not specified as endpoint
grouping categories for animal toxicology studies in the PFAS protocol (Appendix A, Table 5-2),
these data were synthesized in the female reproductive health effects section (Section 3.2.7) as well
as considered when interpreting the developmental health effects (Section 3.2.2) in the
toxicological review. As stated on pg 5-3, lines 7-11, the endpoint groupings outlined in Tables 5-1
and 5-2: "are meantto serve as a starting place for consistency in presentation and analysis across
studies and assessments, although assessment-specific deviations are possible (e.g., depending on
the assessment-specific database of endpoints in the available studies or PFAS-specific
understanding of mechanistic relationships across outcomes)."
Comment: EPA should consider including a list of documents relevant to PFHxA risk
characterization that have been developed by state and international regulatory agencies in the
literature searches and in resulting databases.
EPA Response: State and international regulatory agency documents related to PFAS are
included in EPA literature searches and managed in HERO. EPA does not generally include a
description of non-EPA judgements in EPA assessments but does use these documents as a
resource for the identification of key science issues and potentially relevant studies that may have
been missed by a database search at early stages of draft assessment development.
Comment: EPA should consider incorporating recently published studies in the
Toxicological Review.
EPA Response: Additional studies were considered for incorporation into the toxicological
review for PFHxA. Of the studies that were considered, a subset was prioritized for inclusion
depending on whether they were expected to inform critical data gaps. Additional details regarding
the studies that were prioritized for inclusion can be found in the docket (see EPA-HO-QRD-2021-
0561-00191.
Comment: For increased transparency and ease of reference, EPA should consider adding
the HAWC animal toxicity study evaluation figure to the main document in addition to including it
in the HAWC.
EPA Response: A copy of animal study evaluation heat map in HAWC has been added to
Section 2.2 of the toxicological review.
E-5
-------
Supplemental Information—PFHxA and Related Salts
Comment: For hepatic effects (Table 3-11), EPA should consider several revisions: 1)
Consider additional tables and/or figures to help readers visualize the coherence of liver
histopathology with liver weight effects since these results are only presented in separate tables in
the document; 2) reconsider whether to include decreases in bilirubin amongst the serum
biomarkers of hepatic injury cited in Table 3-11 based on the Loveless etal. f20091 and Hall et al.
f20121 studies; and 3) in characterizing the strength of this evidence, reconfirm that the significant
variability of responses across studies and sexes was considered and weighed, as well as the
magnitude (frequently modest) and direction of change in the cases where there was a change in
one of the serum enzyme biomarkers (in many cases there were decreases).
EPA Response: The hepatic evidence is discussed considering the Hall etal. f20121 criteria
in Section 3.2.1 under the subheading "Considerations for Potentially Adaptive Versus Adverse
Responses". The criteria for considering the adversity of hepatic effects according to Hall is listed
and a summary of the hepatic findings are included below the Hall criteria where decreases in
bilirubin, globulin, and total protein are also summarized. The serum enzyme findings were
clarified to indicate the magnitude of change and a statement summarizing the different clearance
rates of PFHxA in rats (faster clearance in females than males) may underlie sex-specific
differences.
Comment: For developmental effects, EPA should consider revisions to further characterize
the mouse dose-response for decreases in postnatal body weight
EPA Response: Additional text was added to Section 3.2.2 clarify that the data from the
mouse study by Iwai and Hoberman (2014) representtwo separate experimental cohorts with
overlapping dose ranges. Although, in general, similar effects are observed across the two cohorts
there is some variability in the dose response pattern which, as now discussed in the assessment,
could be explained by normal variation across the control body weights in two experiments or a
survivor bias at the higher doses (e.g., higher mortality among low body weight pups in the higher
doses).
Comment: For hematopoietic effects, EPA should consider revisions to: 1) add a table
and/or figure to help readers visualize the coherence of these effects since these results are
presented in separate figures and tables in the document; and 2) add information on the results of
several chronic studies which are an important exception to the cited "consistent treatment related
effect on platelet levels."
EPA Response: Findings from the chronic study are made available in the draft on
hematological effects for all hematological findings that are described throughout Section 3.2.4.
The reviewer may have been referring to time points beyond 52 weeks of age that were not
considered based on as quantitative measures of hematology measures beyond 52 weeks may be
complicated by natural diseases occurring in rodents and test variability leading to decreased
sensitivity and increasing variability with the results fAACC. 19921. The collection of blood findings
are summarized in a visualization available in HAWC (linked here:
E-6
-------
Supplemental Information—PFHxA and Related Salts
https://hawc.epa.gOv/summary/data-pivot/assessment/100500070/pfhxa-animal-toxicology-
hematologv-effects-ervth /] as well as in the evidence integration section to help readers
understand the coherence of effects.
E.1.2. Public Comments on Systematic Review and Documentation
Comment: Several public commenters noted a lack of clarity regarding the literature search
and screening results, including inconsistencies in the screening results shown in HERO, HAWC,
and within the Toxicological Review, lack of clarity on how potentially relevant supplemental
information and newly identified studies would be incorporated in the Toxicological Review. Some
public commenters provided specific references or additional data that were not included in the
public comment draft of the PFHxA Toxicological Review.
EPA Response: EPA has taken several steps to clarify the literature search and screening
results for PFHxA that are now resolved and are available for viewing in HAWC and are available
using the following link:
https://hawc.epa.gOv/lit/assessment/100500070/references/visualization/
E.2. CHARGE QUESTION 3: NONCANCER HAZARD IDENTIFICATION
3) For each health effect considered in the assessment and outlined below, please comment on
whether the available data have been clearly and appropriately synthesized to describe the
strengths and limitations. For each, please also comment on whether the weight-of-
evidence decisions for hazard identification are scientifically justified and clearly described.
a. For hepatic effects, the Toxicological Review concludes the available evidence indicates
PFHxA likely causes hepatic effects in humans under relevant exposure circumstances.
This conclusion is based on studies of rats showing increased liver weight,
hepatocellular hypertrophy, increased serum enzymes, and decreased serum globulins.
The hepatic findings for PFHxA were similar for other PFAS and determined to be
adverse and relevant to humans.
i) Additional considerations influenced the hepatic effects hazard identification decisions.
Appendix A (Systematic Review Protocol for the PFAS IRIS Assessments) outlines the
human relevance of hepatic effects in animals that involve PPARa receptors as a key
science issue. To the extent supported by the PFHxA literature (and to a lesser extent,
literature for other PFAS), the Toxicological Review evaluates the evidence relevant to
the potential involvement of PPARa and non-PPARa pathways with respect to the
reported hepatic effects. The Toxicological Review ultimately concludes evidence from
in vivo (including genetic mouse models) and in vitro studies support a potential role
for multiple pathways operant in the induction of hepatic effects from PFHxA exposure,
but those pathways cannot be specifically determined. Please comment on whether the
conclusions regarding the available animal and mechanistic studies are scientifically
justified and clearly described. The hepatic findings for PFHxA were similar for other
PFAS and determined to be adverse and relevant to humans.
E-7
-------
Supplemental Information—PFHxA and Related Salts
b. For developmental effects, the Toxicological Review concludes the available evidence
indicates PFHxA likely causes developmental effects in humans under relevant exposure
circumstances. This judgment is based primarily on gestational exposure experiments
in mice, with supportive findings in rats exposed throughout gestation and lactation,
showing increased perinatal mortality, decreased offspring body weight, and delayed
eye opening. These effects are similar to those observed for other PFAS following
developmental exposure and were determined to be adverse and relevant to humans.
c. For hematopoietic effects, the Toxicological Review concludes the available evidence
indicates PFHxA likely causes hematopoietic effects in humans under relevant exposure
circumstances. This judgment is based on consistent findings, including decreased red
blood cells [RBCs], hematocrit, and hemoglobin, across study designs that, when
interpreted together, signifies PFHxA-related hematological effects such as anemia.
These findings were determined to be adverse and relevant to humans.
d. For endocrine effects, the Toxicological Review concludes the available evidence
suggests, but is not sufficient to infer, that PFHxA may cause endocrine effects in
humans under relevant exposure circumstances. This conclusion is based on some
evidence of thyroid effects based on hormone and histopathological changes in two rat
studies; however, the data is limited, lacking consistency across studies, and
histopathological changes may be explained by non-thyroid related effects
e. For all other potential health effects (i.e., renal, male and female reproductive, immune,
and nervous system), the Toxicological Review concluded the available evidence is
inadequate to assess whether PFHxA may cause effects in humans under relevant
exposure circumstances. In general, these conclusions were driven by sparse evidence
bases or data that were largely null.
E.2.1. External Peer Review Comments on Hepatic Effects
All seven reviewers were in agreement with the draft Toxicological Review that the data has
been "clearly and appropriately synthesized in order to describe the strengths and limitations of
the data" and that the weight-of evidence decisions used for hazard identification were
"scientifically justified and clearly described" for the hepatic effects from PFHxA. One reviewer also
noted, "importantly, recommendations of the Hall etal. T20121 paper were considered by the EPA
in assessing the adversity of observed hepatic effects," while another reviewer applauded inclusion
of this discussion and outcome stating that it was, "a compelling narrative, which compares point by
point the PFHxA responses against this guide concludes that these responses are adverse, human
relevant and of concern for such biological effects of necrosis". One reviewer noted that, "the IRIS
draft report included a section that discussed "evidence from other PFAS" ...was especially
important for interpreting the PFHxA results and by structural analogy that PFHxA would also
work via both PPAR alpha and non PPAR alpha response pathways. These comparisons showed
that the involvement of other non PPAR alpha receptors in the response to PFAS and by structural
relationship relevance for PFHxA." Several Tier 2 Suggestions are described below.
E-8
-------
Supplemental Information—PFHxA and Related Salts
Tier 1 Recommendations
Reviewers had no Tier 1 recommendations.
Tier 2 Suggestions
Comment: To improve clarity, EPA should revise the text (page 2-3) stating, "All outcomes
rated low confidence or higher were used for evidence synthesis and integration." The reviewer
commented that it may be unclear how this statement can be consistent with the statement on page
1-12 that "no low confidence studies were used in the evidence syntheses for PFHxA included in the
narrative," since low confidence studies may presumably have outcomes that would also be rated
as low confidence, which might be assumed to be included in evidence synthesis and integration
based on the first sentence cited.
EPA Response: The text in Section 1.2.4 has been edited. It now correctly states that "all
studies meeting PECO criteria were used for evidence synthesis and included in the narrative."
Comment: EPA should consider utilizing information on other PFAS compounds (e.g., PFBA)
to supplement and bolster the evidence consistent with the adversity of PFHxA-induced hepatic
effects.
EPA Response: In addition to the text in the Hepatic outcome Section (Section 3.2.1), the
assessment now includes an additional section in Section 4 including narrative description and
table summary (Table 4-1) comparing the hazard conclusions across published EPA PFAS
Assessments.
Comment: A reviewer noted an inconsistency in discussions of necrosis in rats and
suggested that EPA revise the wording to be consistent
EPA Response: The text has been edited to correct the inconsistency. The synthesis in
Section 3.2.1 now correctly states that necrosis was observed in females but not males.
Comment: In the "Evidence from other PFAS" section, EPA should consider emphasizing
that the observations of PPARa independent and dependent pathways from the four other PFAS are
consistent for both short-chain (e.g., PFBA) and long-chain (e.g., PFNA) substances, increasing the
plausibility that it also applies to PFHxA.
EPA Response: Although there is no evidence specifically challenging the role of PPARa in
PFHxA-mediated hepatotoxicity, based on PFHxA structural similarity with other PFAS, most
notably PFBA, it is reasonable to infer that PFHxA exposure in genetic mouse model systems would
elicit similar effects as structurally similar PFAS. Therefore, text was added to Section 3.2.1
specifically stating evidence from structurally similar PFAS, including PFBA, suggest PPARa
independent and dependent pathways also apply to PFHxA (Evidence from other PFAS
subheading).
Comment: A reviewer commented that while the interpretation of both epidemiologic
studies is reasonable, it is not clear why the potential for confounding is considered to be so
substantial without some indication of the rationale for expecting that serum PFHxA levels are
E-9
-------
Supplemental Information—PFHxA and Related Salts
associated with the confounding factors. EPA should consider including stronger reasoning as to
why such confounding would be expected. This comment applies to all health effect sections.
EPA Response: Text was added to human studies sections to further explain the concerns
for confounding. While not explicitly described in each section, studies were rated as "critically
deficient" for confounding and "uninformative" overall when there was no consideration (e.g.,
adjustment, exclusion, stratification) for potential confounders in heterogeneous populations.
There is particular concern with PFAS that lack of adjustment for age and sex would lead to
substantial bias given that these variables are associated with both PFAS exposure and most of the
outcomes of interest
E.2.2. Public Comments on Hepatic Effects
Comment: Public comments on EPA's conclusions made regarding hepatic effects in the
toxicological review were mixed. One commenter agreed with the overall conclusion and supported
EPA's position that the hepatic effects are adverse and relevant to humans. In contrast, two public
commenters expressed concerns about the human relevance of mechanistic support for hepatic
effects via PPARa mediated pathways and the adversity of the observed hepatic effects in animal
toxicity studies. Based on these concerns the commenters felt that the hepatic effects should be
considered to be inadequate to assess whether PFHxA may cause hepatic effects in humans.
EPA Response: The conclusions in the draft Toxicological Review regarding hepatic effects
were supported by the external peer review committee who provided tier 2 recommendations that
were addressed in the assessment The text includes evidence from other PFAS (short and long
chain) in models challenging the role of PPARa in PFHxA-mediated hepatotoxicity indicating roles
for PPARa dependent and independent pathways. The evidence is considered relevant to PFHxA
considering the PFAS evaluated are of similar carbon chain length and structure. The conclusions
were also supported by supplemental mechanistic evidence indicating the human PPARa binds and
is activated by PFHxA at similar or lower concentrations than rodent PPARa. Further, evaluation of
the available evidence was considered in the context of the Hall etal. (2012) criteria. While PFHxA
exposure does not clearly lead to cancer there is evidence for hepatic toxicity rather than
adaptation in rodents. The overall evidence is considered to be adverse and relevant to humans.
E.2.3. External Peer Review Comments on Developmental Effects
All seven reviewers agreed with the assessment conclusions for developmental effects. One
reviewer stated that, "The integration of available animal data, based on two high quality animal
studies (with three experiments) and on plausibility for human relevance, supports the finding that
PFHxA likely causes developmental effects in humans." Another reviewer stated that "The Agency's
logic was clear and transparent, and their conclusions scientifically justified." Several Tier 2
Suggestions are described below.
E-10
-------
Supplemental Information—PFHxA and Related Salts
Tier 1 Recommendations
Reviewers had no Tier 1 recommendations.
Tier 2 Suggestions
Comment: EPA should consider improving the discussion of human relevance such as by
adding information on the conserved biological processes or similarities in anatomy and physiology
between rodents and humans that EPA considers relevant to the observed developmental effects, or
whether rodents (particularly the mouse) have been shown to be good laboratory animal models
for assessing potential human developmental effects.
EPA Response: The text has been edited in Section 3.2.2. "These findings are interpreted as
relevant to humans in the absence of evidence to the contrary. This assumption is based on
Guidelines for Developmental Toxicity Risk Assessment fU.S. EPA. 19911." The assumption in the
EPA Guidelines is based on data for known developmental toxicants which have shown that animal
models are largely predictive of effects in humans.
E.2.4. Public Comments on Developmental Effects
Comment: Public comments on EPA's conclusions made regarding developmental effects in
the draft toxicological review were mixed. One commenter agreed with the overall conclusion that
PFHxA is likely to cause developmental effects in humans. In contrast, two public commenters
expressed concerns about strength of the evidence base to support the conclusion, specifically the
small evidence base (two animal toxicology studies) to inform developmental effects of PFHxA.
Concern about the adversity of decreased offspring body weight and these effects may be secondary
to maternal toxicity rather than a direct effect on development. Based on these concerns these
commenters felt that EPA should reconsider the conclusions for this health effect.
EPA Response: The conclusions in the draft Toxicological Review regarding developmental
effects were supported by the external peer review committee and retained in the revised
assessment. The evidence integration narrative in 3.2.2 discusses the potential impacts of maternal
toxicity on the interpretation of the animal evidence based and the rationale for why maternal
toxicity was not expected to be a primary driver of the observed developmental effects.
E.2.5. External Peer Review Comments on Hematopoietic Effects
Six of seven reviewers supported the overall conclusions of the hematopoietic effects
section, while one reviewer recommended clarifying, and possibly strengthening, the animal
evidence synthesis judgment (see Tier 1 Recommendation below). One reviewer commented that
"the weight-of evidence decisions used for hazard identification were scientifically justified and
clearly described and that when the rat studies are examined as a collective of study results, they
provide compelling evidence for PFHxA causing macrocytic anemia (low hemoglobin and large
RBC) and could be expected to cause serious harm in humans". While one reviewer stated that "the
findings are consistent with similar effects for multiple other PFAS and are reasonably determined
E-ll
-------
Supplemental Information—PFHxA and Related Salts
to be adverse and relevant to humans" a separate reviewer suggested "EPA should consider adding
additional information supporting the human relevance of hematopoietic effects observed in rats."
(see Tier 2 Suggestion below).
Tier 1 Recommendations
Comment: EPA should clarify why the animal evidence is "moderate" rather than "robust"
given that all four animal studies were assessed high confidence and there was agreement across
study findings and doses. The reviewer noted that this clarification would provide context for what
drives the "moderate" decision, and it will help to align with the conclusion that "the currently
available evidence indicates that PFHxA likely causes hematopoietic effects in humans."
EPA Response: Based on external peer review input and further review of the evidence, it
was determined that there is robust animal evidence for hematopoietic effects and the judgment
was changed in the assessment. This did not change the overall evidence integration judgement
based on based on identification of only one uninformative human study and uncertainty around
the human relevance of the rodent findings. Specifically, rodent hematological parameters differ
from humas by smaller erythrocytes, higher percentage of circulating reticulocytes (or
polychromasia), physiologic splenic hematopoiesis and iron storage, and more numerous and
shorter-lived erythrocytes and platelets fO'Connell et al.. 20151. These differences could explain the
possible regenerative response in the spleen and bone and the increase in reticulocytes (i.e.,
erythrogenesis and RBC turnover more rapid in rodent vs. human). Therefore, the currently
available evidence indicates that PFHxA likely causes hematopoietic effects in humans given
sufficient exposure conditions.
Tier 2 Suggestions
Comment: EPA should consider improving the discussion of human relevance such as by
adding information on the conserved biological processes between rats and humans that EPA
considers relevant to the observed hematopoietic effects, or whether rodents (particularly the
mouse) have been shown to be good laboratory animal models for assessing potential human
hematopoietic effects.
EPA Response: A discussion of the human relevance of hematopoietic effects (Section 3.1.4)
in rodents was added to the integration narrative. This additional discussion included background
information on the rodent model strain and origin (all animal models were obtained from the same
outbred population and supplier). Additional text also included a comparison between murine and
human hematological parameters. Specifically, rodent hematological parameters differ from humas
by smaller erythrocytes, higher percentage of circulating reticulocytes (or polychromasia),
physiologic splenic hematopoiesis and iron storage, and more numerous and shorter-lived
erythrocytes and platelets fO'Connell etal.. 2015). These differences could explain the possible
regenerative response in the spleen and bone and the increase in reticulocytes (i.e., erythrogenesis
and RBC turnover more rapid in rodent vs. human).
E-12
-------
Supplemental Information—PFHxA and Related Salts
E.2.6. Public Comments on Hematopoietic Effects
Comment: There were mixed responses to the conclusions made regarding hematopoietic
effects in the toxicological review. One commenter agreed with the overall conclusion and
recommended additional text be drafted discussing similarity of effects observed across related
PFAS. In contrast, two public commenters expressed concerns about strength of the evidence base
to support the conclusion, specifically citing the lack of mechanistic and informative human data
and questioning the adversity and biological significance of findings in animals. Based on these
concerns these commenters felt that EPA should reconsider the conclusions for this health effect.
EPA Response: EPA added additional discussion on the human relevance of the hematologic
effects (Section 3.2.4) observed in rodents and determined that, while there is robust evidence
available from the animal data, there is indeterminate human evidence and some residual
uncertainty around the human relevance of the observed effects; therefore, the evidence indicates
PFHxA likely causes hematopoietic effects, a judgement that was supported by external peer
review. Please see the responses above to peer reviewer comments.
E.2.7. External Peer Review Comments on Endocrine Effects
There were mixed responses from the committee on the conclusions made regarding
endocrine effects in the toxicological review. Three reviewers agreed with the conclusion that "the
currently available evidence suggests, but is not sufficient to infer, that PFHxA might cause
endocrine effects in humans under relevant exposure circumstances." One reviewer stated,
"Overall, the critical available data on endocrine effects are clearly and appropriately synthesized to
describe the strengths and limitations. In this reviewer's opinion, the weight-of-evidence decision
for endocrine effects is scientifically justified." In contrast, three reviewers recommended EPA
reconsider the conclusion on endocrine effects and their specifics comments are outlined in the Tier
1 Recommendations below. One reviewer did not comment on the overall conclusions or provide
other specific recommendations or suggestions in response to this charge question.
Tier 1 Recommendations
Comment: Two reviewers recommended that EPA strengthen the evidence integration
judgment and conclude that the available evidence indicates that PFHxA exposure is likely to cause
thyroid toxicity in humans given relevant exposure circumstances, primarily based on short-term
studies in rats reporting a consistent and coherent pattern of effects on thyroid hormones following
PFHxA exposure, but also drawing from the consistency of effects when considering evidence from
structurally related PFAS. A third reviewer recommended EPA re-examine the part of the statement
that says, "but is not sufficient to infer" that PFHxA could cause endocrine effects in humans.
EPA Response: Based on the Tier 1 recommendation from the external peer review
committee to reconsider the endocrine effects evidence in light of information on thyroid hormone
biology provided by the committee and findings for related PFAS, the overall evidence integration
judgement for endocrine effects was changed from evidence suggests but is not sufficient to infer
E-13
-------
Supplemental Information—PFHxA and Related Salts
to evidence indicates (likely). The evidence synthesis and integration text in Section 3.2.5 has also
been updated to reflect newly identified mechanistic evidence. Consistent with the
recommendations from the external peer review committee, decreased serum total T4 from the 28-
day rat study by NTP f20181 was advanced for dose response analysis.
Comment: One reviewer recommended EPA delete or provide better justification for the
statement, "some of these inconsistencies could be explained by differences in the test article (i.e.,
PFHxA vs. PFHxA salts)" since both the acids and salts will dissociate at biologically relevant pH to
form the identical anion.
EPA Response: This statement was deleted from the text.
Tier 2 Suggestions
Reviewers had no Tier 2 suggestions.
E.2.8. Public Comments on Endocrine Effects
Comment: One commenter disagreed with the conclusion on Endocrine effects, suggesting
that the integration judgement should be changed to evidence is inadequate, citing a weak evidence
base (i.e., inadequate human evidence and slight animal evidence).
EPA Response: For the reasons described in the EPA response to the tier 1 recommendation
from external peer reviewers above, EPA strengthened the overall evidence integration judgment
from evidence suggests to evidence indicates based on concluding that the animal evidence is
moderate rather than slight
E.2.9. External Peer Review Comments on All Other Potential Health Effects
All seven reviewers agreed with the conclusions of the draft Toxicological review for all
other potential health effects in the assessment. One reviewer stated they "agreed that the data has
been "clearly and appropriately synthesized in order to describe the strengths and limitations of
the data" and would in general agree with the comment that these endpoints did not have adequate
data to determine impact or not." Another reviewer commented that, "The Agency clearly
characterized both strength and weaknesses of these studies and the conclusion that there is
inadequate information to assess whether PFHxA affects these physiological domains is
scientifically justified." Their Tier 1 Recommendations and Tier 2 Suggestions are provided below.
Tier 1 Recommendations
Comment: EPA should improve transparency by including observations across other PFAS
compounds for the broad list of potential endpoints in this section, either by each endpoint listed in
charge question 3(e) or by providing an overall summary table of input from evaluation of other
PFOS compounds for these endpoints.
EPA Response: Table 4-1 has been added to the assessment (see Section 4.1) to facilitate
comparisons of toxicity hazard conclusions across EPA PFAS assessments.
E-14
-------
Supplemental Information—PFHxA and Related Salts
Comment: For renal effects, EPA should note reverse causality as a concern in the Seo et al.
(2018) study and provide a clearer justification for considering Zhang et al. (2019)as
"uninformative."
EPA Response: The evidence synthesis text was edited in Section 3.2.3 to clarify the
rationale underlying the study evaluation judgments for these studies. The potential for reverse
causality was added as a factor that decreases certainty to the evidence integration table.
Tier 2 Suggestions
Comment: EPA should consider re-examining the respiratory effects observed in the 28-day
NTP (2018) study and the 90-day Loveless et al. (2009) study for potential incorporation in the
Toxicological Review.
EPA Response: The nasal lesions described in the 28-day NTP f20181 study and the 90-day
Loveless etal. f20091 study were presumed to be driven by irritation stemming from inadvertent
aspiration of the gavage dose. On this basis, the results were considered by EPA to have unclear
toxicological relevance and not prioritized for synthesis and integration, however the results are
summarized in the animal literature inventory. This rationale is described and a link to the animal
literature inventory is provided in Section 3.2 of the Toxicological Review.
Comment: For renal effects, EPA should consider several revisions to Table 3-19: 1)
Consider noting the potential for reverse causality as a factor that decreases certainty for the
association of PFHxA with decrease in estimated eGFR; 2) consider adding "weak, no, or
inconsistent dose-response" as a factor that decreases certainty for organ weight; 3) as a factor that
decreases certainty, consider adding that "blood biomarkers of renal function were inconsistent";
and 4) as another factor that decreases certainty, consider adding difficulty in interpreting the
observed effects as adverse or non-adverse.
EPA Response: Edits were made to Table 3-19 (Renal profile table for PFHxA) to reflect
these suggestions.
Comment: For immune effects, EPA could improve clarity by moving asthma to its own
Pulmonary Effects section, since the one human asthma study examined was mostly of non-immune
mediated outcomes.
EPA Response: Asthma can be driven by both immune and respiratory effects. Since there is
no respiratory effects section in the PFHxA Toxicological Review and the study included evaluation
of immune related markers of asthma, the data from this study are retained in the immune effects
section.
Comment: In Table 3-37 in the nervous system effects section, EPA should indicate the
"preferred metric" for brain weight is absolute brain weight to be consistent with Table 3-31.
EPA Response: This text has been added to Table 3-37 in the nervous system effects section.
Comment: For nervous system effects, zebrafish studies are common for PFAS and should
be considered as useful supplemental data to inform evaluations. The reviewer also commented
E-15
-------
Supplemental Information—PFHxA and Related Salts
that this section could benefit from discussion of known impacts of other PFAS that might inform
design of future studies
EPA Response: Mechanistic information from newly identified studies, including two early
life stage zebrafish studies, have been added to the nervous system effects synthesis and
integration in section 3.2.9. While these data did not change the conclusions of the assessment, they
were included on the basis that they help to inform critical data gaps for nervous system effects.
E.2.10. Public Comments on All Other Potential Health Effects
Comment: One commenter disagreed with the conclusion that the evidence is inadequate
for renal effects. They suggested thathistopathological changes observed in the kidney
(i.e., papillary necrosis and tubular degeneration) in the chronic rat study by Klaunig (2015) are
adverse and should be used as the basis for derivation of the RfD. One commenter agreed with the
conclusions for immune and nervous system effects.
EPA Response: The conclusions in the draft Toxicological Review regarding other effects
were supported by the external peer review committee. The EPR also supported the conclusions in
the draft regarding the renal effects, although there were Tier 1 and 2 comments that were
addressed above. The EPA is aware of the report prepared by Luz etal. f20191. and the author
conclusion for an RfD based on papillary necrosis in female rats exposed to 200 mg/kg-d PFHxA
from the chronic study (Klaunig, 2015, 2850075). While the histopathological renal effects
observed by Klaunig etal. (2015) in females were the most significant effect, there were
inconsistencies across studies at similar observations times and doses and lack of coherence with
other renal findings. Therefore, the judgment of slight animal evidence was retained in the revised
assessment. The decision in the draft assessment that overall the renal evidence is inadequate is
similarly retained and renal endpoints were not advanced for RfD derivation.
E.3. CHARGE QUESTIONS 4 AND 5: NONCANCER TOXICITY VALUES DATA
SELECTION
4) For PFHxA, no RfC was derived. The study chosen for use in deriving the RfD is the Loveless
et al. (2009) one-generation reproductive toxicity study based on decreased offspring body
weight in rats exposed continuously throughout gestation and lactation to PFHxA sodium
salt via the dam. Is the selection of this study and these effects for use in deriving the RfD for
PFHxA scientifically justified and clearly described?
a. If yes, please provide an explanation.
b. If no, please provide an alternative study(ies) or effect(s) that should be used to support
the derivation of the RfD and detail the rationale for use of such an alternative.
c. As part of the responses in "a" or "b" above, please comment on whether the effects
selected are appropriate for use in deriving the RfD, including considerations regarding
E-16
-------
Supplemental Information—PFHxA and Related Salts
adversity (or appropriateness in representing an adverse change) and the scientific
support for their selection.
d. Given the lack of studies on inhalation exposure to PFHxA, no reference concentration
(RfC) is derived. Please comment on this decision.
5) In addition, for PFHxA, an RfD for less-than-lifetime ("subchronic") exposures is derived. No
"subchronic" RfC was derived. The same study and outcome were chosen for use in deriving
the RfD. Is the selection of this study and these effects for the derivation of the subchronic
RfD for PFHxA scientifically justified and clearly described?
a. If yes, please provide an explanation.
b. If no, please provide an alternative study(ies) and/or effect(s) that should be used to
support the derivation of the subchronic RfD and detail the rationale for use of such an
alternative.
c. As part of the responses in "a" or "b" above, please comment on whether the effects
selected are appropriate for use in deriving the RfD, including considerations regarding
adversity (or appropriateness in representing an adverse change) and the scientific
support for their selection.
d. Given the lack of studies on inhalation exposure to PFHxA, no "subchronic" RfC is
derived. Please comment on this decision.
E.3.1. External Peer Review Comments on Noncancer Toxicity Values Data Selection
For charge question 4, three reviewers concurred with the selection of the Loveless et al.
(2009) study and the effect of decreased offspring body weight as scientifically justified for
derivation of an RfD for PFHxA. Two reviewers recommended the NTP f20181 study with serum T4
as an endpointbe used as an alternative." These comments are described below. Two reviewers
noted that the topic is "outside of their expertise" with one stating that "the reasoning presented for
RfD derivation appeared sound" while the other declined to comment. "All reviewers who provided
comments agreed with the decision to not derive a reference concentration."
For charge question 5, "reviewers' comments on the charge questions related to the
derivation of the subchronic RfD were similar to those made for the chronic RfD." Four reviewers
"concurred with the selection of the Loveless etal. f20091 study and the selected effect as
scientifically justified for derivation of the subchronic RfD for PFHxA. As with the chronic RfD, one
reviewer "suggested using the NTP f20181 study with the endpoint of T4 suppression, although
they did not include this comment as a tiered recommendation." Two reviewers noted that the topic
is "outside of their expertise" with one stating that "the reasoning presented for RfD derivation
appeared sound" while the other declined to comment "All reviewers who provided comments
agreed with the decision to not derive a subchronic reference concentration."
E-17
-------
Supplemental Information—PFHxA and Related Salts
Tier 1 Recommendations
Comment: Two reviewers commented that the EPA should calculate an osRfD using the T4
endpoint from the NTP (2018). Of these, one reviewer recommended EPA "...also calculate this
value using the T4 endpoint from the NTP, 2018 study and to determine if this has significant
impact on the calculation of the RfD." This reviewer recommended that, if it does "have a significant
impact, then EPA should prioritize the use of the T4 endpoint" for the RfD. The second reviewer
recommended EPA use serum T4 as an endpoint should be used as an alternative to support the
derivation of an RfD.
EPA Response: Decreases in free and total T4 observed in male rats in the 28 day study by
NTP f20181. Total T4 was advanced for derivation of a POD for endocrine effects over free T4
because of concerns about the measurement method variability of the assay used to measure free
T4 (see Section 5.2.1). The POD for total T4 was higher than that selected for the developmental
osRfD (see Table 5-5). This endpoint was not considered for derivation of a lifetime toxicity value
due to the high level of uncertainty associated with use of a short-term study to protect against the
effects of a chronic, lifetime exposure. Therefore, total T4 was prioritized for subchronic candidate
value derivation.
Calculation of a candidate subchronic toxicity value for total T4 did not affect the overall
subchronic RfD selection for the Toxicological Review. As described in Section 5.2.1 (Selection of
Subchronic RfD and Confidence Statement), "a subchronic RfD of 5 x 10-4 mg/kg-day based on
decreased postnatal body weight is selected for less-than-lifetime exposure. The confidence in the
selected subchronic RfD is equivalent to that of the hepatic subchronic RfDs but lower than the
hematopoietic subchronic RfD. The developmental subchronic RfD is expected to be protective of
all life stages. The UFc (see Table 5-13) is lower than or equivalentto the other subchronic osRfDs
and the endpoint has the lowest PODhed (0.048 mg/kg-day, see Table 5-11). The decision to select
the developmental subchronic RfD was based on all of the available subchronic osRfDs in addition
to overall confidence and composite uncertainty for those subchronic osRfDs."
Tier 2 Suggestions
Comment: EPA should consider adding text to the organ-specific narrative for hepatic
effects and for developmental impacts, regarding adversity versus adaptation that may be relevant
for the study selection justification and health impacts to the human population. The reviewer
noted that these studies were either medium or high confidence studies with good annotation and
discussion of observations, and the quantitative estimates resulting from these calculations indicate
that these are sensitive hence protective endpoints for use in the RfD development The reviewer
also noted that these endpoint choices for the RfD are highly relevant for human populations.
EPA Response: Additional text was added to Sections 3.2.1, 3.2.2, 4.1 and 5.2.1 to clarify
how the observations for hepatic and developmental effects are expected to be adverse, potentially
E-18
-------
Supplemental Information—PFHxA and Related Salts
relevant to humans, and coherent across different layers of biology (i.e., from chemical -molecular
interaction) to organ level effects (e.g., increased liver weight and necrosis for hepatic effects).
E.3.2. Public Comments on Noncancer Toxicity Values Data Selection
Comment: One commenter agreed with the endpoints selected for POD derivation and a
second agreed with the decision to use high confidence animal studies.
EPA Response: The external peer review panel supported the selection of the endpoints for
POD derivation in the draft, although a tier 1 recommendation was made to add decreased T4 for
endocrine effects. The assessment has been updated to reflect the panel recommendations.
Comment: Two commenters disagreed with the selection of decreased postnatal body
weight for the RfD on the basis that these effects could be driven by maternal toxicity rather than a
direct developmental effect and the outcome is non-specific. One also suggested that the RfD should
be based on renal effects (papillary necrosis and tubular degradation) from the Klaunig (2015)
study.
EPA Response: EPA considered the available evidence base, key science questions, and
extensive peer review to develop and justify conclusions that are based on the PFAS protocol and
the IRIS Handbook that was favorably reviewed by the National Science Academy. Additional
justification and documentation for the rationale underlying this decision is provided in Section 5.
E.4. CHARGE QUESTIONS 6, 7, AND 8: NONCANCER TOXICITY VALUE
DERIVATION
6) EPA used benchmark dose modeling (U.S. EPA, 2012,1239433) to identify points-of-
departure (PODs) for oral exposure to PFHxA. Are the modeling approaches used, selection
and justification of benchmark response levels, and the selected models used to identify
each POD for toxicity value derivation scientifically justified and clearly described?
7) Appendix A identifies the potential for pharmacokinetic differences across species and
sexes as a key science issue and lays out a hierarchy for using relevant pharmacokinetic
data in extrapolating oral doses between laboratory animals and humans. Section 5.2.1
describes the various approaches considered and the rationale for the selected approach.
Given what is known and not known about the potential interspecies differences in PFHxA
pharmacokinetics, EPA used the ratio of human-to-animal serum clearance values assuming
the volume of distribution (Vd) in humans is equivalent to that in monkeys to adjust the POD
to estimate a human equivalent dose (HED) in the derivation of the respective RfDs.
a. Is applying the ratio of human-to-animal serum clearance values for PFHxA scientifically
justified and clearly described? If not, please provide an explanation and detail the
preferred alternative approach.
b. Does the Toxicological Review clearly describe the uncertainties in evaluating the
pharmacokinetic differences between the experimental animal data and humans?
E-19
-------
Supplemental Information—PFHxA and Related Salts
8) EPA has evaluated and applied uncertainty factors to account for intraspecies variability
(UFh), interspecies differences (UFa), database limitations (UFd), exposure duration (UFs),
and LOAEL-to-NOAEL extrapolation (UFl) for PFHxA.
a. Is uncertainty in the derivation of the toxicity values scientifically justified and clearly
described? Please describe and provide comments, if needed.
b. For uncertainty in interspecies differences (UFa), a value of 3 is applied to account for
remaining uncertainty in characterizing the pharmacokinetic and pharmacodynamic
differences between laboratory animals and humans after calculation of the HED. For
developmental and hematopoietic outcomes, the evidence base lacked chemical-and
species-specific information that would have been useful for informing the UFA; for
hepatic outcomes, however, available mechanistic and supplemental information was
useful for further evaluating the interspecies uncertainty factor. Some data indicate a
PPARa-dependent pathway that might support a UFa of 1. Evidence for non-PPARa
modes of action, however, is available in the PFHxA (and larger PFAS) database. Thus,
uncertainty remains regarding the potential differences in sensitivity across species due
to the involvement of both PPARa-dependent and-independent pathways. Further, data
are lacking to determine with confidence the relative contribution of each of these
pathways. As such, the Toxicological Review concludes the available data are not
adequate to determine if humans are likely to be equally or less sensitive than
laboratory animals with respect to the observed hepatic effects and that a value of
UFa = 3 is warranted to account for the residual uncertainty in pharmacodynamic
differences across species. Please comment on whether the available animal and
mechanistic studies support this conclusion and whether the analysis presented in the
Toxicological Review is scientifically justified and clearly described.
c. To inform uncertainty in intraspecies variability (UFh), the assessment evaluates and
considers the available evidence on potential susceptibility to PFHxA within different
populations or lifestages, including any potential human health impacts from early life
exposure. Are the available information and data appropriately considered and the
resultant UFh values scientifically justified and clearly described?
d. Are the provided rationales for the remaining uncertainty factors (UFl, UFd, UFs)
scientifically justified and clearly described? If not, please explain.
E.4.1. External Peer Review Comments on Noncancer Toxicity Value Derivation
For charge question 6, as summarized in the contractor report, "all reviewers who provided
responses to this charge question concurred that the approaches used, and the identification of
PODs were scientifically justified and clearly described. Faustman was impressed with the details
provided to identify the PODs for exposure to PFHxA and found the tables very easy to use." Two
reviewers "declined to comment, stating that this topic was outside of their area of expertise."
For charge question 7, "reviewers who provided responses to this charge question generally
concurred that the approach used for potential interspecies differences in PFHxA pharmacokinetics
was scientifically justified and clearly described. The same reviewers stated that the Toxicological
Review clearly described the uncertainties. Several reviewers provided recommendations for
E-20
-------
Supplemental Information—PFHxA and Related Salts
improving the clarity." These are described below. Two reviewers "declined to comment, stating
that this topic was outside of their area of expertise."
For charge question 8, "reviewers had mixed responses when commenting on the UFa of 3.
All reviewers who responded to Charge Question 8c concurred with a UFh of 10." Two reviewers
"declined to comment, stating that this topic was outside of their area of expertise." Four reviewers
"provided several comments related to the remaining uncertainty factors." These comments are
described below.
Tier 1 Recommendations
Comment: If models that do not provide adequate fit are included in the tables summarizing
benchmark dose modeling results for different endpoints (in Appendix B), they should be
marked/identified as such in these tables (e.g., by placing the model names and associated
estimates in parentheses)
EPA Response: Appendix B has been edited to provide additional clarity and transparency
on the modeling results and decisions. All model results are provided in the summary tables and
footnotes indicate whether data sets were determined to be inappropriate for modeling or when
there was inadequate fit for all models. In instances where there was adequate fit for one or more
models, bolded text indicates the selected model and associated values (explained in footnote).
Comment: In Table B-25, the selected model (indicated by bold type in the table and shown
in the proceeding figure) has neither the lowest AIC nor lowest BMDL. While an explanation of this
was provided by EPA during the peer review meeting, EPA should provide an explanation in the
modeling appendix.
EPA Response: After additional review and discussion with BMD modelers, it was
determined that this data set is not appropriate for BMD modeling because there is a single dose
group showing a high incidence response (50%) in contrast to no response in all other groups. On
this basis, a NOAEL approach was used for POD derivation for this endpoint. Tables 5-5 and 5-10
which show the PODs considered for derivation of the RfD and subchronic have been updated to
reflect this change in the toxicity value derivation approach. Because this endpoint was not
prioritized for derivation of the hepatic chronic or subchronic osRfD selections or the overall RfD or
subchronic RfD there was no effect on the derived toxicity values for the Toxicological Review.
Comment: The pharmacokinetic assumptions and parameterizations used by EPA in the
httk: High-Throughput Toxicokinetics package should be briefly mentioned/discussed in the
Toxicological Review (since httk is a publicly available EPA "product") and the context for making
comparisons with the assumptions and parameterizations of the pharmacokinetic modeling
performed for this Review should be clarified.
EPA Response: httk is a tool for rapid risk ranking to identify chemicals for which more in-
depth, chemical-specific analyses should be conducted. The httk project at EPA advises against
using this approach for this IRIS assessment, noting that the in vitro data used as inputs do not
E-21
-------
Supplemental Information—PFHxA and Related Salts
capture the large sex differences seen for many PFAS. On this basis, EPA determined that such an
evaluation of httk would not be an appropriate addition to this assessment
Comment: The reasoning behind using CL as opposed to tl/2 uses two conflicting lines of
reasoning and clarification is needed.
EPA Response: Clarifying text was added to Section 5.2.1 (Approach for Animal-Human
Extrapolation of PFHxA Dosimetry) that included explanation of EPA's guidelines on using
allometric scaling for deriving oral reference doses and that while there was not PBPK data
available for PFHxA, there was TK data.
Tier 2 Suggestions
Comment: Given the lack of sex differences observed in human studies, EPA should consider
clarifying the text implying that female human and male human equivalent doses will be calculated
on the basis of sex-specific PODs in animals.
EPA Response: Clarifying text was added to Section 5.2.1 (Approach for Animal-Human
Extrapolation of PFHxA Dosimetry).
Comment: Discussion of the Perez et al. study should note that some of the results were
called into question for PBFA and some of these issues could also apply to PFHxA. EPA should also
consider avoiding use of the Perez study as supplemental information, or if used, to include a caveat
per the additional studies referenced by the reviewer.
EPA Response: Clarifying text was added to Section 3.1.2 (Distribution in Humans).
Comment: The reference to slower elimination at higher concentrations (Dzierlenga et al.)
was noted as opposite the expectation of saturable renal absorption (mediated by Oatplal). The
reviewer noted that Han et al. mentions other transporters that have been tested for activity with
PFAS and suggested EPA consider adding a clarification such as: "While saturation of reabsorption
transporters would lead to decreased half-life, there are also transporters responsible for
elimination of PFAS to urine, and saturation of these transporters, such as Oat 1 and 0at3, could
lead to an increase in observed half-life and could thereby help explain the observations of
Dzierlenga etal."
EPA Response: Clarifying text was added to Section 5.2.1 (Approach for Animal-Human
Extrapolation of PFHxA Dosimetry).
Comment: If EPA decides to maintain a value of 3 for UFa, then a value of 10 should be
adopted for UFd.
EPA Response: As described in EPA's A Review of the Reference Dose and Reference
Concentration Processes (U.S. EPA. 2002). the interspecies uncertainty factor (UFa) is applied to
account for extrapolation of animal data to humans; it accounts for uncertainty regarding the
pharmacokinetic and pharmacodynamic differences across species. Although the pharmacokinetic
uncertainty is mostly addressed through the application of dosimetric approaches for estimating
human equivalent doses, there is residual uncertainty around the pharmacokinetics and the
uncertainty surrounding pharmacodynamics. Typically, a threefold UF is applied for this
E-22
-------
Supplemental Information—PFHxA and Related Salts
uncertainty in the absence of chemical-specific information. This is the case for the hematopoietic
and developmental endpoints. For the hepatic endpoints, known species differences exist between
rodent and human hepatic response to hepatotoxicant, particularly for effects mediated by PPARa
fHall etal.. 20121. Although the available evidence from PFAS structurally similar PFHxA were
available, experiments specifically challenging the role of PPARa in PFHxA mediated hepatotoxicity
were not available. Thus, based on the residual uncertainty surrounding the interspecies
differences in pharmacodynamics described above, a factor of 3 is applied to account for the
pharmacodynamic uncertainty of the UFA for all potential health effect consequences of PFHxA
exposure.
A UFd of 3 is applied because the evidence base for hepatic, hematopoietic, and
developmental endpoints included two subchronic studies and one chronic study in Sprague
Dawley rats and developmental/reproductive studies in Sprague Dawley rats and Crl:CDl mice.
Limitations, as described in U.S. EPA (20021 were used as the basis for a UFd = 3. These limitations
included a lack of informative human studies for most outcomes, subchronic or chronic toxicity
studies in more than one species, multigenerational study, a developmental neurotoxicity study.
Additionally, the data to inform effects on thyroid hormones is limited to a single short term study.
Additional justification has been provided in the draft in Section 5 to clearly document the
rationale for UF selection.
Comment: For the UFs for hepatocellular hypertrophy, EPA should consider including a
discussion of the specific study results justifying the specific UFs value proposed (i.e., 3 instead
of 10).
EPA Response: Additional text was added to Table 5-6 to clarify EPA's rationale for
selection of a UFs = 3 for hepatic effects. Briefly, hepatocellular hypertrophy observed in the
subchronic study is expected to represent a less severe adverse hepatic response than would be
expected to occur with chronic exposure. This is expected to reduce the uncertainty with use of a
subchronic study.
Comment: For the UFd, EPA should consider modifying Table 5-6 to delete "the dose
received by the pups is unclear and might be significantly less than that administered to the dams"
as a cited factor that in a meaningful way diminishes confidence in the database relevant to deriving
the RfD. Otherwise, since developing organism (e.g., pup) doses are commonly unknown, by EPA's
reasoning a UFd of 3 might automatically be applied any time the basis for an RfD or candidate RfD
is developmental effects. Moreover, it is not needed as the EPA cites other considerations that are
sufficient to support a UFd of 3.
EPA Response: This text was removed.
Comment: EPA should consider adding a more explicit description of the reasoning for
choosing a UFa of 3 instead of 1 or 10.
EPA Response: Clarifying text was added to Section 5.2.1 (Derivation of Candidate Toxicity
Values for the RfD) to support the rationale for the UFa = 3.
E-23
-------
Supplemental Information—PFHxA and Related Salts
Comment: EPA should consider revising the UFs of 1 to 10. The reviewer stated that the UFs
of 1 does not seem to consider the data showing that PFHxA exposure causes a reduction in serum
thyroid hormone, but there is little information beyond that Moreover, there is data suggesting that
eye-opening is delayed by PFHxA exposure, which is a potential thyroid endpoint, but this
relationship is not evaluated empirically. Considering this, the UFs of 1 does not appear to cover this
level of uncertainty for development
EPA Response: The reviewers concerns seem to be more directly related to lack of
additional data to inform the endocrine effects of PFHxA exposure. Limitations of the evidence base
are accounted for with the database uncertainty factor (UFd). Additional text was added to the UFd
justification in Table 5-6 to address the data gap described by the reviewer. Additionally, in the
current version of the draft, decreased serum free T4 was brought forward for dose response
analysis in support of the subchronic reference dose (RfD) and a UFs = 10 was applied to this
endpoint.
E.4.2. Public Comments on Noncancer Toxicity Value Derivation
Comment: Two commenters agreed with the data-driven HED approach used in the
toxicological review. One commenter disagreed with the selected approach on the basis that there
is insufficient evidence to support the data driven approach used in the assessment and suggest
that the BW3/4 approach should be used for calculation of the PODhed-
EPA Response: The HED approach in the draft Toxicological Review was supported by the
external peer review committee. Some clarifying text was added to the draft in response to some
tier 1 and tier 2 recommendations from the panel.
Comment: One commenter indicated that they supported selection of a UFs = 1 for
hematopoietic effects but suggested EPA reconsider a UFs = 3 for hepatic effects. Specifically, they
state that "the rationale for application of a UFs should consider whether adverse effects occur at a
lower dose with longer exposure, not only that adverse effects at a certain dose become more
severe with longer exposure" They suggest that a UFs = 1 may be appropriate for hepatic effects
similar to thatused by DWQI (2017) for PFOA. Another commenter suggested that it was
inappropriate to EPA should not use a subchronic study and an uncertainty factor for derivation of
a lifetime toxicity value when data are available from a chronic study.
EPA Response: As explained in the assessment, a UFs of 3 is applied to hepatocellular
hypertrophy for the purpose of deriving a lifetime RfD. Although the endpoint was derived from a
90-day subchronic study fLoveless etal.. 20091. which would typically warrant application of a
UFs = 10, there are some other sparse data that mitigate this uncertainty, to an extent. Specifically,
significant hepatocellular hypertrophy was not observed in the chronic study in male or female rats
(Klaunig etal.. 20151. However, a UFs = 1 was not applied as the evidence supports a pathway
where hepatocellular hypertrophy is an adverse event leading to more severe outcomes with
longer exposure durations, such as the necrosis that was observed in female rats in the chronic
study. Additionally, the highest dose levels used in the chronic study were at or below the LOAEL
E-24
-------
Supplemental Information—PFHxA and Related Salts
for this effect in the available subchronic studies (see Section 3.2.1). Thus, some uncertainty
remains and a UFs of 3 is applied.
A UFs of 3 is also applied to the hematopoietic endpoint (i.e., decreased RBCs) from the 90-
day subchronic study fChengelis etal.. 2009bl Specifically, a UFs lower than 10 was warranted as
more significant effects on RBCs were not observed after chronic exposure at the same PFHxA
doses (RBCs decreases of the same magnitude were observed at matched doses and sexes across
exposure durations see Section 3.2.4); however, uncertainty remains when considering the doses
tested in the chronic as compared to the subchronic study. Further, the subchronic study may
poorly predict a chronic exposure setting across multiple RBC life cycles (one cycle is ~60 days),
which could reflect cumulative effects as greater proportions of RBCs across stages are affected, or
possibly even reduced effects (compensatory responses) warranting a UFs higher than 1. Thus, a
UFs of 3 was applied.
Comment: One commenter suggested that the lack of informative human data should not be
accounted for in the UFd = 3 and indicates questionable human relevance of the animal findings.
EPA Response: EPA's A Review of the Reference Dose and Reference Concentration Processes
(U.S. EPA. 2002) states that the "database UF is intended to account for the potential for deriving an
underprotective RfD/RfC as a result of an incomplete characterization of the chemical's toxicity."
The document recommends "...the assessor should consider both the data lacking and the data
available for particular organ systems as well as life stages" when determining the value of the UFd.
Because reliable human data is the most relevant for assessing risk of an exposure to humans, the
lack of informative human studies presents an important gap in the evidence base.
Comment: One commenter agreed with the BMR selections. Another suggested that
additional support including references are needed to support the BMR justifications.
EPA Response: The modeling approach in the draft Toxicological Review was supported by
the external peer review committee. In addition, the assessment cites and follows EPA guidance on
BMR selection, which includes additional references and information.
Comment: One commenter disagrees with the decision to use a BMD approach for datasets
where there is a response only at one dose (e.g., the highest dose group) regardless of whether the
software states that models are viable suggesting that a NOAEL/LOAEL approach be used in these
instances. They specifically cite the following endpoints but indicate this list may not be exhaustive:
decreased RBC in female rats (Klaunig 2015); decreased hemoglobin in female rats (Klaunig 2015;
Loveless 2009); and increased hepatocellular hypertrophy in female rats (Loveless 2009)
EPA Response: All datasets were reviewed for the appropriateness of a BMD approach for
POD derivation. For decreased hemoglobin in male rats from Chengelis etal. (2009b) it was
determined that the data set was not appropriate for BMD modeling on the basis that the response
in the high dose group was much larger than the BMR and there was no response in all other dose
groups thus the NOAEL/LOAEL approach was applied to this endpoint (Appendix B, Section B.5).
For increased hepatocellular hypertrophy in female rats from Loveless etal. f20091 the dataset was
E-25
-------
Supplemental Information—PFHxA and Related Salts
not appropriate for BMD modeling on the basis that the response in the high dose group (50%) is
much larger than the BMR and there was no response in all other dose groups (Appendix B, Section
B.16). Additional text has been added to modeling appendix has been updated to provide clarify
situations in which a NOAEL/LOAEL approach would be preferred over a BMD approach (Appendix
B, Section B.3).
Comment: One commenter expressed concern that the osRfD for hematopoietic effects is
higher than the subchronic osRfD on the basis that, "It is not logical or supportable to conclude that
a specific toxicological effect will occur at a much lower dose from subchronic exposure than from
chronic exposure."
EPA Response: The EPA carefully evaluated the hematopoietic endpoints for toxicity value
derivation considering several factors that formed the basis for the overall osRfD and subchronic
osRfD. The specific toxicological endpoint, decreased red blood cells, was available from both the
chronic and subchronic studies, however it was noted that while the subchronic osRfD is lower
(~7-fold) than the chronic osRfD and both subchronic and chronic exposure designs and study
durations include the life cycle of a red blood cell (~60 days in rats), the subchronic study duration
may miss longer term (or even compensatory) effects on RBCs (i.e., regeneration) that would be
observable in a chronic study. Further, confidence in the quantification of the POD for the
subchronic osRfD is low given the POD was far below the NOAEL (50 mg/kg-d) and the osRfD is far
below toxicity values derived for the same finding from other subchronic studies suggesting some
underlying variability is driving the POD lower. These weaknesses and uncertainties in the ability
to reliably estimate toxicity values for the hematopoietic effects observed in the 90-day study by
(Chengelis etal.. 2009b) reduce the confidence in those estimates, which is reflected in two ways.
First, there is less confidence in the (Chengelis etal.. 2009b) candidate value for lifetime exposure,
as compared to the value based on the chronic study, and, although the candidate value from the
subchronic study is lower, the higher confidence data from the longer-term study is selected for the
lifetime osRfD. Second, for the subchronic osRfD, although data from the fChengelis etal.. 2009bl is
ultimately selected because the chronic study is not applicable and the POD from (Chengelis etal..
2009b) was much lower and more protective than PODs from the other subchronic studies, this
value was interpreted as medium-low confidence overall given the aforementioned uncertainties.
Based on this lower confidence determination, this subchronic osRfD is not used to support the
overall subchronic RfD. The EPA added this additional clarification to Section 3.2.1, and Section
5.2.1.
E.5. CHARGE QUESTION 9 AND 10: CARCINOGENICITY HAZARD
IDENTIFICATION AND TOXICITY VALUE DERIVATION
9) The Toxicological Review concludes that there is inadequate information to assess
carcinogenic potential for PFHxA and that this descriptor applies to oral and inhalation
routes of human exposure. Please comment on whether the available animal and
E-26
-------
Supplemental Information—PFHxA and Related Salts
mechanistic studies and the analysis presented in the Toxicological Review are scientifically
justified and clearly described.
10) Given the conclusion there was inadequate information to assess carcinogenic potential for
PFHxA (Charge Question 5), the Toxicological Review does not derive quantitative
estimates for cancer effects for either oral or inhalation exposures. Is this decision
scientifically justified and clearly described?
E.5.1. External Peer Review Comments on Carcinogenicity Hazard Identification and
Toxicity Value Derivation
"All reviewers concurred that the analysis presented in the Toxicological Review was
scientifically justified and clearly described."
Tier 1 Recommendations and Tier 2 Suggestions
Reviewers had no Tier 1 recommendations or Tier 2 suggestions.
E.5.2. Public Comments on Carcinogenicity Hazard Identification and Toxicity Value
Derivation
Comment: One commenter agreed with the conclusion that there is inadequate information
to assess carcinogenic potential of PFHxA, noting that carcinogenicity has only been evaluated in a
single study of one species (rat).
EPA Response: NA
E.6. ADDITIONAL COMMENTS
E.6.1. Additional External Peer Review Comments
Two reviewers "provided additional comments separately from their responses to the
charge questions. These included the following tiered comments not already covered in their
responses to charge questions."
Tier 1 Recommendations
Reviewers had no Tier 1 recommendations.
Tier 2 Suggestions
Comment: EPA should consider how data from other PFAS either support or differ from
PFHxA observations and how those could be explained by structure-activity relationships
(e.g., chain length vs. half-live observations) as well as how data from other model systems
(e.g., zebrafish) could help to fill data gaps.
EPA Response: Additional comparisons from other PFAS were added to the draft
toxicological review, specifically drawing from other observations in PFAS. Examples include the
discussion in the hepatic effects Section 3.2.1, subsection Considerations Related to Human
relevance, and Table 4-1 in Section 4.1. Note that information from the supplemental evidence that
E-27
-------
Supplemental Information—PFHxA and Related Salts
was determined to be potentially impactful to assessment conclusions (that may evaluate effects on
model systems other than animal and human) were captured and incorporated into the assessment
Comment: EPA should consider harmonizing the discussion of supporting evidence across
the different endpoints considered. For example, if structure-activity relationship information is
available for hepatic effects and the document includes text on what should be expected for PFHxA
based on observations for other PFAS, then under developmental effects, the document should state
whether similar structure-activity relationships could be considered or if such information is not
available
EPA Response: Similar to the comment above, the EPA included additional information from
other PFAS particularly in Section 4, and summarized in Table 4.1, into the assessment. The
evidence from other PFAS in models evaluating effects similar to PFHxA were considered.
Comment: EPA should consider adding context on reliability for the information presented
in Table 1-1 on the available physicochemical properties of PFHxA. The reviewer noted, for
example, that water solubility of ammonium vs. sodium salts varies five orders of magnitude and
stated that "clearly one of these values is wrong as once dissociated these should behave similarly."
Similarly, the reviewer noted that the same is true for the bioconcentration factor.
EPA Response: Text was added to Section 1.1.1 to clarify thatthe data in the table represent
both experimental and predicted values and that the predicted values may be less reliable.
Footnotes are used in Table 1-1 to indicate which values are experimental and which are predicted.
Comment: In the pharmacokinetics background (Section 3.1) of the Toxicological Review,
EPA should consider clarifying how "substantial binding" to serum proteins is defined (see page 3-
5, lines 6-7). The reviewer noted that PFHxA has been shown in in vitro studies to bind less strongly
than long-chain PFAS.
EPA Response: The text was edited to indicate the percent binding reported in the study
(>99% bound to serum albumin).
E.6.2. Additional Public Comments
Comment: One commenter suggested that EPA expand the background information section
on sources and relative contributions of PFHxA sources to better support the case for human
exposures. They also recommend that EPA present exposure information for PFHxA relative to
other PFAS compounds.
EPA Response: Comprehensive evaluation of exposure is outside the scope of IRIS
assessments. The background information described in Section 1.1 is an overview and is not
intended to provide a comprehensive description of the available information on PFHxA and
related salts, and information on human exposure is tagged as supplemental information during the
screening process.
Comment: One commenter suggested changes to the integration judgement language
(i.e., "the available evidence indicates that PFHxA exposure is likely to cause X effects in humans,
given relevant exposure circumstances") on the basis that it implies causation. Alternative language
E-28
-------
Supplemental Information—PFHxA and Related Salts
was suggested. Specifically, "the available evidence indicates that PFHxA could potentially result in
an increased risk of X in humans if exposure to exceeds XXX on a mg/kg-day basis."
EPA Response: The integration judgement language in the Toxicological Review is
consistent with the peer-reviewed IRIS Handbook.
Comment: One commenter suggested that EPA apply an additional uncertainty factor to
account for potential additive effects of exposure to multiple PFAS chemicals.
EPA Response: EPA applied uncertainty factors to account for five possible areas of
uncertainty as described in "Derivation of Candidate Values for the RfD," and in U.S. EPA (20021.
This assessment is specific to PFHxA and its related salts and the consideration of a potential
additive effect of exposure to multiple PFAS chemicals would not be appropriate for a scientific
document developed for one PFAS. The consideration of potential additive effects of exposure to
multiple PFAS would be part of the risk assessment and risk management activities such as the
application of this assessment (once finalized) along with other relevant assessments by risk
managers addressing human exposure to multiple PFAS. Thus, this is outside of the scope of the
IRIS Program.
E-29
-------
Supplemental Information—PFHxA and Related Salts
REFERENCES
AACC (American Association for Clinical Chemistry). (1992). Clinical pathology testing
recommendations for nonclinical toxicity and safety studies. Toxicol Pathol 20: 539-543.
Chengelis. CP: Kirkpatrick. IB: Myers. NR: Shinohara. M: Stetson. PL: Sved. DW. (2009a).
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.
Chengelis. CP: Kirkpatrick. IB: Radovskv. A: Shinohara. M. (2009b). A 90-day repeated dose oral
(gavage) toxicity study of perfluorohexanoic acid (PFHxA) in rats (with functional
observational battery and motor activity determinations). Reprod Toxicol 27: 342-351.
http://dx.doi.Org/10.1016/i.reprotox.2009.01.006.
Crump. KS. (1995). Calculation of benchmark doses from continuous data. Risk Anal 15: 79-89.
http://dx.doi.Org/10.llll/i.1539-6924.1995.tb00095.x.
Daikin Industries (Daikin Industries Limited). (2010). Oral (gavage) acute pharmacokinetic study of
PFH ammonium salt (ammonium salt of perflourinated hexanoic acid) in mice. Osaka, Japan.
Dzierlenga. AL: Robinson. VG: Waidvanatha. S: Devito. MI: Eifrid. MA: Gibbs. ST: Granville. CA:
Blvstone. CR. (2019). Toxicokinetics of perfluorohexanoic acid (PFHxA), perfluorooctanoic
acid (PFOA) and perfluorodecanoic acid (PFDA) in male and female Hsd:Sprague dawley SD
rats following intravenous or gavage administration. Xenobiotica 50: 1-11.
http://dx.doi.Org/10.1080/00498254.2019.1683776.
Gannon. SA: Tohnson. T: Nabb. PL: Serex. TL: Buck. RC: Loveless. SE. (2011). Absorption,
distribution, metabolism, and excretion of [l-14C]-perfluorohexanoate ([14C]-PFHx) in rats
and mice. Toxicology 283: 55-62. http://dx.doi.Org/10.1016/i.tox.2011.02.004.
Hall. AP: Elcombe. CR: Foster. TR: Harada. T: Kaufmann. W: Knippel. A: Kiittler. K: Malarkev. DE:
Maronpot. RR: Nishikawa. A: Nolte. T: Schulte. A: Strauss. V: York. Ml. (2012). Liver
hypertrophy: a review of adaptive (adverse and non-adverse) changes-conclusions from
the 3rd International ESTP Expert Workshop [Review], Toxicol Pathol 40: 971-994.
http://dx.doi.org/10.1177/0192623312448935.
Hill. AB. (1965). The environment and disease: Association or causation? Proc R Soc Med 58: 295-
300.
Iwabuchi. K: Senzaki. N: Mazawa. D: Sato. I: Hara. M: Ueda. F: Liu. W: Tsuda. S. (2017). Tissue
toxicokinetics of perfluoro compounds with single and chronic low doses in male rats. J
Toxicol Sci 42: 301-317. http://dx.doi.org/10.2131/its.42.301.
Iwai. H: Hoberman. AM. (2014). Oral (Gavage) Combined Developmental and Perinatal/Postnatal
Reproduction Toxicity Study of Ammonium Salt of Perfluorinated Hexanoic Acid in Mice. Int
J Toxicol 33: 219-237. http: //dx.doi.org/10.1177/1091581814529449.
Klaunig. IE: Shinohara. M: Iwai. H: Chengelis. CP: Kirkpatrick. IB: Wang. Z: Brunei'. RH. (2015).
Evaluation of the chronic toxicity and carcinogenicity of perfluorohexanoic acid (PFHxA) in
Sprague-Dawley rats. Toxicol Pathol 43: 209-220.
http: //dx.doi.org/10.1177/0192623314530532.
Loveless. SE: Slezak. B: Serex. T: Lewis. 1: Mukerii. P: O'Connor. TC: Donner. EM: Frame. S. R:
Korzeniowski. SH: Buck. RC. (2009). Toxicological evaluation of sodium perfluorohexanoate.
Toxicology 264: 32-44. http://dx.doi.Org/10.1016/i.tox.2009.07.011.
R-l
-------
Supplemental Information—PFHxA and Related Salts
Luz. AL: Anderson. TK: Goodrum. P: Durda. T. (2019). Perfluorohexanoic acid toxicity, part I:
Development of a chronic human health toxicity value for use in risk assessment Regul
Toxicol Pharmacol 103: 41-55. http://dx.doi.Org/10.1016/i.yrtph.2019.01.019.
Nilsson. H: Karrman. A: Rotander. A: van Bavel. B: Lindstrom. G: Westberg. H. (2013). Professional
ski waxers' exposure to PFAS and aerosol concentrations in gas phase and different particle
size fractions. Environ Sci Process Impacts 15: 814-822.
http://dx.doi.org/10.1039/c3em30739e.
NTP (National Toxicology Program). (2018). 28-day evaluation of the toxicity (C20613) of
perfluorohexanoic acid (PFHxA) (307-24-4) on Harlan Sprague-Dawley rats exposed via
gavage. Available online at
https://tools.niehs.nih.gov/cebs3/views/?action=main.dataReview&bin id=3879 (accessed
O'Connell. KE: Mikkola. AM: Stepanek. AM: Vernet. A: Hall. AD: Sun. CC: Yildirim. E: Staropoli. IF:
Lee. IT: Brown. DE. (2015). Practical murine hematopathology: a comparative review and
implications for research. Comp Med 65: 96-113.
U.S. EPA (U.S. Environmental Protection Agency). (1991). Guidelines for developmental toxicity risk
assessment Fed Reg 56: 63798-63826.
U.S. EPA (U.S. Environmental Protection Agency). (2002). A review of the reference dose and
reference concentration processes. (EPA630P02002F). Washington, DC.
https://www.epa.gov/sites/production/files/2014-12/documents/rfd-final.pdf.
U.S. EPA (U.S. Environmental Protection Agency). (2012). Benchmark dose technical guidance [EPA
Report], (EPA100R12001). 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). (2015). Peer review handbook [EPA Report] (4th
ed.). (EPA/100/B-15/001). Washington, DC: U.S. Environmental Protection Agency, Science
Policy Council, https: //www.epa.gov/osa/peer-review-handbook-4th-edition-2015.
------- |