vvEPA
EPA-823-P-18-001
Public Comment Draft
Human Health Toxicity Values for
Hexafluoropropylene Oxide (HFPO) Dimer Acid and Its
Ammonium Salt (CASRN 13252-13-6 and CASRN
62037-80-3)
Also Known as "GenX Chemicals"
This document is a Public Comment draft. It has not been formally released by the U.S.
Environmental Protection Agency and should not at this stage be construed to represent Agency
policy. This information is distributed solely for the purpose of public review.
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Human Health Toxicity Values for Hexafluoropropylene Oxide (HFPO)
Dimer Acid and Its Ammonium Salt (CASRN 13252-13-6 and CASRN 62037-
80-3)
Also Known as "Gen\ Chemicals"
Prepared by.
U.S. En\ironmeiilal Protection Auency
Office of Wilier (4.i<>4T)
llenllh and Ecological Criteria I)i\ ision
Washington. DC 2d4(iO
I-PA DiK'iinienl \nniher SZ.i-P-1X-001
\OVKiVlBKR2Ul8
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Disclaimer
This document is a public comment draft for review purposes only. This information is
distributed solely for the purpose of public comment. It has not been formally disseminated by
EPA. It does not represent and should not be construed to represent any Agency determination
or policy. Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Acknowledgments
This document was prepared by the Health and Ecological Criteria Division, Office of Science
and Technology, Office of Water (OW) of the U.S. Environmental Protection Agency (EPA).
OW leads for the assessment include Brittany Jacobs, PhD; Greg Miller, MS; and Jamie Strong,
PhD. OW scientists who provided valuable contributions to the development of this assessment
include Joyce Donohue, PhD and Barbara Soares, PhD. The Agency gratefully acknowledges the
valuable contributions of EPA scientists from the Office of Pollution Prevention and Toxics,
including Catherine Aubee; Amy Babcock, MPH, DABT, MRSB; Amy Benson, MS, DABT;
Tracy Behrsing, PhD; Chris Brinkerhoff, PhD; Tala Henry. PhD: and Laurence Libelo, PhD.
This document was provided for review by staff in the fol low i nu I ¦ I'A Program Offices and
Regions:
• Office of Water
• Office of Chemical Safety and Pollution he\ enlion. Office of Pesticide Programs
• Office of Chemical Safety and Pollution he\ ention, Office of Pollution Prevention and
Toxics
• Office of Chemical Safety and Pollution he\ enlion. Office of Science Coordination and
Policy
• Office of Land and Emergency Management
• Office of Air and Radiation. Office of Transportation and ,\ir Quality
• Office of Air and Radiation, Office of Air Oualilv Planning and Standards
• Office of Policy
• Office of Children's I leallh Protection
• Office of Research and De\ elopmenl
• Regions 1-1 n
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Contents
1.0 Introduction and Background 1
1.1 History of Assessment of GenX Chemicals 1
1.2 Uses of GenX Chemicals under TSCA 2
1.3 Occurrence 3
1.4 Other Assessments of GenX Chemicals 4
1.4.1 North Carolina Assessment 4
1.4.2 Report by the National Institute for Public I lealih and the Environment 4
2.0 Nature of the Stressor 5
2.1 Chemical/Physical Properties 5
2.2 Environmental Fate 8
2.2.1 Water 8
2.2.2 Air 8
2.2.3 Sediments and Soils 9
2.2.4 Biodegradation 9
2.2.5 Incineration 9
2.2.6 Bioaccumulation 9
2.3 Toxicokinetics 10
2.3.1 Absorption . . . 10
2.3.2 Distribution 12
2.3.3 Distribution during Gestation and Lactation.. 13
2.3.4 Metabolism 14
2.3 5 l-\civtion 14
2 3 (•> Pharmacokinetic Clearance and I lull-life Data 15
3.0 Problem l-'orimilalion 20
3.1 Conceptual Model 20
3.2 ()\ eralI Scientific Objecti\es 22
3.3 Methods 23
3.3.1 Literalure Search Strategy and Results 23
3.3.2 Study Screening Process and Study Evaluation 24
3.3.3 Approach lor Derivation of Reference Values 25
3.3.4 Measures of Effect 26
4.0 Study Summaries 27
4.1 Acute Toxicity Studies 27
4.2 Short-Term Toxicity Studies 29
4.3 Subchronic Toxicity Studies 32
4.4 Chronic Toxicity and Carcinogenicity Studies 34
4.5 Reproductive and Developmental Toxicity Studies 37
4.6 Other Studies 40
iii
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
4.6.1 Immunotoxicity Studies 40
4.6.2 Mechanistic Studies 40
4.6.3 Genotoxicity Studies 41
4.7 Mode of Action 42
5.0 Summary of Hazard 43
5.1 Hepatic 43
5.2 Hematological 44
5.3 Renal 45
5.4 Developmental/Reproductive 46
5.5 Immune System 46
5.6 Cancer 47
6.0 Dose Response Assessment 48
6.1 Study and Endpoint Selection 48
6.2 BMD Modeling 53
6.3 Dosimetric Adjustment of the Experimental Animal-Based POD to PODhed 53
6.4 Derivation of the RfD 54
6.4.1 Critical Study and Effect. ... 54
6.4.2 Uncertainty Factors 55
6.5 Subchronic RfD 58
6.6 Chronic RID 58
7.0 Discussion of I nccrlainties 59
7.1 Uncertainty and Variability 59
7.2 Composition of Test Substance 60
7.3 I se of l)ata-l)eri\ed l-Aliapolalion factors 61
7.4 I.imilcd Data on Carcinogenicity 61
7.5 l-flects on Bilirubin 61
7.(-> Susceptible Populations and l.ile Stages 62
8.0 References 64
Appendix A: Literature Search Strategy A-l
Appendix B: Systematic Review of DuPont/Chemours Submissions B-l
Appendix C: Acute and 7-l)av Study Summaries C-l
Appendix D: Genotoxicity Study Summary D-l
Appendix E: Benchmark Dose Modeling E-l
iv
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Figures
Figure 1. Structure of HFPO Dimer Acid and HFPO Dimer Acid Ammonium Salt 5
Figure 2. Conceptual Diagram for HFPO Dimer Acid and/or Its Ammonium Salt 21
Figure E-l. Plot of Incidence Rate by Dose with Fitted Curve for the Selected Model
Quantal-Linear Model for Single-Cell Necrosis in Male Mice; Dose Shown in
mg/kg/day E-2
Figure E-2. Plot of Incidence Rate by Dose with Fitted Curve for the Selected Model
Multistage 2° Model for Single-Cell Necrosis in Male Mice; Dose Shown in
mg/kg/day E-2
Figure E-3. Plot of Incidence Rate by Dose with Fitted Cui\ e lor the Selected Multistage
2° Model for Single-Cell Necrosis in Male Mice. Dose Shown in mg/kg/day E-4
Tables
Table 1. Chemical and Physical Properties of] II PO Dimer Acid and III PO Dimer Acid
Ammonium Salt 6
Table 2. Plasma Concentration in Crl CDHTCRj Mice at 2 I loin s after the First Gavage
Exposure to HFPO Dimer Acid Ammonium Suit" 11
Table 3. Clearance Times in Male and l-'emalc Rats and Mice I-'dIlowing a Single Oral
Dosea 15
Table 4. T1/2 Estimates IV0111 the lntia\ enons Injection in S|">iauuc I)awley Rats and
Cynomoluus Monkeys'1 . 17
Table 5. T1/2 Estimates in S|">iauuc Daw ley Rats and Crl CD 1 (ICR) Mice Exposed to a
Single Oral Dose of III PO Dimer Acid Ammonium Salta 18
Table 6 Mean Plasma Concentrations with Standard Deviations (SD) of Dosing with
III PO Dimer Acid Ammonium Salt lor at I .east 90 Daysa 19
Table 7 Summary of Study \()AI-I.S LOAELS 49
Table 8. Summary of Determination of PODhed 54
Table A-l. Summary of Core Database Search Results A-l
Table A-2. Database Search Strings A-2
Table A-3. Processes Used to Augment the Search of Core Databases for HFPO Dimer
Acid (13252-13-0) A-5
Table A-4. Processes Used to Augment the Search of Core Databases for HFPO Dimer
Acid Ammonium Salt (62037-80-3) A-7
Table A-5. Inclusion-Exclusion Criteria for HFPO Dimer Acid and HFPO Dimer Acid
Ammonium Salt A-9
Table B-l. Definition of Overall Quality Levels and Corresponding Quality Scores B-3
Table B-2. Types of Animal and In Vitro Toxicity Data B-5
v
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Table B-3. Data Evaluation Domains and Definitions B-6
Table B-4. Data Evaluation Domains and Metrics for Animal Toxicity Studies B-7
Table B-5. Data Evaluation Domains and Metrics for In Vitro Toxicity Studies B-8
Table B-6. Animal Toxicity Metrics with Greater Importance in the Evaluation and
Rationale for Selection B-10
Table B-7. In Vitro Toxicity Metrics with Greater Importance in the Evaluation and
Rationale for Selection B-10
Table B-8. Metric Weighting Factors and Range of Weighted Metric Scores for Animal
Toxicity B-12
Table B-9. Metric Weighting Factors and Range of Weighted Metric Scores for In Vitro
Toxicity Studies B-13
Table B-10. Scoring Example for Animal Toxicity Study with all Metrics Scored B-14
Table B-l 1. Scoring Example for Animal Toxicity Study with Some Metrics Not
Rated/Not Applicable B-l5
Table B-12. Scoring Example for In Vitro Study with all Metrics Scored B-16
Table B-13. Scoring Example for In I 'uro Study with Some Metrics Not Rated/Not
Applicable B-l 7
Table B-14. Serious Flaws that Won Id Make Animal Toxicity Studies Unacceptable B-18
Table B-15. Data Quality Criteria lor Animal Toxicity Studies B-21
Table B-16. Serious Maws that Would Make /// I uro Toxicity Studies Unacceptable B-36
Table B-17. Data Quality Criteria lor In I uro Toxicity Studies B-38
TableD-l (ienotoxicity Study Summary D-l
Tablel>l Single-Cell Necrosis in thcLher Selected for Dose-Response Modeling E-l
Table E-2 Summary of IJMI) Modeling Results for Single-Cell Necrosis in Male Mice E-l
Table E-3. Single-Cell Necrosis in the Liver Selected for Dose-Response Modeling E-3
Table E-4. Summary of BMI) Modeling Results for Single-Cell Necrosis in Male Mice E-3
Table E-5. Goodness of I-'it TaMe of BMD Modeling Results for Single-Cell Necrosis in
Male Mice E-4
vi
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Acronyms and Abbreviations
A/G Albumin to globulin
AIC Akaike information criterion
ALD Approximate lethal dose
ALP Alkaline phosphatase
ALT Alanine aminotransferase
AST Aspartate aminotransferase
BAF Bioaccumulation factor
BCF Bioconcentration factor
BBDR Biologically based dose-response
BMD Benchmark dose
BMDL Benchmark dose lower limil
BMR Benchmark response
BOD Biochemical oxygen demand
BUN Blood urea nitrogen
BW Body weight
CASRN Chemical Abstracts Service Registry Number
CFR Code of federal Regulations
CoA Coenzyme A
COV Coefficient of\ariation
DAF Dosimetric adjustment factor
DWEL Drinking water equi\alent level
DWTP Drinking water treatment plant
El I lepliillLioroprop\ I 1.2.2.2-tetrafluorethyl ether
EPA I S I¦ n\ironmental Protection Agency
FABP fatty acid-binding protein
FIFRA federal Insecticide, Fungicide, and Rodenticide Act
GD (iestation day
GLP (iood Laboratory Practices
H3O+ 11 yd milium ion
HERO I leallh & Environmental Research Online
HED 11uman equivalent dose
HFPO liexalluoropropylene oxide
hL-FABP Human liver fatty acid-binding protein
HPLC High-performance liquid chromatography
ICR Institute of Cancer Research
LC50 Median lethal concentration
LD Lactation day
LD50 Median lethal dose
LLNA Local lymph node assay
LOAEL Lowest-ob served-adverse-effect level
vii
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
LOQ Limit of quantification
|ig/L Microgram per liter
mg/L Milligram per liter
mg/kg Milligram per kilogram
mg/kg/day Milligram per kilogram per day
MO A Mode of action
NC DHHS North Carolina Department of Health and Human Services
ng/g nanograms per gram
ng/mL Nanograms per milliliter
NHANES National Health and Nutrition I l\aminaLion Survey
NOAEL No-observed-adverse-eflccl le\el
OECD Organization for Economic Cooperation and Development
OPPT Office of Pollution Pre\ ention and Toxics
PBPK Physiologically based pharmacokinetic
PFAS Per-and polyfluoroalkyl substances
PFOA Perfluorooctanoic acid
PFOS Perfluorooctane sulfonate
PMN Premanufaclurc notice
PND Postnatal day
POD Point of departure
PODhed Point of departure human e<.|ui\alent dose
PPARa Peroxisome piolilcratoi-acli\ ated receptor alpha
RBC Red blood cell
RfD Rclcrcnce dose
RIVM National Institute for Public I lealth and the Environment
(Rijksinstiluul \oor Yolksuexondheid en Milieu)
SDH Sorbitol dehydrogenase
T1/2 I lall-lile
TDAR T cell-dependent antibody response
TG Test guideline
TSCA Toxic Substances Control Act
UF I ncertainty factor(s)
UFa Interspecies uncertainty factor
UFd Database uncertainty factor
UFh Intraspecies uncertainty factor
UFl LOAEL to NOAEL extrapolation uncertainty factor
UFs Extrapolation from subchronic to a chronic exposure duration
WOS Web of Science
viii
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Executive Summary
The U.S. Environmental Protection Agency (EPA) is issuing draft subchronic and chronic oral
toxicity values (i.e., reference doses, or RfDs) for 2,3,3,3-tetrafluoro-2-(l,1,2,2,3,3,3-
heptafluoropropoxy)propanoic acid (Chemical Abstracts Service Registry Number (CASRN)
13252-13-6)—or hexafluoropropylene oxide (HFPO) dimer acid—and 2,3,3,3-tetrafluoro-2-
(l,l,2,2,3,3,3-heptafluoropropoxy)propanoate (CASRN 62037-80-3)—or HFPO dimer acid
ammonium salt for public comment. These chemicals are also known as "GenX chemicals"
because they are the two major chemicals associated with GenX processing aid technology. The
toxicity assessment for GenX chemicals is a scientific and technical report that includes toxicity
values associated with potential noncancer health effects following oral exposure (in this case,
oral reference doses (RfDs). This assessment evaluates human health hazards. The toxicity
assessment and the values contained within is not a risk assessment as it does not include an
exposure assessment nor an overall risk characterization I'ni lher. the toxicity assessment does
not address the legal, political, social, economic, or technical considerations involved in risk
management. When final, the GenX chemicals toxicity assessment can lx- used by EPA, states,
tribes, and local communities, along with specific exposure and other rele\ ant information, to
determine, under the appropriate regulations and statutes, if. and when, it is necessary to take
action to address potential risk associated with human exposures to GenX chemicals.
These GenX chemicals are fluorinaled organic chemicals that are part of a larger group of
chemicals referred to as "per- and polylluoroalkyl subslances "" In 2006, the PA initiated a
stewardship program with the goal of eliminating chemical emissions of perfluorooctanoic acid
(PFOA) and related chemicals by 2015. GenX chemicals are replacements for PFOA.
Specifically, GenX is a trade name for a processing aid technology that enables the creation of
fluoropolymers without the use of PFOA Muoropolymers are used in many applications,
including the manufacture of nonstick coatings lor cookware, water repellent garments, and other
specialty agrochemical and pharmaceutical applications.
For III PO dimer acid and its ammonium sail, oral animal toxicity studies of acute, short-term,
subchronic. and chronic duration are available in rats and mice. Limited information identifying
health effects from inhalation or dermal exposures to GenX chemicals in animals is available.
Repeated-dose toxicity data are available lor oral exposure, but not for the other exposure routes
(inhalation and dermal exposures). Thus, this assessment applies only to the oral route of
exposure. One oral reproductive and developmental toxicity study in mice and one prenatal
developmental toxicity study in rats are available. These studies report liver toxicity (increased
relative liver weight, hepatocellular hypertrophy, and single-cell necrosis), kidney toxicity
(increased relative kidney weight), immune effects (antibody suppression), developmental
effects (increased early deliveries and delays in genital development), and cancer (liver and
pancreatic tumors). Overall, the available toxicity studies demonstrate that the liver is
particularly sensitive to HFPO dimer acid- and HFPO dimer acid ammonium salt-induced
toxicity.
The EPA followed the general guidelines for risk assessment set forth by the National Research
Council (1983) and the EPA's Framework for Human Health Risk Assessment to Inform
Decision Making (2014a) in determining the point of departure (POD) for the derivation of the
RfDs for these chemicals. Consistent with the recommendations presented in the EPA's A
Review of the Reference Dose and Reference Concentration Processes (USEPA 2002), the
vii
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Agency applied uncertainty factors (UF) to address intraspecies variability, interspecies
variability, and extrapolation from a subchronic to a chronic exposure duration.
The critical study chosen for determining the subchronic and chronic RfDs for HFPO dimer acid
and/or its ammonium salt is the oral reproductive/developmental toxicity study in mice with a
no-observed-adverse-effect level of 0.1 milligrams per kilogram per day (mg/kg/day) based on
liver effects (single-cell necrosis in males) (DuPont-18405-1037, 2010). Using the EPA's
Benchmark Dose Technical Guidance Document (2012), benchmark dose modeling was used to
empirically model the dose-response relationship in the range of observed data. Additionally, the
EPA's Recommended Use of Body Weight3/4 as the Default Method in Derivation of the Oral
Reference Dose (201 lb) was used to allometrically scale a toxicologically equivalent dose of
orally administered agents from adult laboratory animals lo adult humans. The use of allometric
scaling addresses some aspects of cross-species extrapolation of toxicokinetic and toxicodynamic
processes (i.e., interspecies uncertainty factor). The resulting I'OI) human equivalent dose is
0.023 mg/kg/day. UF applied include a 10 for intraspecies variability. 3 for interspecies
differences, and 3 for database deficiencies, including immune effects and additional
developmental studies, to yield a subchronic Rll) of <> 0002 mg/kg/day In addition to those
above, aUF of 3 was also applied for extrapolation lYom a subchronic lo a chronic duration in
the derivation of the chronic RfD of 0.00008 mg kg day
viii
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
1.0 Introduction and Background
1.1 History of Assessment of GenX Chemicals
In 2008, DuPont submitted two premanufacture notices (PMNs) to the U.S. Environmental
Protection Agency (EPA) under the Toxic Substances Control Act (TSCA) (Title 15 of the
United States Code § 2601 etseq.) for two chemicals—2,3,3,3-tetrafluoro-2-(l,1,2,2,3,3,3-
heptafluoropropoxy)propanoic acid (Chemical Abstracts Service Registry Number (CASRN)
13252-13-6)—or hexafluoropropylene oxide (HFPO) dimer acid—and 2,3,3,3-tetrafluoro-2-
(l,l,2,2,3,3,3-heptafluoropropoxy)propanoate (CASRN 62037-80-3)—or HFPO dimer acid
ammonium salt—that are part of the GenX processing aid technology they developed. It should
be noted that in July 2015, DuPont announced it had separated its Performance Chemicals
segment through the creation of The Chemours Company The GenX processing technology and
associated chemicals are now products of The Chemours Company (Chemours, 2018). Because
the submitted studies were conducted prior to the 2' > I 5 separation, most of the studies are
referenced with DuPont identifiers.
Upon receipt, the EPA assigned these PMNs case numbers P-08-0508 and P-<>8-0509, and they
were reviewed by the New Chemicals Program in the Office of Pollution Pre\ cnlion and Toxics
(OPPT) and posted in the Federal Re: for puMie comment (I (SEPA, 2008). A PMN
assessment was completed and included a hazard assessment based on EPA review of test data
submitted to the Agency with the PMNs (including two 28-day oral (gavage) toxicity studies in
mice (DuPont-24459, 2008) and rats (DiiPonl-24447. 2008)). as well as publicly available
literature and TSCA confidential business information on other per- and polyfluoroalkyl
substances (PFAS). Submitted test data on HFPO dimer acid and or its ammonium salt were
available for numerous endpoints such as acute toxicity, metabolism and toxicokinetics,
genotoxicity, and systemic toxicity in mice and rats with dosing durations of up to 28 days.
The EPA OPPT evaluated the methods and data submitted and deemed the studies acceptable to
the Agency The studies submitted in 2<)()S with the PMNs formed the primary basis of the
EPAs hazard assessment at that time. The 28-day toxicity study in mice, from which the EPA
OPPT deri\ ed the point of departure (POD) of 0.1 milligrams per kilogram per day (mg/kg/day),
was conducted according to Organization for Economic Cooperation and Development (OECD)
Test Guideline (TG) 407 and followed Good Laboratory Practices (GLP) (DuPont-24459, 2008;
OECD, 2008). The submitted studies were also used, in concert with information on other PFAS
chemicals, to inform the decision for further testing included in the Consent Order that
concluded the PMN re\ iew (I SI-PA, 2009).
The Consent Order included, among other things, additional testing pertaining to human health.
The tests were identified in the Consent Order according to OECD TG numbers and/or EPA
health effects TGs for pesticides and toxic substances numbers. The studies included in the
Consent Order relevant to human health and this assessment are listed below:
• Repeated dose metabolism and pharmacokinetics studies (OPPTS 870.7485) in mice and
rats (Dupont-18405-1017, 2011)
• Modified Oral (Gavage) Reproduction/Developmental Toxicity Study in Mice (OECD
TG 421) (Dupont-18405-1037, 2010)
1
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
• 90-Day Oral (Gavage) Toxicity Study (OECD TG 408) (species not specified): both mice
(DuPont-18405-1307, 2010) and rats (Dupont-17751-1026, 2009) were submitted
• Combined Chronic Toxicity/ Oncogenicity Study in Rats (OECD TG 453) (Dupont-
18405-1238, 2013)
It is noted that the OECD TGs are accepted internationally as standard methods for safety testing
and:
.. .are covered by the Mutual Acceptance of Data, implying that data generated in the
testing of chemicals in an OECD member country, or a partner country having adhered to
the Decision, in accordance with OECD Test (iuidelines and Principles of GLP, be
accepted in other OECD countries and partner countries ha\ ing adhered to the Decision,
for the purposes of assessment and other uses relating to the protection of human health
and the environment (OECD, 2018).
Specifically, for the required oral reproductive/de\ elopmental toxicity test, the EPA OPPT
included requirements for specific modifications to the test to increase robustness of the study for
this class of chemicals (DuPont-184<)5-l<)37. 2010; Ol-('l). 2<)|6). These modifications are
stated in the Consent Order (USEPA. 20(»l->) and were followed by the testing laboratory as
outlined in the study report (DuPont-184<)5-|()37. 2010) For the required combined chronic
toxicity/oncogenicity study, the EPA reviewed and concurred with protocols submitted to the
Agency prior to the study being conducted (DuPont-IX4<)5-l238, 2013). In addition, the
submitter consulted with the l-PA on study findings to determine the need for additional data
(e.g., further toxicokinetic testing bused on results of the first tierOPPTS 870.7485 study).
Finally, while not specifically required under the Consent Order, additional OECD TG studies
were conducted and submitted lor Agency review (e.g., the prenatal and developmental toxicity
study in rats (OIX'I) TG 414) (l)uPont-lX4<)5-K4l, 2010).
1.2 Uses of GenX Chemicals under TSCA
GenX is a trade name for a processing aid technology developed by DuPont to make high-
performance fluoropolymers without the use of perfluorooctanoic acid (PFOA) (Chemours,
2018). Transition to GenX processing aid technology began in 2009 as part of the company's
commitment under the 2< > I <) 2d I 5 PFOA Stewardship Program to work toward the elimination of
these chemicals from emissions and products by 2015. Although production of most long-chain
PFAS (i.e., six or more carbons)1) has been phased out in the United States and has been
generally replaced by production of shorter chain PFAS, the EPA is aware of ongoing uses by
companies that did not participate in the PFOA Stewardship Program and ongoing uses of long-
chain PFAS that are available in existing stocks or are being newly introduced via imports.
Fluoropolymers are used in many applications because of their unique physical properties such
as resistance to high and low temperatures, resistance to chemical and environmental
degradation, and nonstick characteristics. Fluoropolymers also have dielectric and fire-resistant
properties that have a wide range of electrical and electronic applications. Applications of
1 Long-chain PFASs comprise two subcategories: (1) long-chain perfluoroalkyl carboxylic acids (PFCAs) with eight
or more carbons, including PFOA, and (2) perfluoroalkane sulfonates (PFSAs) with six or more carbons, including
perfluorohexane sulfonic acid (PFHxS) and periluorooctane sulfonic acid (PFOS). https://www.epa.gov/assessing-
and-managing-chemieals-under-tsca/risk-management-and-pol.vfluoroalkvl-substances-pfass#tab-3.
2
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
fluoropolymers include architecture, fabrics, automotive uses, cabling materials, food processing,
pharmaceutical and biotech manufacturing, and semiconductor manufacturing (DuPont, 2013).
One of the PMNs the EPA received in 2008, P-08-0508, was for HFPO dimer acid, a chemical
used as an intermediate to make the polymerization aid HFPO dimer acid ammonium salt. The
PMN for HFPO dimer acid ammonium salt was received by the EPA under PMN P-08-0509, and
it is used as a replacement for PFOA in the manufacture of fluoropolymers. The GenX resin
manufacturing process includes the thermal transformation of the HFPO dimer acid ammonium
salt processing aid into a hydrophobic hydride. HFPO is used in the manufacture of the HFPO
dimer acid, HFPO dimer acid ammonium salt, other HFPO dimer acid derivatives,
fluoropolymers (including polyethers), and other specially agrocheinical and pharmaceutical
applications. When in water, both HFPO dimer acid and I II PO dimer acid ammonium salt
dissociate to form the HFPO dimer acid anion (HFPO") as a common analyte. HFPO is
manufactured from hexafluoropropene. HFPO dimer acid can react with additional HFPO to
form the HFPO trimer acid and longer polymer fluorides. There are oilier PFAS chemicals that
might be part of the GenX processing aid technology, hut HFPO dimer acid and its ammonium
salt are the major chemicals associated with this technology.
1.3 Occurrence
GenX chemicals were identified in North Carolina's Cape I 'ear River and its tributaries in the
summer of 2012 (Strynar et al., 2015) following this disco\ cry. between June and December
2013, Sun et al. (2016) sampled source water at three drinking water treatment plants (DWTPs)
(identified as DWTPs A. li. and C) treating surface water from the Cape Fear River watershed.
The mean concentration of I II PO dimer acid in the finished drinking water treated by DWTP C
was 0.631 micrograms per liter (uu L) (Sun et al . 2<)|r>) In a separate experiment to look at
removal efficiency of DWTP C. water samples were taken during August 2014 from the raw
water intake and after each treatment process step used by DWTP C (i.e., coagulation/
flocculation/sedimenlation. raw and settled water ozonation, biological activated carbon
filtration, and disinfection by medium-pressure ultraviolet lamps and free chlorine). GenX
chemicals were found at concentrations of 0 4 n 5 |ig/L at all steps of the treatment process,
indicating that the concentrations of HFPO dimer acid were only slightly decreased by the
conventional and advanced water treatment processes used at this DWTP.
The publication of these data prompted the North Carolina Department of Environmental Quality
to sample sites for GenX chemicals along the Cape Fear River and in private wells close to the
Chemours facility. In certain samples of surface water, ground water, and finished drinking
water, GenX chemicals were delected above 0.140 |ig/L, which is North Carolina's drinking
water health goal for GenX chemicals. Chemours has indicated that GenX chemicals have been
discharged into the Cape Fear River for several decades as a byproduct of other manufacturing
processes (NCDEQ, 2017).
GenX chemicals have been identified in other media, including rainwater and air emissions.
Estimates from the Chemours Fayetteville Works plant (in the North Carolina Cape Fear
watershed) indicate that Chemours' annual emissions of GenX chemicals could exceed 2,700
pounds per year (NCDEQ, 2018a). Additional details on air emissions of GenX chemicals at the
Fayetteville Works plant can be found here. Rainwater samples were collected between February
28 and March 2, 2018 up to 7 miles from the North Carolina plant (NCDEQ, 2018b). The highest
3
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
concentration of GenX chemicals in a rainwater sample (0.810 |ig/L) was detected 5 miles from
the Fayetteville Works facility center. The three samples collected 7 miles from the plant ranged
from 0.045 to 0.060 |ig/L (NCDEQ, 2018b).
GenX chemicals also have been detected in three on-site production wells and one on-site
drinking water well at the Chemours Washington Works facility in Parkersburg, West Virginia.
The EPA subsequently requested that Chemours test for GenX chemicals in both raw and
finished water at four public drinking water systems and 10 private drinking water wells.
Chemours agreed to the testing and completed sampling during February 2018. The results from
these samples can be found here and range before treatment from less than 0.010-0.081 |ig/L in
the public drinking water systems and less than 0.010 o <>52 uu I. in the private drinking water
wells. All samples were less than 0.010 |ig/L after treat men l (I SI-PA, 2018a).
Finally, low concentrations of HFPO dimer acid (O.oo.i n 1104 uu |.) were detected in the
Delaware River, as reported in the recent publication In Pan et al (2d 18)
Globally, GenX chemical occurrence has been reported in Germany (I leydebreck et al., 2015;
Pan et al., 2018), China (Heydebreck et al., 2<)| 5. Pan et al. 2<~>17, 201S. Song et al., 2018), the
Netherlands (Heydebreck et al., 2015; Gebbink et al . 2<)|7. Pan et al., 201S). the United
Kingdom (Pan et al., 2018), South Korea (Pan et al.. 2d IK). and Sweden (Pan et al., 2018).
1.4 Other Assessments of <
1.4.1 North Carolina Assessment
The North Carolina Department of Health and Human Ser\ ices (\C DHHS) released a health
assessment and provisional drinking water health goal lor (ienX chemicals in July 2017. North
Carolina defines "health goal" as a nonrcgulaloiv. non-enforceable level of contamination below
which no adverse health effects would he expected over a lifetime of exposure. The provisional
health goal for exposure to GenX chemicals in drinking water is 0.140 |ig/L, which is intended to
protect the most sensiti\ e population, namely bottle-led infants. The state selected bottle-fed
infants as the most sensi ti \ e population because they drink the largest volume of water per body
weight (IJW )
North Carolina's provisional health goal is based on a reference dose (RiD) derived from a no-
observed-ad verse-effect level (NO VEL) of 0.1 mg/kg/day for liver effects (single-cell necrosis)
in mice (DuPonl-24451), 2008. DuPont-18405-1037, 2010). The total uncertainty factor (UF)
applied was 1,000, including indi\ idual factors to account for interspecies variability (10),
intraspecies variability (MM. and extrapolation from a subchronic to a chronic exposure duration
(10). This RiD of 0.0001 mg kg/day was used to derive a drinking water equivalent level
(DWEL), which considers exposure. The DWEL was calculated using BW and drinking water
intake for bottle-fed infants and a relative source contribution of 20% to account for potential
exposure to GenX chemicals from other media and routes, including air, soil, dust, and food.
Additional details are available at NC DHHS.
1.4.2 Report by the National Institute for Public Health and the Environment
The National Institute for Public Health and the Environment (RIVM) in the Netherlands
evaluated the data for GenX chemicals to set a safe limit for air. RIVM's assessment focused on
the precursor 2,3,3,3-tetrafluoro-2-(heptafluoropropoxy)-propanoic acid (FRD-903 (a synonym
4
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
for HFPO dimer acid)), the processing agent ammonium 2,3,3,3-tetrafluoro-2-
(heptafluoropropoxy) propanoate (FRD-902 (a synonym for HFPO dimer acid ammonium salt)),
and the transformation product heptafluoropropyl 1,2,2,2-tetrafluorethyl ether (El). Overall,
RIVM concluded that there is no health risk expected for people living near plants from
emissions of FRD-902 and FRD-903 at a limit of 73 nanograms per cubic meter (insufficient
data are available to determine the toxicity of El) (Beekman et al., 2016). RIVM used the oral
carcinogenicity study in rats as the critical study (DuPont-18405-1238, 2013) and concluded that
the study NOAEL was 0.1 mg/kg/day, based on increased albumin and the albumin-to-globulin
(A/G) ratio observed at 12 months in males dosed with 1 mg/kg/day, an effect that indicates the
potential for immunotoxicity. Using route-to-route extrapolation. RIVM converted this NOAEL
to an air concentration to be used as the POD. UF to account lor intraspecies differences (10) and
interspecies differences (1.8), and an additional factor to account for uncertainty in the human
elimination of GenX chemicals (66) were applied to the POD to determine the chronic inhalation
limit. Additional details are available in the 1
2.0 Nature of the Stressor
2.1 Chemical/Physical Properties
HFPO dimer acid and its ammonium salt are fluorinaled organic compounds (I 'iuure 1).
F F F
F F F F
F
I
OH
¦FO
F F F
F F F F
F0
H
H—N—H
H
I II PO dimer acid I ll'I'O dimer acid ammonium salt
CASRN 13252-13-0 CASRN 62037-80-3
Figure I. Structure ol' IIM'O Dimer Acid and HFPO Dimer Acid Ammonium Salt
HFPO dimer acid is a liquid u hcrcas its ammonium salt is a solid at room temperature. Both are
highly soluble in water Iacc|M in very acidic solvents (pH < 3), the acid will dissolve and be
present as the acid anion with a - I charge. The associated cation ion will be a hydronium ion
(H30+) in water if other hydrogen ion acceptors are absent. Both compounds can volatilize from
water to air, where they will dissolve in aerosolized water droplets or bind to suspended
particulate matter. In soils, they will migrate with the aqueous phase or bind to the soil particle
surfaces with areas of positive charge. The organic portion of the HFPO dimer acid and its
ammonium salt are stable to environmental degradation. Table 1 presents the chemical and
physical properties of HFPO dimer acid and its ammonium salt.
5
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Table 1. Chemical and Physical Properties of HFPO Dimer Acid and HFPO Dimer Acid
Ammonium Salt
Property
HFPO Dimer Acid
HFPO Dimer Acid
Ammonium Salt
Source
Chemical Abstracts
13252-13-6
62037-80-3
Chemical Abstracts Service.
Service Registry
Number (CASRN)
CAS Index Name
Propanoic acid, 2,3,3,3-
tetrafluoro-2-(l,l,2,2,3,3,3-
heptafluoropropoxy)
Propanoic acid, 2,3,3,3-
tetrafluoro-2-( 1,1,2,2,3,3,3 -
hcptafluoropropoxy)-
ammonium sail (1 h
Chemical Abstracts Service.
IUPAC Name
Synonyms
GenX Acid
FRD903
H-28307
C3 Dimer acid
2,3,3,3 -tetrafluoro-2-
(heptafluoropropoxy)-
propanoic acid
GenX salt3<>S
FRD 902
FDR 'Jo:! )X
tt-:i:i(.
ii-:_5:^
ii-:s()-:
n-:s^r
H-2S'OS
H-2854S
IIFPO dimer ammonium salt
C3 dimer sail
\miikiiinim. 2. \ \ '-
k'liaHiKii'i)-2-
(licpi;illiK>i\>pi'i>|x>xy)-
piopaiKiale
DuPont.
Chemical Formula
( III' ()
( II l; \()
Molecular Weight
11() ()(¦ u inol
'4~ us u iikS ( i as S(i"„ salt solution in
walen
No measuiement available for
salt form
DuPont-24637, 2008;
DuPont-24698, 2008.
Melting Phi ill
4UO (
-21.0 °C (as 86% salt solution
in water)
No measurement available for
salt form
DuPont-24637, 2008;
DuPont-24698, 2008.
Vapor Pressure
'11(1 |';| (2 " nun 1 Ig)
(20 Ci
No measurement available
DuPont-24128, 2008;
DuPont-24129, 2008.
Henry's Law
<2.5 \ In aim-m3/mol
No measurement available
Calculated from measured
Constant
vapor pressure and highest
measured solubility. Water
solubility is reported to be
"infinite" (DuPont-24128,
2008; DuPont-24129,
2008), so the actual Kh is
expected to be much lower.
These values should not be
used to estimate partitioning
between water and air.
Pka
2.84 (20 °C)
3.82
DuPont-26349, 2008.
6
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Property
HFPO Dimer Acid
HFPO Dimer Acid
Ammonium Salt
Source
Pkb
8.1
8.1
DuPont-24198, 2008
(HFPO dimer acid
ammonium salt).
Koc
Soil -12 L/Kg (log 1.10)
Sludge -12.6 L/Kg (log = 1.08)
DuPont-17568-1675, 2008.
K-OW
Not applicable3
Not applicable3
Solubility in Water
@20 °C
>751 g/L
>739 g/L
Highest tested values.
Actual solubility not
determined but described as
"infinite" (DuPont-24128,
2008; DuPont-24129,
2008).
Half-life (Ti/2) in
Stable
Stable
Measured hydrolysis values
Water
(25 °C)
for salt. No degradation in 5
days at 50 °C, pH 4, 7, and
(DuPont-24199, 2008).
Half-life (Ti/2) in
Stable
Siahle
Ultraviolet-Visible and
Air
Visible Spectrophotometry
spectra for acid showed
little absorption above 240
nm (DuPont-26349, 2008).
The EPA concurs with
DuPont's assessment that
the salt is assumed to be
similar. Measured OH-
reaction rate for El reaction
product indicates TVi > 37
years.
Biodegradntion
liindemadalKiii was nni
likidemadalinii was not
DuPont-A080558, 2009;
nhscr\cd mi reads
nhsei'\ ed mi ready
DuPont-1388231-
hkidcuradalimi and inherent
hkideuradalkHi and inherent
R2009NC03 l(a)-02, 2010;
hkidcuradalinii lesls
hkidemadalkHi lests
DuPont-1388231-
R2009NC03 l(s)-02, 2010.
Biocoiiceiuralkiii
'"(Inn li
< 30 (log < 1)
Measured BCFb < 30 at
(Fish
bioconcei lira lion
0.02 mg/L and < 3 at
0.2 mg/L in Medaka 28
factor (B( In
days (DuPont-A080560,
2009).
Bioaccuniiilalkiii
in
< 10
Pan etal., 2017.°
(Field
bioaccuniiilalinn
factor (BAF))
a Surfactants are surface acting agents that lower the interfacial tension between two liquids. Their amphiphilic nature (i.e., they
contain both a hydrophilic part and a hydrophobic part) causes them to accumulate at interfaces such as the water-air interface,
the water-food interface, and glass walls, which hampers the determination of their aqueous concentration. These surfactant
properties present difficulties in applying existing methods for the experimental determination of log K0w and produce
unreliable results.
b The concentration of the propionate ion was not quantified in the bioconcentration factor (BCF) study so the values are limits
based on the limit of quantification (LOQ) for the analytical technique employed in the study.
c Pan et al. (2017) quantified the propionate ion and found that the concentrations were low in the tissues expected to most likely
accumulate perfluorinated compounds (e.g., muscle, blood, and so forth). The tissue values indicate a bioaccumulation factor
(BAF) less than 10. Lipid tissue concentrations are not the basis for this BAF as is common for "traditional" organic
compounds.
7
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
2.2 Environmental Fate
The HFPO dimer acid and its ammonium salt are stable to photolysis, hydrolysis, and
biodegradation. The degradation data suggest that the substances will be persistent (i.e., half-life
(T1/2) longer than 6 months) in air, water, soil, and sediments. Based on measured physical-
chemical and sorption data, they are expected to run off into surface water and to rapidly leach to
ground water from soil and landfills. As seen with PFOA and chemicals with similar properties,
HFPO dimer acid and its ammonium salt might undergo long-range atmospheric transport in the
vapor phase and associate with particulate matter. They are not expected to be removed during
conventional wastewater treatment or conventional drinking water treatment. They have low
potential to bioaccumulate in fish (Table 1).
When released to the fresh water environment, the HFPO dimer acid will dissociate to the HFPO
carboxylate anion and H30+. The ammonium salt will dissoK e lo the HFPO carboxylate anion
and the ammonium cation (NH4+). Both have high solubilities in water and are expected to
remain in water with low sorption to sediment or soil. Given the vapor pressure, the acid can
partition to air as well as to water. The salt can also be transported in air. although the
mechanism of vapor phase transport is not understood (DuPont CCAS. 2<)i)1)) Tn the vapor
phase, the acid and salt are expected to be stable to direct photolysis and will undergo hydroxyl
radical catalyzed indirect photolysis \ cry slowly.
2.2.1 Water
Measured data for the HFPO dimer acid and or ammonium salt show that they are highly soluble
in water (Table 1). The measured base dissociation constants (pKb) indicate that the chemicals
will exist primarily as the propionate ion at most en\ ironmental pH levels.
The chemicals are stable to hydrolysis. A hydrolysis study on the ammonium salt found no
degradation at pH 4, 7, and 0 at 5<) degrees Celsius (°C) in 5 days, indicating a hydrolysis T1/2 of
greaterthan 1 year at 2<) °C (l)ul,ont-24100. 200K) Calculated Henry's Law constants (Table 1)
suggest that partitioning from water to air might occur. Experimental data on the transfer of the
acid and salt from water to air indicates that partitioning from surface water to the vapor phase
might occur, and some transfer from surface water to air is expected (DuPont CCAS, 2009).
Water-air transport of these chemicals, however, is not well understood. Their surfactant
properties, equilibrium between chemical forms as a function of pH, and interaction with
dissolved cations make it difficult to accurately predict how the chemicals will behave in the
aquatic environment
2.2.2 Air
The acid was described as having "a significant vapor pressure" (DuPont CCAS, 2009). As
observed with PFOA and other perfluorochemicals, these chemicals could be transported in the
vapor phase or could associate with particulate material and be transported with the solids when
released or partitioned into air.
When released to air or volatilized from water, the chemicals are stable and long-range transport
could occur. Removal from air is expected to occur through scavenging by water droplets and
attachment to particulates followed by precipitation and settling. Studies regarding long-range
transport or air removal rates are not available.
8
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
2.2.3 Sediments and Soils
Organic carbon normalized sorption coefficients were measured by high-performance liquid
chromatography (HPLC) (following OECD TG 121). The sorption of the HFPO dimer acid
ammonium salt to soil and sludge were 12.0 L/kg (log = 1.10) and 12.6 L/kg (log = 1.08),
respectively (DuPont-17568-1675, 2008; OECD, 2001a). Their high water solubility and low
sorption potential indicate that the chemicals will tend to remain largely in water with little
partitioning to soil or sediment. If applied or deposited to soil, they will run off or leach to
ground water and, as indicated by the vapor pressure, could \ olatilize to air.
2.2.4 Biodegradation
The GenX chemicals are resistant to biodegradation; no degradation was observed in
standardized internationally recognized test methods for biodegradability The aerobic aquatic
biodegradation T1/2 is on the order of years based 011 no measured inherent biodegradation of the
acid or ammonium salt in OECD 302C, modified Ministry of International Trade and Industry
studies (DuPont-1388231-R2009NC031 (a)-02, l)uPont-1388231-R2<)
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
In a 4-day trout hepatocyte bioaccumulation screening test (non-TG) with the HFPO dimer acid
ammonium salt, no metabolism was observed, suggesting that in vivo metabolism does not
significantly affect potential bioaccumulation (DuPont-23459, 2007).
2.3 Toxicokinetics
In rats and mice, HFPO dimer acid and its ammonium salt are both absorbed from the
gastrointestinal tract at levels that are proportional to dose following acute oral exposures.
Transfer from plasma/serum to the liver, but not adipose tissue, was demonstrated in the few
studies that conducted tissue analysis. The potential for maternal transport to the fetus during
development and to the neonate during lactation was noted in one study (DuPont-18405-1037,
2010). Urine is the primary pathway for excretion. Based on data from studies of acute, single-
dose, gavage, or intravenous exposures, T1/2S in the beta (elimination) phase are longer in male
rats and mice than in females. The male rats T1/2S in the beta (elimination) phase are relatively
comparable to the male and female monkeys, whereas the female rats" T1/2S are shorter.
The HFPO dimer acid is a strong acid (pKa = 2 84) and will be predominantly ionize in aqueous
solutions with pHs higher than 4 and in both plasma and serum (DuPonl-20349, 2008). Once in
solution, the cation that counterbalances the HFPO dimer anion will vary with the salt used or
the mineral ion composition of the solvent, plasma, serum, intercellular, and intracellular fluids.
Based on the physical and chemical properties of I II PO dimer acid and its ammonium salt, once
these chemicals enter physiologic compartments with pi Is higher than 4 (e.g., water, serum, or
blood), they will either dissociate (acid) or dissoK e (ammonium salt) to yield the carboxylate
anion. Thus, what is being measured in the studies outlined below is the HFPO dimer acid anion
concentration regardless of w hether animals are dosed with the HFPO dimer acid or its
ammonium salt.
2.3.1 Absorption
Oral. Sprague Daw ley rats (li\ e of each sex (5/se\)) were administered (via gavage) a single oral
dose of 3d milligram per kilogram (mg kg) I II PO dimer acid ammonium salt in aqueous solution
(purity 84"<>) in a study conducted according to I-IPA TG OPPTS 870.7485. Two animals of each
sex sen ed as controls. Urine was collected and pooled for the first 12 hours and again for 12 to
168 hours after dosing (blood/serum was not sampled). The 12-hour urine collections accounted
for a mean of <> to 97% of the dose, supporting a conclusion that these Gen-X chemicals are
rapidly absorbed from the G1 tract by male and female rats (DuPont-18405-1017 RV1, 2011).
In a similar guideline study with Crl/CD-1(ICR) mice (5/sex) (OPPTS 870.7485), the animals
were administered a single oral dose of 3 mg/kg HFPO dimer acid ammonium salt (purity 84%)
by gavage in aqueous solution (DuPont-18647-1017 RV1, 2011). Two animals of each sex
served as controls. In the 12-hour pooled urine, 31% (mean) of the substance was found for the
males and 39% (mean) for the females. By 168 hours postdosing, the mean urine values
accounted for 90% and 92% of the total dose for male and female mice, respectively, indicating
that both rats and mice extensively absorb HFPO dimer acid ammonium salt. This study
additionally shows mice either absorb HFPO dimer acid ammonium salt slower or eliminate it in
urine slower than rats. In mice, the HFPO dimer in urine was found as both the protonated HFPO
dimer acid and its sodium salt. The authors report the sodium salt as being a product of the
analytical method. Thus, the authors considered the recovered compound to be the dosed ionized
HFPO acid (DuPont-18647-1017 RV1, 2011).
10
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
A 28-day gavage study by Rushing et al. (2017) indicates a potentially more complex
toxicokinetic profile for HFPO dimer acid when dosing occurs over multiple days. Groups of six
male and six female C57BL/6 mice were given doses of 1, 10, or 100 mg/kg/day HFPO dimer
acid daily for 28 days. Serum concentrations were measured at intervals of 1, 5, 14, and 28 days,
and urine concentrations were measured on days 1, 2, 3, 5, 10, and 14. At each time point, serum
levels reflected the magnitude of the dose, but not the exposure duration. The peak concentration
occurred at day 5 for all but the high-dose males, where it occurred at day 14. Serum
measurements for the 1- and 10-mg/kg/day doses were lower on days 14 and 28 than on day 5.
The differences in serum concentration between days 5, 14, and 28 are not explained by the
study authors, but could possibly indicate changes in absorption, tissue storage, or elimination
after repeated dosing. The males exposed to 10 and 100 mg/kg/day had higher serum
concentrations and urine concentrations than the females, as described in section 2.3.5
(Excretion). Based on the higher serum and urine concentrations. there appeared to be greater
absorption in males than in females.
In a repeated-dose study following OECD T(i 4<)S. I LFPO dimer acid ammonium salt (purity
84%) was administered to Crl:CDl(ICR) mice lor (males) or 9b (females) consecutive days
via gavage at doses of 0, 0.1, 0.5, and 5 mg/kg/day (DiiPonl-l S405-1307, 2<)|0) Ten animals per
sex per group (animals/sex/group) were included for toxicity e\ aluation, and an additional 15
animals/sex/group were included for quantitation of the test substance concentration 2 hours after
dosing on day 0 (the first day of dosing) (5/sex/dose). pix>\ idinu a measure of postdosing
absorption (Table 2). Overall, concentrations increased in a dose-related manner, with broad
standard deviations indicative of considerable interanimal \ ariability in the absorption. The doses
evaluated differ from those used by Rushing et al (2<)|7). limiting comparisons of the
postexposure data. The sex difference seen In Rushing et al (2< 117) (i.e., where male uptake to
serum for the 1 and l<> mg kg day doses at the end of day I was greater than female uptake) is
not as apparent at 2-honr postdosing in this dataset.
Table 2. Plasma ('oncenlralion in (rl:('l)l(l('U) Mice at 2 Hours after the First Gavage
Exposure lo UI-'PO Dimer Acid Ammonium Salt"
Dose
mg/kg/(l:i>
Miilos
Females
ng/ml.
SD
ng/mL
SD
0
0
0
0
0
0.1
~ 1(1
99
824
72
0.5
\XU(.
1,175
3,608
1,308
5
42.5X0
5,214
35,340
9,362
Notes'. ng/mL = nanograms per milliliter; SD = standard deviation.
a Adapted from Dupont-18405-1307 (2010)
Inhalation. There are no studies investigating HFPO dimer acid or its ammonium salt's uptake
following inhalation exposures of aerosols. In a study conducted by Dupont (17751-723, 2009),
groups of three young male and female Crl:CD(SD) rats were exposed to aerosols containing 0,
13, and 100 milligrams per cubic meter (mg/m3) of HFPO dimer acid ammonium salt (84%
purity) for 4 hours per day for 2 days. There were no measurements of the chemical in serum or
plasma, however, to support an estimate of absorption by way of the respiratory tract.
11
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Dermal. Absorption of HFPO dimer acid ammonium salt through the skin was determined in
vitro with rat and human skin specimens (DuPont-25292, 2008).
Penetration rates were 70.3 ± 5.3 and 6.2 ± 5.3 micrograms per square centimeter per hour,
respectively, and these have dermal permeability coefficients (Kp) of 5.71E-4 ± 4.3E-5 and
5.02E-5 ± 4.3E-5 centimeters per hour for rats and humans. These dermal kinetic parameters
demonstrate dermal absorption occurs, but at a relatively slower rate than chemicals that are well
absorbed dermally.
2.3.2 Distribution
Crl:CD(SD) rats (3/sex/dose) were administered a single oral dose of 10 or 30 mg/kg by gavage
in aqueous solution of either HFPO dimer acid ammonium sail (purity 84%) or HFPO dimer acid
(purity 98%) (DuPont-24281, 2008; DuPont-24286, 2<)i)K) Plasma samples were collected at 0,
0.25, 0.5, 1, 2, 4, 8, 12, 24, 48, 72, 96, 120, 144, and I^S hours, as described in section 2.3.6
(Pharmacokinetic Clearance and Half-life Data) I ,i\ or and lai samples were collected for
analysis after sacrifice. The mean liver tissue-to-plasma concentration ratio was higher in males
for the ammonium salt (2.19) than for the acid (<> M) at the low dose (I <> mg/kg); however, this
ratio equilibrated at the high dose (30 mg/kg) for the ammonium salt (0.78) and the acid (0.71).
At both doses, females had a lower accumulation of I II PO dimer acid and its ammonium salt in
the liver. In females dosed with HI PO dimer acid ammonium salt at the low dose (10 mg/kg),
liver HFPO dimer acid anion concentrations above the limit of quantification (LOQ) (20
nanograms per gram (ng/g)) were ohscr\ cd in two of three animals (20.6 and 54.1 ng/g), while
none were observed at the high dose (30 mu kg). Females dosed with HFPO dimer acid did not
have liver anion concentrations above the I.OQ (2<) nu u) No I II PO dimer acid anion was
detected in the fat tissue samples of any of the rats ui\ en I II PO dimer acid or HFPO dimer acid
ammonium salt (LOO 2<) nu u) (l)uPont-242S 1. 2<)<)S; DuPont-24286, 2008).
Crl:CDl(TCR) mice (3, sex dose) were administered a single oral dose of 10 or 30 mg/kg by
gavage in aqueous solution of I II PO dimer acid ammonium salt (purity 86%) (DuPont-25300,
2008) I illikc the rat studies, the IIFPO dimer acid was not evaluated in the mice. Plasma
samples were collected at 0, 0.25, 0.5, 1, 2, 4, 8, 12, 24, 48, 72, 96, 120, 144, and 168 hours, as
described in section 2.3.6 (Pharmacokinetic Clearance and Half-life Data). Liver and fat samples
as well as plasma were collected for analysis after sacrifice. In males, the mean concentration of
HFPO dimer acid anion in the liver was 384 ng/g (ranging from 90.7 to 929 ng/g) for the low
dose (10 mg/kg) and 457 ng/g (ranging from 87.9 to 750 ng/g) for the high dose (30 mg/kg). The
mean concentration in fat tissue was 31.6 ng/g in males for the high dose (30 mg/kg) and less
than the LOQ (20 ng/g) for the low dose (10 mg/kg) and for both doses in females. In males the
mean concentration in plasma was 756 nanograms per milliliter (ng/mL) (ranging from 139 to
1330 ng/mL) for the high dose (30 mg/kg) and 830 ng/g (ranging from 174 to 1850 ng/mL) for
the low dose (10 mg/kg). In females only one of three mice in each dose group had a plasma
concentration above the LOQ, which was 23.2 ng/mL for the high dose (30 mg/kg) and 29.2 ng/g
for the low dose (10 mg/kg). Based on the plasma and liver concentrations reported in the study,
a liver-to-plasma ratio was calculated for males, but not for females because the females did not
have liver concentrations above the LOQ. At the low dose (10 mg/kg), the liver-to-plasma ratio
was 0.50 to 0.53, and at the high dose (30 mg/kg), it was 0.52 to 0.63.
12
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
To the extent that the perfluorinated ether portion of the HFPO dimer acid is comparable to the
perfluorinated alkane acids (e.g., PFOA), HFPO dimer acid and its ammonium salt are
anticipated to be transported in serum either freely dissolved or bound to serum protein (Gomis
et al., 2018). No studies investigating albumin binding to HFPO dimer acid or its ammonium salt
were identified. Although experimental data demonstrating that the HFPO dimer acid anion
interacts with albumin are lacking, albumin is the serum protein likely to provide the highest
number of primary binding sites for perfluorinated compounds because it accounts for about
60% of total serum protein. Additionally, albumin contains about 15% positively charged amino
acids (Spahr and Edsall, 1964), which are likely to be at the hydrophilic surface in contact with
serum, where they would electrochemically attract the JTFPO dimer acid anion. Some support for
the albumin-binding hypothesis comes from the increased alhuniin and A/G ratio observed in the
subchronic and chronic mice and rat studies (DuPont-2445^. 2<)i)X. DuPont-24447, 2008;
DuPont-17751-1026, 2009; DuPont-18405-1238, 2013)
A study by Sheng et al. (2018) recently identified lluit the I II PO dimer acid anion binds to fatty
acid-binding protein (FABP). FABPs are intracellular lipid carrier proteins that reversibly bind
long-chain fatty acids, phospholipids, and a \ ariely of chemicals that induce peroxisome
proliferation (Erol et al., 2003). They constitute 2".. 5°.. of the cytosolic protein in the liver.
FABPs can be synthesized in the gastrointestinal tract and act as a systemic carrier of long-chain
fatty acids in plasma and serum (Storch and McDermott. 2<)i)1)) Thus, FABPs might play a role
in the systemic distribution of HFPO dimer acid in both its neutral and ionized forms.
2.3.3 Distribution during Gestation and
HFPO dimer acid ammonium salt can be transferred (distributed) li om a pregnant animal to the
fetus during gestation, as demonstrated in an OFCI) TG 421 reproduction/developmental toxicity
study (DuPont-18405-1037. 2<)|<)) Pregnant Crl CI) l(ICR) mice (25/sex/group) were
administered, by gavage. <>. 0.1. n 5. or 5 mg kg day HFPO dimer acid ammonium salt from
premating day 14 to lactation day (I.I)) 2d Blood was collected from the dams 2 hours after
dosing on 1.1)21 (scheduled termination) and pooled. Trunk blood was collected from the culled
pups on postnatal day (PND) 4 and pooled. III P( > dimer acid anion was present in the pooled
plasma of PND 4 pups at concentrations approximately two to four times lower than the
concentrations in the dams on LD 21. These results indicate that there is transfer of HFPO dimer
acid anion from maternal serum. It cannot be determined from this study, however, whether
transfer occurs during gestation, during lactation, or both (DuPont-18405-1037, 2010).
Dosing of five dams per dose group continued following pup delivery (PND 1) until LD 20 to
enable quantification of uptake from maternal milk by the pups in the DuPont-18405-1037
(2010) study. Blood samples were collected from the dams and pups on LD 21. The plasma
levels in the pups were 10 to 32 times lower than the concentrations in pups on PND 4 and were
40 to 60 times lower than those measured in the dams on LD 21. The declining plasma
concentrations from PND 4 to LD 21 suggest that there is little-to-no lactational transfer via
maternal milk. The F1 pups were then dosed daily from PND 21 to PND 40 by gavage with 0,
0.1, 0.5, or 5 mg/kg/day HFPO dimer acid ammonium salt. The plasma levels in the pups
following the 20 days of direct dosing were comparable to those of the dams. Overall, pup
plasma serum concentrations increased in a dose-related manner from PND 21 to PND 40, after
which time they were comparable to those of the dams. Sex differences in HFPO dimer acid
anion concentrations were not observed in the offspring (DuPont-18405-1037, 2010).
13
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Transfer of HFPO dimer acid anion to the fetus was also demonstrated in groups of five
Crl:CD(SD) rats exposed to doses of 0, 5, 10, 100, or 1,000 mg/kg/day from GD 6 to GD 20
(Dupont-18405-849 RV1, 2011). On GD 20, blood was collected from individual dams 2 hours
after dosing and trunk blood was collected from the fetuses and pooled for analysis. The plasma
concentration in the dam blood samples was three times higher than plasma concentration in the
pooled blood of their fetuses. The detection of HFPO dimer acid anion in the pooled fetus
plasma demonstrates gestational transfer from dam to fetus.
In the studies of rats dosed during pregnancy in which plasma concentrations in both the dams
and fetuses were measured at GD 20 (Dupont-18405-849 RV1. 2<~>11) the HFPO dimer acid
anion plasma concentration ratio for dams to fetuses is approximately three for the rat. In the
study of mice dosed during pregnancy (Dupont 18405-1037, 2010), plasma concentrations were
measured in dams on LD 21 and in pups on PNDs 4, 21. and 4<) If the plasma concentrations in
dams on LD 21 are assumed to be representative of those on 1.1)4. the comparison to pup plasma
concentrations on PND 4 indicate a dam-to-pup plasma concentration ratio of two to four.
Together these data indicate the efficiency of transfer in rats and mice is of a similar magnitude.
2.3.4 Metabolism
In an in vitro study, hepatocytes (1 x 106 cells/ml.) from male and female Sprague Dawley rats
were incubated with 5 micrometers of I II PO dimer acid ammonium salt for a total of 120
minutes (DuPont-23460, 2007). Samples were analyzed for I ll'PO dimer acid and suspected
metabolites at 0, 30, 45, 60, 90, and 12<> minutes I leat inacti\ ated hepatocytes were used as
negative controls. There was no difference in the concentration of IIFPO dimer acid between the
viable and heat-inacti\ ated hepatocytes. indicating that I ll'PO dimer acid ammonium salt is not
metabolized by rat hepatocytes No metabolites were detected In the single oral (gavage) study
of rats described in section 2 3 I (Absorption), the total accumulated amount of HFPO dimer
acid ammonium salt at I OS hours postdosinu in the urine collections accounted for 103% +
2.73% and 99 8% _ 6,41".. of the administered dose lor males and females, respectively, and
there was no detection of metabolites (DuPont-l K4<)5-1 "17 RV1, 2011).
2.3.5 t
Urine. Studies in rats, mice, and monkeys indicate that urine is the primary excretory pathway
for these Gen-X chemicals In the DuPont-18405-1017 RV1 study (2011), Sprague Dawley rats
(5/sex) administered a single oral (gavage) dose of 10 mg/kg HFPO dimer acid ammonium salt
excreted 95% to 97".. of the dose in urine within 12 hours and by 168 hours, the pooled urine
collections accounted for \ irtnally all of the substance administered with no evidence of
metabolic alteration. In a companion study, Crl/CD1(ICR) mice (5/sex) were administered a
single oral (gavage) dose of 3 mg/kg HFPO dimer acid ammonium salt (purity 84%) (DuPont-
18647-1017 RV1, 2011). Urinary elimination in mice was possibly less efficient than in the rats
given that only 31% (mean) and 39% (mean) of the dose material was found in the 12-hour
pooled urine for the males and females, respectively. At 168 hours postdosing, the mean values
for the pooled urine samples accounted for 90% and 92% of the total dose for the male and
female mice, respectively (DuPont-18647-1017 RV1, 2011). Based on the amounts in urine and
the clearance from blood, mice appear to have a lower ability to clear the HFPO dimer acid anion
by transferring it to urine in the early postexposure period than rats. The differences in the results
of these studies might have been influenced by the different doses given to the rats (30 mg/kg)
and the mice (3 mg/kg) (DuPont-18647-1017 RV1, 2011; DuPont-18405-1017 RV1, 2011).
14
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
The dynamic relationship across dose, exposure duration, and excretion observed in serum
measurements from the Rushing et al. study (2017) is also reflected in their data on urinary
excretion. Urine concentrations were monitored on exposure days 1, 2, 3, 5, 10, and 14. For the
1- and 10-mg/kg/day doses, urinary concentration peaked on day 3 and declined monotonically.
Males had higher urine concentrations than females at each time point, consistent with their
higher serum concentrations. For the 100-mg/kg/day dose group, the concentrations in urine
peaked at day 2 in males while females declined more slowly than at the lower doses.
Feces. Fecal elimination of HFPO dimer acid appears to be minor in rats and mice after acute,
subacute, subchronic, and chronic exposures. The data for combined fecal matter and cage-wash
(dried fecal matter) suggest that mice might lose slightly more III PO dimer acid through fecal
matter than rats. Low fecal excretion could reflect low le\ els of hepatic loss via biliary excretion
(DuPont-18405-1017 RV1, 2011; DuPont-18647-101 7 RV1. 2< > I 1)
2.3,6 Pharmacokinetic Clearance and
Clearance time. In multiple study reports, the study authors did not calculate pharmacokinetic
parameters such as Ty2 or area under the cur\ e and i nsiead defined the metric "clearance time" as
the time when 98.4% of the anion from the HFPO dimer acid ammonium salt was cleared from
the plasma.
A total of 12 Crl:CD(SD) rats, 3 per sex per dose. recei\ ed a single oral dose of 0, 10, or 30
mg/kg/day HFPO dimer acid ammonium salt (84 0% purity ) In uavage (Dupont-24281, 2008).
Plasma samples were collected from animals serially at <>. <> 25. n 5. 1,2, 4, 8, 12, 24, 48, 72, 96,
120, 144, and 168 hours In males, plasma levels peaked within the lirst 1-2 hours after dosing for
the low dose, and within the first 30 minutes to 1 hour for the high dose. By days 4 to 5, plasma
concentrations were less than 1% of the peak. In females, the plasma levels peaked at 1 hour for
the low dose and had usually declined to the LOO (2" ng/mL) by 24 hours. At the 30-mg/kg dose,
the plasma levels peaked at one-half to I-hour post-dosing and declined to theLOQ (20 ng/mL) by
24 or 28 hours for males and females, respectively In male rats, the authors identified 12 hours as
the clearance time at the low dose and 22 hours at the high dose (Table 3). In female rats, the
clearance \ alues were 4 hours and S hours for the low dose and high dose, respectively.
Table 3. Clearance Times in Male and Female Rats and Mice Following a Single Oral Dosea
Chemical
Male Rat
Male Mouse
Female Rat
Female Mouse
10 mg/kg
HFPO dimer acid ammonium sail
12 hr
143 hr
4 hr
57 hr
HFPO dimer acid
28 hr
ND
8 hr
ND
30 mg/kg
HFPO dimer acid ammonium salt
22 hr
139 hr
8 hr
62 hr
HFPO dimer acid
22 hr
ND
4 hr
ND
Notes', hr = hour; ND = no data.
a Adapted from Dupont-24281 (2008), Dupont-24286 (2008), and Dupont-25300 (2008) where clearance time is defined as the
time when 98.4% of the HFPO dimer acid ammonium salt was cleared from the plasma.
15
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
The same protocol was followed using the HFPO dimer acid (98% purity) (Dupont-24286,
2008). At the low dose, plasma concentrations peaked at 1 hour in both male and female rats,
while at the high dose, the peak plasma concentrations occurred in males at 1 or 2 hours and in
females at 15 minutes. The clearance times in males were 28 hours and 22 hours for the low dose
and high dose, respectively. The clearance times in females were 8 hours and 4 hours for the low
dose and high dose, respectively (Table 3).
The protocol outlined above was also followed with mice with a total of 12 Crl:CD(ICR) mice, 3
per sex per dose, receiving a single oral dose of 10, or 30 mg/kg/day HFPO dimer acid
ammonium salt (86% purity) by gavage (Dupont-25300 2<~>ns) Plasma samples were collected
from animals serially at 0, 0.25, 0.5, 1, 2, 4, 8, 12, 24, 4S. 72. 120, 144, and 168 hours
postdosing. Peak plasma HFPO dimer acid anion concentrations were reached within 8 hours for
the males and 4 hours for the females at the 10-mg/ku dose At the 30-mg/kg dose, the peak
HFPO dimer acid anion concentrations were reached within 2 hours for both males and females.
The mean clearance time was slower in the males (143 hours and I >> hours at the low dose and
high dose, respectively) than in the females (57 hours and 62 hours at the low dose and high
dose, respectively) (Table 3).
In the oral toxicokinetic studies, the clearance times were more rapid in rats than in mice and
were most rapid in female rats compared to male rats for both anions from the HFPO dimer acid
and its ammonium salt. In rats at the 10-mg/kg dose, the HFPO dimer acid took longer to clear
than its ammonium salt in both male and female rats. At 30 mg/kg dose, however, both the
HFPO dimer acid and its ammonium salt had the same clearance times in male rats, but the
HFPO dimer acid ammonium sail took longer to clear in female rats.
In a cross-species pharmacokinetic study, Sprague Dawley rats (3/sex) and Cynomolgus
monkeys (3/sex) were administered a single intravenous dose of the HFPO dimer acid
ammonium salt (10 mg ku) (DuPont-1775 1 -1579 RV1, 2009). Plasma samples were collected at
intervals o\ er the first 24 hours postdosing and once per day for the subsequent 7 days in the rats
and 21 days in the monkeys In the rats, the plasma concentrations were consistently higher for
the males than the females In approximately one to two orders of magnitude, consistent with the
indication that female rats ha\ e more rapid elimination. In the monkeys, the plasma levels were
nearly identical for the males and females over the first 24 hours. Beyond that point, the plasma
concentrations in the males were slightly higher than in the females. The levels of the anion from
HFPO dimer acid ammonium salt were very low at 168 hours in male (4 ng/mL) and female
(1 ng/mL) monkeys for 4<)S hours and beyond, concentrations were below the LOQ, which was
1 ng/mL.
Additionally, in this study, another group of Sprague Dawley rats (3/sex) received a 50-mg/kg
intravenous dose of HFPO dimer acid ammonium salt. Over the first 72 hours after the
intravenous injection, the HFPO dimer acid anion plasma levels in females were consistently
lower than in males by approximately one to two orders of magnitude (6703 ng/mL for males
versus 269 ng/mL for females at 12 hours and 776 ng/mL for males versus 7 ng/mL for females
at 24 hours). The standard deviations on each serum mean were broad, indicative of wide
differences across the three males and three females evaluated at that dose (Dupont-17751-1579
RV1, 2009).
16
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Half-lives. In Gannon et al. (2016), the goodness of fit was calculated for the plasma
concentrations after oral and intravenous dosing using one- and two-compartment models, and
the two-compartment model had a better fit. Pharmacokinetic parameters identified by Gannon et
al. (2016) are presented for the intravenous studies in Table 4 and for the oral studies in Table 5.
The alpha phase T1/2 represents the plasma concentration in the early postinjection period and is
considered to reflect the plasma distribution phase (Klaassen, 1996). The beta phase T1/2
represents the period during which the chemical in the plasma has established an equilibrium
with the levels in the body tissues and represents the elimination phase. The two-compartment
model is a refinement to the prior pharmacokinetic analysis in which the clearance time was
calculated. The two-compartment model better fits the data and separates distribution and
elimination phases; therefore, generally for comparisons across the datasets, the T1/2S are
preferred.
Table 4. T1/2 Estimates from the Intravenous Injection in Sprague Dawley Rats and
Cynomolgus Monkeys"
Tl/2
Intravenous Exposu res
Male Rat
Male Monkey
Female Kal
l-emale Monkey
Alpha (distribution) Phase (hours)
\6
2 ^
04
1.9
Beta (elimination) Phase (hours)
X<> 1
(4 1
22.6
79.6
Note: T1/2 = half-life.
a Adapted from Gannon et al. (2016).
In the intravenous injection studies, the Ti : of the alpha phase of distribution is similar (about 2
hours) for male and female monkeys The Ti : of the beta (elimination) phase in female monkeys
is longer than it is in the female rats, u liich could be a result of female monkeys having higher
tissue stores than female rats or clearance of I II PO dimer acid anion from their tissues might be
slower. There are 110 studies. ho\\e\ er. to distinguish these explanations such as a study of tissue
concentrations o\ er time. In rats, both the alpha and beta phases are shorter in females than in
males, u hile the beta phase Ti ; is about four times longer in males, suggesting higher levels in
tissues of males or slower clearance of HFPO dimer acid anion from their tissues (Gannon et al.,
2016). The crude metric ratio of the beta (elimination) phase T1/2 from intravenous dosing in
female rats (approximately I day) to female monkeys (3.3 days) is approximately 3, which
suggests a smaller magnitude of concern for rodent-to-primate interspecies differences in
toxicokinetics than is associated with other perfluorinated compounds (e.g., PFOA).
Gannon et al. (2016) also used the data from the single oral dose studies in rats and mice to
derive estimates of alpha and beta phase T1/2S to represent the distribution and elimination
phases. The oral exposure data are not ideal for this calculation because the chemical is not
directly injected into the blood. Because intestinal uptake of HFPO dimer acid anion from the
ammonium salt is believed to be rapid, and there appears to be no metabolism, the estimates are
reasonable for a two-compartment model.
In rats, following oral exposure, the alpha (distribution) T1/2 phase is more rapid in females than
in males and the beta (elimination) phase T1/2 is comparable for both sexes (Table 5). In mice, the
T1/2 estimates for both the alpha and beta phases are similar for both sexes and the clearance
times are shorter for females than for males (Table 5). The T1/2 estimated for the beta phase in
17
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
female rats is shorter from the intravenous data (22.6 hours) than from the oral gavage data (67.4
hours), while the other estimates of T1/2 from the intravenous and oral gavage data for males and
females are similar. While further study would be needed to better understand the differences in
toxicokinetics between males and females as well as between species, these differences are
generally threefold or less (e.g., the ratio of beta phase T1/2S in female rats and monkeys
described above).
Table 5. T1/2 Estimates in Sprague Dawley Rats and Crl/CD1(ICR) Mice Exposed to a
Single Oral Dose of HFPO Dimer Acid Ammonium Salt3
Tl/2
Oi'iil l.\|)osurcs
Male Rat
Female Kal
M ale Mouse
Female Mouse
Alpha (distribution) Phase (hours)
2.8
0.2
5.8
4.6
Beta (elimination) Phase (hours)
72.2
(."4
36.9
24.2
Note: T1/2 = half-life.
a Adapted from Gannon et al. (2016).
The time that it takes to achieve a balance between gastrointestinal uptake and excretion (i.e.,
steady state) following daily gavage exposures to the I II TO dimer acid anion is dependent on the
T1/2S of the alpha and beta phases. When the data are well described by a muliicompartmental
model, the steady state (more than 90%) is a function of the multiple T1/2S for the
intercompartmental distribution and elimination; however, at later times the elimination Tu is
expected to dominate the time to steady state and to be reached approximately within four T1/2S,
or 6.15 days, for male mice This was calculated by multiplying the beta phase T1/2 (36.9 hours)
by 4 and dividing that product by 24 hours The data from Rushing et al. (2017) for male mice
clearly demonstrate a lack of serum steady state for male mice after receiving doses of 1, 10, and
100 mg/kg/day for 28 days because the serum concentrations do not remain constant after the
expected 6 days T11 fact, the III TO dimer acid concentrations continue to change between 5 and
14 days and 14 and 2K days These continual changes in plasma concentration after 6 days
indicate dynamics o\ er multiple days that are not represented by typical multicompartment
models and. therefore, are not appropriate for modeling the complexity of the pharmacokinetics
of HFPO dimer acid and its ammonium salt.
Repeated-dose study. In a repeated-dose study with Crl:CDl(ICR) mice dosed with 0, 0.1, 0.5,
or 5 mg/kg/day lor al least l><) days, plasma measurements were made at 2 hours, 28 days, and 95
days (Dupont 18405-1 >7. 2<)|i)) Plasma concentrations increased less than twofold between the
2-hour and the 28-day measurements for both the males and females in all dose groups (Table 6).
Unfortunately, the study provides no measurements between the 2-hour and 28-day time points
to allow for a determination regarding steady state. As mentioned above, however, the Rushing
et al. study (2017) in mice provides measurements in serum at 1, 5, 14, and 28 days following
daily gavage dosing of C57BL/6 mice that clearly establish the lack of steady-state conditions,
which supports development of a more complex model to represent these data.
18
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Table 6. Mean Plasma Concentrations with Standard Deviations (SD) of Dosing with
HFPO Dimer Acid Ammonium Salt for at Least 90 Days"
Dose
mg/kg/day
Day 0
Day 28
Day 95
ng/mL
SD
COV
ng/mL
SD
COV
ng/mL
SD
COV
Males
0
ND
NA
NA
ND
NA
NA
ND
NA
NA
0.1
736
99
13%
1,124
238
21%
1,276
309
24%
0.5
3,806
1,197
31%
7,192
3,055
42%
7,068
2,398
34%
5
42,580
5,214
12%
52,240
16,725
32%
67,980
13,717
20%
Females
0
ND
NA
NA
ND
\ \
\ \
ND
NA
NA
0.1
824
72
9%
"04
'5ii
50" „
740
282
38%
0.5
3,606
1,308
36%
4.1'JS
1.239
30%
5.438
1,696
31%
5
35,340
9,262
26%
4<>.5X<)
16,842
36%
45,580
5741
13%
Note: COV = coefficient of variation (SD / mean); ND = not deluded: NA = not applicable.
a Adapted from Dupont 18405-1307 (2010).
Plasma concentrations remained relali\ ely constant between 2S days and 95 days for male and
female mice administered the 0.1-niu kg day dose in the Dupont 18405-1307 study (2010)
(Table 6). At the 0.5-mg/kg/day dose, plasma concentrations are relatively constant from day 28
to 95 days for the males, but the females" plasma concentrations increased from 4,198 ng/mL to
5,438 ng/mL (a 30°.. increase) This indicates that the I II PO dimer acid anion does not appear to
accumulate at 0.1 mg ku day. however, it might ha\ e accumulation potential at 0.5 mg/kg/day.
Interestingly, this increase in female plasma concentrations from 28 days to 95 days is equal to
the coefficient of variation (COV) in the 28-day measurement, thus the difference between days
28 and 95 could be due to interanimal differences in response to the same dose. Also interesting
is that, at the 5-mg'kg/day dose, female plasma le\ els returned to approximately the same levels
at 28 and l>5 days (46,580 and 45.580 ng/mL. respectively) (Table 6). In the males, the plasma
levels at 2S days increased from 52.240 ng/mL to 67,980 ng/mL at 95 days (a 30% increase),
again equaling the COV in the 2S-day measurement. Thus, the difference between days 28 and
95 could be due to \ ariability in these measurements as a result of interanimal differences and
might not necessarily reflect accumulation of HFPO dimer acid anion.
19
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
3.0 Problem Formulation
3.1 Conceptual Model
The conceptual model provides useful publicly available information to characterize and
communicate the potential health hazards related to oral exposure to HFPO dimer acid and its
ammonium salt. Figure 2 depicts in a conceptual diagram the sources of these GenX chemicals,
the routes of exposure to biological receptors of concern (e.g., human activities related to
ingested tap water such as drinking, food preparation, and consumption), the potential
assessment endpoints (e.g., effects such as liver toxicity), and populations at risk of exposure to
HFPO dimer acid and its ammonium salt. As outlined in the legend at the bottom right of Figure
2, the green boxes indicate where there are data available for these GenX chemicals. This
includes quantitative data for oral exposure to HFPO dimer acid and its ammonium salt, as well
as the limited data available for some of the potential sources of exposure to these chemicals.
The white boxes indicate that there are no data publicly available lo allow for determining if
these GenX chemicals are found in certain sources and that no human toxicity data exist.
20
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
NOVEMBER 2018
STRESSORS
POTENTIAL
SOURCES OF
EXPOSURE
EXPOSURE
ROUTES
ORGANS/
SYSTEMS
AFFECTED
HFPO Dimer Acid and its
Ammonium Salt
Drinking
Water
POTENTIAL
RECEPTORS IN
GENERAL
POPULATION
Ambient
Water
Industrial
Uses
Air
Food
Dust
Consumer
Products
Soil
Oral
Dermal
Inhalation
Liver Effects
Hematological
Effects
Developmental/
Reproductive
Effects
Kidnev Effects
Immune
Effects
Cancel-
Adults
Children
Pregnant Women
and Fetuses
Lactating Women
Fire
Fighting
Foams
LEGEND
Data Available
Data Unavailable
Figure 2. Conceptual Diagram for HFPO Dimer Acid and/or Its Ammonium Salt
21
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
3.2 Overall Scientific Objectives
This document provides the health effects basis for the development of oral RfDs for subchronic
and chronic durations for GenX chemicals, including the science-based decisions providing the
basis for estimating the POD. This section discusses the factors the EPA considers in the process
of developing a POD (depicted in Figure 2).
Stressors: Uses of GenX chemicals include as intermediates and as polymerization aids in the
production of fluoropolymers. These chemicals are two of several replacements for PFOA and its
ammonium salt and could have many applications in consumer products (e.g., stain- and water-
repellant textiles) and industrial processes (e.g., pharmaceutical and semiconductor
manufacturing). Publicly available data, although limited, indicate that sources of exposure to
GenX chemicals include both ground and surface waters used for drinking. Many other
potentially important sources of exposure to GenX chemicals exist given their use as a
replacement for PFOA, including foods, indoor dusl in a home or work environment, indoor and
outdoor air, soil, biosolids, and consumer products within the home, workplace, children's
schools, and daycare centers. Very little quantitati\ e information on these sources of exposure,
however, is available.
Routes of exposure: Nonoccupational exposure to GenX chemicals in water can occur through
oral exposure (i.e., drinking water, cooking with water, and incidental ingestion from showering)
and is expected to occur by dermal exposure (i e . contact of exposed parts of the body with
water containing GenX chemicals during hathing or showering, and dishwashing) and inhalation
exposure (e.g., volatilization of the GenX chemicals from the water during bathing or showering,
or while using a humidifier or \ aporizer). There is limited information identifying health effects
from inhalation or dermal exposures to GenX chemicals in animals Specifically, there are only
two acute dermal toxicity tests one dermal irritation study in rabbits, and one acute inhalation
toxicity test in rats. Repeated-dose toxicity data are available for oral exposure, but not for the
other exposure routes (inhalation and dermal exposures). The only quantitative data available for
HFPO dimer acid and its ammonium salt are for oral routes of exposure. Thus, this assessment
applies only to the oral route of exposure.
Receptors: The receptors are those in the general population (i.e., adults, the elderly, women of
childbearing age. pregnant women, and fetuses, infants, and children) who could be exposed to
GenX chemicals in tap water through ingestion, dermal contact, or inhalation at their homes,
workplaces, schools, and daycare centers. In the conceptual model in Figure 2, the box for adults
also includes sensiti\ e life stages (e.g., women of childbearing age and the elderly).
Endpoints: No human epidemiological studies for GenX chemicals are available. Oral exposure
studies of acute, subchronic, and chronic duration are available in rodent species, including rats
and mice. The recommended definitions of study duration were applied as outlined in A Review
of the Reference Dose and Reference Concentration Processes (USEPA, 2002). By this
approach, the employed study durations are as follows:
• Acute: Exposure by the oral, dermal, or inhalation route for 24 hours or less.
• Short-term: Repeated exposure by the oral, dermal, or inhalation route for more than 24
hours, up to 30 days.
22
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
• Subchronic: Repeated exposure by the oral, dermal, or inhalation route for more than
30 days, up to approximately 10% of the life span in humans (more than 30 days up to
approximately 90 days in typically used laboratory animal species).
• Chronic: Repeated exposure by the oral, dermal, or inhalation route for more than
approximately 10% of the life span in humans (more than approximately 90 days to
2 years in typically used laboratory animal species).
Adverse effects observed following exposure to HFPO dimer acid and/or its ammonium salt
include liver toxicity (e.g., hypertrophy, single-cell necrosis, peroxisome proliferation, and
increased liver weight relative to BW), hematological cffccls (eg. decreased red blood cell
(RBC) count, hemoglobin, and hematocrit), kidney toxicity (eg . increased kidney weight,
necrosis, and hyperplasia), developmental effects (e.g . Ii\\ changes), immune effects
(e.g., T cell-dependent antibody response (TDAR) suppression and lymphocyte increases), and
suggestive evidence of tumor formation (e.g., lh er and pancreatic acinar cell tumors).
In most of the available animal studies, hepatocellular hypertrophy and necrosis of the liver
appear to be the most sensitive effects obsen ed The increases in relali\ e I i \ er weight,
hepatocellular hypertrophy, and peroxisome acli\ ilv (e g . peroxisomal bela-oxielation induction)
can be associated with activation of cellular peroxisome piolilerator-activaled receptor-alpha
(PPARa) receptors, making it difficult to determine if this change is a reflection of PPARa
activation or an indication of GenX chemical toxicity. This is important because the PPARa
response could be more relevant to rodents than humans. The N\\ evaluated liver effects
resulting from exposure to GenX chemicals in the context of the I lull criteria (Hall et al., 2012),
where liver effects can he considered adverse when changes in li \ er weight or hepatocellular
hypertrophy are accompanied with necrosis, inflammation, and/or fibrosis. The observance of
liver necrosis indicates that cytotoxicity also could he a mode of action (MOA) for liver damage.
The toxicity values for this assessment include a chronic oral RfD (chronic RfD) and a
subclironic oral RfD (subchronic Rll)) lor I II PO dimer acid and its ammonium salt. An RfD is
an estimate of the concentration or dose of a substance (with uncertainty spanning perhaps an
order of magnitude) to u liich a human population (including sensitive subgroups) can be
exposed that is likely to he without an appreciable risk of deleterious effects during a lifetime. In
addition to chronic RfDs, other durations of exposure can be considered, including subchronic
exposures. RlDs are derived for noncarcinogenic toxicological endpoints of concern.
3.3 Methoc^
3,3,1 Literature and Results
To derive the RfD, the EPA assembled available information on toxicokinetics; acute, short-
term, subchronic, and chronic toxicity; developmental and reproductive toxicity; neurotoxicity;
immunotoxicity; genotoxicity, and cancer in animals. Most of the available data for HFPO dimer
acid and its ammonium salt were submitted to the EPA by DuPont/Chemours, the manufacturer
of GenX chemicals, under TSCA, including with PMNs, as required pursuant to a consent order
(USEPA, 2009) or as required under TSCA reporting requirements (e.g., section 8(e)).
To identify public literature available for HFPO dimer acid and its ammonium salt, a
comprehensive contractor-led search was conducted of four databases (PubMed, Toxline, Web
of Science (WOS), and TSCATS) using CASRN, synonyms, and additional relevant search
23
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
strings (see appendix A for a full list). Because the results of this core search were so limited,
additional databases were searched for physicochemical property information, health effects,
toxicokinetics, and mechanistic information. A list of the additional databases searched is
provided in appendix A. Combined, these database searches returned 27 studies for HFPO dimer
acid and its ammonium salt, after accounting for duplicates. The submitted studies and literature
identified by the search of publicly available sources are available through the EPA's Health &
Environmental Research Online (HERO) website at
https://hero.epa.eov/hero/index.cfm/proiect/page/proiect id/2627.
3.3,2 Study Screening Process and Study Evalu
For the publicly available literature, inclusion and exclusion criteria (outlined in appendix A)
reduced the database to potentially relevant studies Potential rele\ ance was based primarily on a
title and abstract screen.
For HFPO dimer acid and its ammonium salt, howe\ er. most of the a\ ai lahlc data were
submitted to the EPA under TSCA. Submitted test data on HFPO dimer acid and its ammonium
salt were available for numerous endpoints such as acute toxicity, metabolism and toxicokinetics,
genotoxicity, and systemic toxicity in mice and rats with dosing durations of up to 2 years. Most
of these submitted studies were conducted according to Ol-('l) TGs and/or EPA health effects
TGs for pesticides and toxic substances, which "are generally intended to meet testing
requirements for human health impacts of chemical substances under the Federal Insecticide,
Fungicide, and Rodenticide Act (FIFRA) and TSCA "" All available studies were considered for
inclusion (see appendices A and li). The majority of the studies selected for this assessment,
however, adhered to the Principles of GLP, and full study reports were submitted for Agency
review. As noted by OI-CI).' the OI X'D TGs are accepted internationally as standard methods
for safety testing and
are co\ ered In the Mutual Acceptance of Data, implying that data generated in the
testing of chemicals in an OECD member country, or a partner country having adhered to
the Decision, in accordance with OECD l est Guidelines and Principles of GLP, be
accepted in other OECD countries and partner countries having adhered to the Decision,
for the purposes of assessment and other uses relating to the protection of human health
and the environment.
In addition to these study quality considerations, the EPA OPPT evaluated all studies considered
for the derivation of the Rll )s usi ng quality criteria for various metrics published in the recent
Application of Systematic Review in TSCA Risk Evaluations (USEPA, 2018b). Studies were
evaluated according to the following domains—test substance, test design, exposure
characterization, test model, outcome assessment, confounding/variable control, and data
presentation and analysis. As discussed in appendix B, data evaluation is a qualitative assessment
of confidence in a study or dataset. A scoring system is applied to ascertain a qualitative rating in
order to provide consistency and transparency to the evaluation process. Applying the scoring
system results in assigning a confidence level rating of high, medium, low, or unacceptable. Any
score falling within the range of the confidence level is associated with the high, medium, or low
confidence levels. The system is not intended to imply precision and/or accuracy of the scoring
results. For example, any score between 1 and 1.7 is considered as a high confidence level, and
3 http://www.oecd.ore/chemicalsafetv/testine/oecdguidelinesforthetestingofchemicals.htm.
24
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
differences within that range should be interpreted with caution, as they might not necessarily
reflect a relative difference across studies (i.e., a study with score of 1.1 might not be different in
terms of quality than a study with a score of 1.3; both studies are judged to have a high
confidence level).
The GenX study reviews were conducted by EPA staff with expertise in toxicology, biology,
pharmacokinetics/modeling, and related fields with understanding of OECD TGs and risk
assessment. Each expert reviewed 1-2 studies. Following individual study reviews, a scientist
performed a final review of all reviewers' evaluations to provide a level of consistency across the
evaluations. The results of data evaluation for these individual studies are provided in appendix
B.
3.3,3 Approach for Derivation of Reference Va
Development of the hazard identification and dose-response assessment for HFPO dimer acid
and its ammonium salt has followed the general guidelines for risk assessment put forth by the
National Research Council (1983) and the EIWs Iramework for Human Health Risk Assessment
to Inform Decision Making (USEPA, 2014a) Additional EPA guidelines and other Agency
reports used in the development of this assessment include the following.
• Guidelines for Developmental / oxicity Risk . Issessment (IJSEPA, 1991)
• Guidelines for Reproductive toxicity Risk . Issessment (I SEP A, 1996)
• Guidelines for Neurotoxicity Risk . Issessment (I SI.IW. 11)98)
• A Review of the Reference / h>se and Reference ('oncentration Processes (USEPA, 2002)
• Guidelines for (\ircinogen Risk . Issessment (I Si; PA. 2t )t >5a)
• Supplemental (iunlance for. Issessing Susceptibility from l'.arly-Life Exposure to
Carcinogens (I SI .IW 2005h)
• A Framework for Assessing Health Risks oj Environmental Exposures to Children
(USEPA, 2006)
• Exposure Factors Handbook ( I SI - P A. 2d I I a)
• Recommended I Jse of Body Weight' 4 as the Default Method in Derivation of the Oral
Reference Dose (USEPA, 201 1 b)
• Benchmark Dose Technical Guidance Document (USEPA, 2012)
• ChiUI-SpeciJic Exposure Scenarios Examples (USEPA, 2014b)
• Guidance jor. Ipplying Quantitative Data to Develop Data-Derived Extrapolation
Factors jor Interspecies andIntraspecies Extrapolation (USEPA, 2014c)
• Application eview in TSCA Risk Evaluations (USEPA. 2018b)
The EPA's A Review oj the Rejerence Dose and Reference Concentration document describes a
multistep approach to dose-response assessment, including analysis in the range of observation
followed by extrapolation to lower levels (USEPA, 2002). The EPA conducted a dose-response
assessment to define a POD and extrapolated from the POD to an RfD. For HFPO dimer acid
and its ammonium salt, the EPA used benchmark dose (BMD) modeling to refine the POD in
deriving the RfD. The steps for deriving an RfD are summarized below.
Step 1: Evaluate the data to identify and characterize endpoints related to exposure to
GenX chemicals. This step involves determining the relevant studies and adverse effects to be
considered for BMD modeling. Once the appropriate data are collected, evaluated for study
25
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
quality, and characterized for adverse outcomes, endpoints are selected that the risk assessor
judges to be relevant and the most sensitive (typically defined by the NOAEL value).
Considerations that might influence selection of endpoints include data with dose-response,
percent change from controls, adversity of effect, and consistency across studies. The OPPT
evaluated all toxicokinetic, repeated-dose toxicity studies of 28 days and longer and potentially
relevant published in vitro studies using the approach describe in Application of Systematic
Review in TSCA Risk Evaluations (USEPA, 2018b). Results are provided in appendix B.
Step 2: Conduct BMD Modeling. Using the EP A's Benchmark Dose Technical Guidance
Document (2012), a benchmark response (BMR) is selected and BMD modeling is applied to the
endpoints selected as most relevant. The BMR is a predetermined change in the response rate of
an adverse effect. It serves as the basis for obtaining the benchmark dose lower limit (BMDL),
which is the 95% lower bound of the BMD. A family of BMD models are fit to the dose-
response data that describes the dataset of the identified adverse effect. From the family of
models, either a best fitting model with the corresponding BMD and BMDL is derived or, if no
adequate models are found, the NOAEL or loucst-ohserved-adversc-elVcci level (LOAEL)
identified in step 1 is used as the POD.
Step 3: Convert the POD to a human equivalent dose (IIKD). or point ol'departure human
equivalent dose (PODhed). The POD (either a BMDL. \OALL, or LOAEL) is then converted
to an HED using the EPA's Recommended I se of Body Weight' * as the Default Method in
Derivation of the Oral Reference Dose (I SI.IW 2<) I lb).
Step 4: Provide rationale for selecting I l\ IJF are selected in accordance with EPA guidelines
considering variations in sensiti\ ily among humans, differences between animals and humans,
the duration of exposure in the critical study compared to the lifetime of the species studied, and
the completeness of the toxicology database
Step 5: Calculate the chronic and suhchronic UfDs. The RlDs are calculated by dividing
POD]ii D by the selected I I
KID = PODhed
Total UF
where:
PODhed Calculated from the BMDL using a BW3 4 allometric scaling approach
consistent with LIW guidance (USEPA, 2011b).
UF = Total UF established in accordance with EPA guidelines considering variations in
sensitivity among humans, differences between animals and humans, the duration of
exposure in the critical study compared to the lifetime of the species studied, and the
completeness of the toxicology database.
3.3.4 Measures of Effect
The available dataset regarding the toxicity of these GenX chemicals includes in vivo and in vitro
studies. The in vivo studies were considered in the dose-response assessment for HFPO dimer
acid and its ammonium salt. The available data indicate that the liver, kidney, RBCs,
immunological responses, BW, and fetal development are adversely impacted by exposure to
26
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
GenX chemicals. In this analysis, all reported changes in relative organ weights were presented
as relative to BW (data relative to brain weight were not included). The endpoints presented in
this assessment represent potentially adverse effects that were statistically significantly different
(p < 0.05 or 0.01) from control unless otherwise noted. Additionally, statistically significant
changes from the control are presented as the percent change from control, unless otherwise
noted.
The animal studies demonstrated dose-related effects on the liver in rodent species (rats and
mice) following exposure to HFPO dimer acid and/or its ammonium salt for durations of 28 days
to 104 weeks. The studies and endpoints reviewed as possible critical studies and effects for
determination of the POD were evaluated for experimental design, data quality, and dose-
response identified through the range of experimental NOAELs/LOAELs. A route-to-route
extrapolation of oral toxicity data to derive an inhalation reference concentration was not
conducted due to data limitations. For example, no toxicokinetic data are available characterizing
the uptake of GenX chemicals through the lung for systemic distribution, and only one acute
inhalation toxicity study is available (DuPonl-l 775 1-723, 2009). This study identifies portal of
entry effects, albeit at a high dose. Tumors wciv also observed following oral exposure to GenX
chemicals; however, the tumor data failed to demonstrate a direct response to dose and thus were
not considered quantitatively.
4.0 Study Summaries
4.1 Acute Toxicity Studies
There are over 10 studies a\ ailable detailing the acute toxicity and irritation studies of HFPO
dimer acid and its ammonium salt. This section summarizes the available acute oral, dermal, and
inhalation toxicity studies for I II PO dimer acid and its ammonium salt as well as dermal and eye
irritation studies. Detailed study summaries are a\ ailahle in appendix C.
Oral Toxicity Se\ eral studies ha\ e e\ aluated oral toxicity in rats and mice from single doses of
theFII PO dimer acid ammonium salt at doses ranging from 1.5 mg/kgto 17,000 mg/kg
(DuPont-22ir,2. 2<)<)7. Dul'ont^lZo. 2007, DuPont-25438 RV1, 2008; DuPont-2-63, 1963;
DuPont-77<)-l)5. 1996). Also, male and female rats were evaluated with doses of 175-5,000
mg/kg III PO dimer acid (I)ul,ont-25S75. 2008). The rats and mice in these studies received a
single dose of the compound and were observed for clinical effects of toxicity for 14 days.
Four studies were conducted according to OECD TG 425 (OPPTS 870.1100) using the Up-and-
Down Procedure (DuPont-ZZ^Z. 2007; DuPont-25438 RV1, 2008; DuPont-25875, 2008;
DuPont-24126, 2007) Two studies that estimated approximate lethal doses (ALDs) did not have
identified TGs (DuPont-2-t>3, 1963; DuPont-770-95, 1996). For the HFPO dimer acid, the oral
median lethal doses (LDsos) were 1,730 mg/kg and 1,750 mg/kg in male rats and female rats,
respectively (DuPont-25875, 2008). For the HFPO dimer acid ammonium salt, the LD50 was
3,129 mg/kg for female rats (DuPont-22932, 2007); 1,030 mg/kg for female mice (DuPont-
24126, 2007); and 1,750 mg/kg for male rats (DuPont-25438 RV1, 2008). The estimated ALD
for male rats for the ammonium salt ranged from 5,000 mg/kg to 7,500 mg/kg (DuPont-770-95,
1996; DuPont-2-63, 1963).
The more common clinical signs observed across studies included wet fur, fur/skin stain or
discoloration, altered posture, and lethargy; changes in BW were also seen (DuPont-22932,
27
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
2007; DuPont-24126, 2007; Dupont-25438 RV1, 2008; DuPont-25875, 2008; DuPont-770-95,
1996). Effects in mice were observed after exposure to HFPO dimer acid ammonium salt (86%
purity) doses at 550 mg/kg and higher. Effects in rats were observed after exposure to either
HFPO dimer acid (98% purity) or its ammonium salt (82.6% to 99% purity) at doses of 175
mg/kg and higher (DuPont-22932, 2007; DuPont-25875, 2008).
Gross evidence of organ or tissue damage included discoloration of lungs, stomach, skin, lymph
nodes, liver, and/or esophagus (DuPont-22932, 2007; DuPont-25438, RV1 2008; DuPont-25875,
2008). Enlarged livers and enlarged hepatocytes were observed in young male rats following
single doses of 2,250, 3,400, or 5,000 mg/kg for HFPO dimer acid ammonium salt (DuPont-2-
63, 1963).
Dermal Toxicity. Two studies reported acute dermal toxicity of IITPO dimer acid ammonium
salt in rats or rabbits following acute dermal exposure (l)iil,ont-24113. 2007; DuPont-839-95,
1996). In an OECD TG402 (OPPTS 870.1200) study. 5,000 mu kg 111 PO dimer acid
ammonium salt (86% purity) was applied to sha\ ed. intact skin of male and female rats under a
semi-occlusive dressing for 24 hours. The dermal LI)-., was greater than 5.i)i)i) mg/kg (both
sexes). Erythema was observed only in females, u hercas hyperkeratosis and ulceration were
observed in some rats of both sexes. All dermal effects cleared In* 13 days post treatment
(DuPont-24113, 2007). In another study (in which no guideline is cited), HFPO dimer acid
ammonium salt (99% purity) was applied to shaved, intact skin of New Zealand white rabbits for
24 hours. The ALD was determined to lx- greater than 5.i)i)i) mu kg In this study, erythema
persisted for 13 days postapplication and was accompanied hy scaling and sloughing of skin.
One of the rabbits also exhibited necrosis lor 2 (•> days postapplication (DuPont-839-95, 1996).
Inhalation Toxicity. One study (conducted using GLI* Compliance Statement in compliance
with Title 40 of the Code of I 'ederal Regulations (CFR) part 792) evaluated acute inhalation
toxicity of HFPO dimer acid ammonium salt (84% purity) in male and female rats following a
single 4-hour nose-only exposure to aerosol concentrations of 0, 13, 100, and 5,200 mg/m3. The
median lethal concentration (LCso) was greater than 5,200 mg/m3. Red discharge from the nose,
eyes, and mouth was ohser\ ed in rats at doses of 100 and 5,200 mg/m3 for up to 2 days
postexposure No gross lesions were observed. Microscopic evaluation of respiratory tract tissue
(lung, larynx pharynx, trachea, and nose) from rats exposed to concentrations of 0, 13, and 100
mg/m3 detected no substance-related effects (DuPont-17751-723, 2009).
Dermal Irritation In an OF.CD TG 404 (OPPTS 870.2500) dermal irritation study, very slight-
to-well-defined erythema was observed in three male New Zealand white rabbits following a
single application of a <> 5-ml. aliquot of HFPO dimer acid ammonium salt (86% purity) in an
area of shaved skin for a period of 4 hours on the day of application. Erythema cleared by
24 hours postexposure (DuPont-24030, 2007).
Eye Irritation. New Zealand white rabbits were administered a single application of a 0.1-mL
aliquot of HFPO dimer acid ammonium salt (86% purity) to the lower conjunctival sac in an eye
irritation study conducted according to OECD TG 405 (OPPTS 870.2400). At 28 hours after
instillation of the compound, necrosis, corneal opacity, iritis, conjunctival chemosis (swelling),
discharge, and corneal injury were observed (DuPont-24114, 2007).
28
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
4.2 Short-Term Toxicity Studies
Seven-Day Toxicity Studies. Hepatic effects were observed in 6-week-old mice and rats of both
sexes in four 7-day studies (in which no TG is cited) evaluating repeated dose oral toxicity of
HFPO dimer acid and its ammonium salt (DuPont-24010, 2008; DuPont-25281, 2008; DuPont-
24116, 2008; DuPont-24009, 2008). Water was used as the vehicle control in all studies. Two 7-
day studies evaluated the toxicity HFPO dimer acid ammonium salt (86.6% purity) and HFPO
dimer acid (99% purity) at doses of 30 mg/kg/day in male mice and rats, respectively. In both
studies, a twofold increase in liver weight relative to control, cell necrosis of hepatocytes, and
hepatocellular hypertrophy were observed in all exposed animals (DuPont-24010, 2008; DuPont-
25281, 2008). A third 7-day study evaluating toxicity of HFPO dimer acid (99% purity) also
detected increased liver weight in male rats (at 30, 100, and 300 mg/kg/day) and in female rats
(at 300 mg/kg/day). Hepatocellular hypertrophy was present in both sexes at all doses (DuPont-
24116, 2008). Hypertrophy and increased liver weight were obser\ ed in another similar 7-day
gavage study evaluating effects of HFPO dimer acid ammonium sail (S6.6% purity). Males
appeared to be more sensitive to hepatic effects because increases in li \ er weight were observed
at 30, 300, and 1,000 mg/kg/day, whereas increased liver weight was observed in females only at
1,000 mg/kg/day. These effects were accompanied by increases in P-oxidation and increases in
cytochrome P450 enzyme activity, biomarkers for activation of PPARa nuclear receptors.
Mild-to-minimal hepatocellular hypertrophy was obser\ed in both sexes at 1,00<) mg/kg/day
(DuPont-24009, 2008).
Twenty-Eight-Day Toxicity Studies. Two 2S-day studies e\ alualing systemic toxicity in rats
and mice are available for HFPO dimer acid ammonium salt
DuPont-24447 (200M
In a study with 7-week-okl Crl ("I)(SI)) rats (1 <> sex group) conducted according to OECD TG
407, HFPO dimer acid ammonium salt (purity SS'\.) was administered on 28 consecutive days
via gavaue (\ ehicle was deioni/ed water) (Ol-('l). 2<>08; DuPont-24447, 2008). Male rats
received doses of 0. n 3. 3. or 3d mg kg day w hile females received 0, 3, 30, or 300 mg/kg/day.
In this study, there were no mortalities and clinical signs were confined to high-dose females
(e.g., urogenital staining)
Hematological e\ aluation re\ ealed statistically significantly decreased RBC count, hemoglobin,
and hematocrit at greater than or equal to 3 mg/kg/day in males. The maximum decreases
compared to control at 4 weeks were observed at the highest dose (30 mg/kg/day) and were 6%,
7%>, and 8% for RBC count, hemoglobin, and hematocrit, respectively. Increases in absolute
reticulocyte counts were also observed in males at all dose levels, but this increase was only
statistically significant from control at the highest dose (27%) at 4 weeks. No statistically
significant hematological effects were observed in the females (DuPont-24447, 2008).
Alterations in serum clinical chemistry parameters were seen in both sexes, but most of the
significant effects were observed in the male rats. Decreases in total globulin and increases in the
A/G ratio were observed in males and females. In males, total serum albumin increased (15% at
30 mg/kg/day) while total globulin decreased 13% and 22% compared to control at 3 mg/kg/day
and 30 mg/kg/day, respectively. This resulted in an increase in the A/G ratio to 16% and 41% in
the 3 mg/kg/day and 30 mg/kg/day males, respectively, most likely due to underproduction of
globulin. Females exhibited a 9% decrease in total globulin and a 20% increase in the A/G ratio
29
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
compared to control at 300 mg/kg/day. Males also showed statistically significant decreases in
serum cholesterol at all doses, with the largest decrease compared to control (28%) in the
30-mg/kg/day group. Triglyceride levels were decreased at all doses, but were significantly
decreased (22%) only at 3 mg/kg/day. Males also exhibited increases in blood urea nitrogen
(BUN) (24%) and glucose (15%) at 30 mg/kg/day when compared to controls (DuPont-24447,
2008).
In males, relative kidney weight was significantly increased (15% compared to control) only at
the highest dose tested. Minimal mineralization of the kidneys was also observed in 1/10 male
rats in the high-dose group. There were no statistically significant changes in kidney weight in
the females; however, there was minimal basophilic staining of some cells in the tubules for 3/10
female mice in the 300-mg/kg/day group, while none were observed in the control group. Dose-
response could not be determined for basophilic tubules because no rats were examined in the 3-
mg/kg/day dose group and only one rat was examined in the 3<)-mg ku'day dose group. No
statistical analyses were completed on these microscopic observations
Relative liver weights were statistically increased in a dose-response manner in males, 19%
and 56%) compared to control at 3 mg/kg/day and 3d mg/kg'day, respecli\ ely These increases
were accompanied by decreases compared to control in sorbitol dehydrogenase (SDH) at 0.3
mg/kg/day (-36%) and 30 mg/kg/day (-21%) in males In females, the only statistically
significant change in liver weight was a 12".) increase compared to control at the highest dose
(300 mg/kg/day). Microscopically. 4 I n and 7 11) male rats exhibited hepatocellular hypertrophy
at 3-mg/kg/day and 30-mg/kg/day doses, respecti\ ely Tn female rats, hepatocellular hypertrophy
was observed in 4/1 n mis in the high-dose group. Hepatocellular necrosis (3/10) and single-cell
necrosis (1/10) were ohser\ ed in males at 30 mg kg day No statistical analyses were completed
on these histological ohser\ ations. The authors note that hepatic peroxisomal P-oxidation activity
was induced in both sexes at the middle and high doses. Specifically, P-oxidation activity was
determined using [ 14CJ palmitoyl coenzyme A (Co.\) as the substrate and total cytochrome P-
450 content as markers of peroxisome proliferation In the males, P-oxidation activity was
significantly increased compared to control at dosages of 0.3 mg/kg/day, 3 mg/kg/day, and 30
mg/kg/day by 42%, 274%, and 772%, respectively, and total cytochrome P-450 content was
significanlly increased by 23".. at 30 mg kg/day (DuPont-24447, 2008). In female rats dosed
with 30 mg kg day and 30<) mg'kg/day, p-oxidation activity was significantly increased
compared to control by 49% and 198%, respectively, while total cytochrome P-450 content
remained unaltered (DuPont-24447, 2008). The EPA identified the NOAEL to be 0.3 mg/kg/day
and the LOAEL to he 3 mu kg day based on hematological (decreased hemoglobin, RBC count,
and hematocrit) and immune (decreased globulin levels) findings in males (DuPont-24447,
2008). These findings were also accompanied by liver effects, including an increase in relative
liver weight and hepatocellular hypertrophy; however, necrosis was observed only at the high
dose (30 mg/kg/day).
DuPont-24459 (2008)
In another repeated-dose study conducted according to OECD TG 407, 7-week-old Crl:CD-l
mice (10/sex/group) were administered 0, 0.1, 3, or 30 mg/kg/day HFPO dimer acid ammonium
salt (purity 88%) for 28 consecutive days via gavage (vehicle was deionized water) (DuPont-
24459, 2008). Increases in mean BW gain were observed at 30 mg/kg/day in both males and
females. In males, increases in mean cumulative BWs were reported as statistically different
30
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
from the control group in the 30-mg/kg/day group during study weeks 1, 2, 3, and 4. In females,
mean cumulative BW gains were significantly increased in the 30-mg/kg/day group during study
weeks 2, 3, and 4.
Similar to the findings observed in the 28-day toxicity study in Crl:CD(SD) rats (DuPont-24447
2008), decreases of 5.0% in hemoglobin and hematocrit were reported at greater than or equal to
3 mg/kg/day, and RBC count was significantly decreased by 7.6% in the Crl:CD-l male mice at
30 mg/kg/day. In both males and females, the A/G ratio was statistically increased compared to
control at greater than or equal to 3 mg/kg/day. Albumin alone was significantly increased by
31.3%) compared to controls in males at 30 mg/kg/day, and globulin alone was decreased in
females at greater than or equal to 3 mg/kg/day by 15.8% and 21 I % at 3 mg/kg/day and
30 mg/kg/day, respectively. Finally, in males, the serum liver enzymes aspartate
aminotransferase (AST) (478%), alanine aminotransferase (ALT) (1,254%), alkaline
phosphatase (ALP) (1,222%), and SDH (1,800°/..) were significantly increased from control at
the 30-mg/kg/day dose.
In male mice, no statistically significant effect was observed on kidney weight. Female kidney
weight findings were equivocal with the mean relali\ e kidney weight showing statistically
significant increases compared to control only at the low dose (8%) and high dose (17%).
Minimal increases in basophilic tubular cells and tubular dilatation were observed in females at
30 mg/kg/day (3 of 10 animals for both effects) (DuPont-24451). 2008).
Macroscopic and microscopic tissue pathology evaluations were conducted for all dose groups.
The inspection of male adrenal cortex at the highest dose found minimal hypertrophy in 8 of 10
tissue samples examined, while females showed mild or minimal adrenal cortex congestion at
only the highest dose (l)nl,ont-24459, 2008). No statistical analyses were completed on these
microscopic observations
Liver effects were also reported in both males and females in this study. In males, relative liver
weights were significantly increased compared to control at 3 mg/kg/day and 30 mg/kg/day by
78%) and I (->3%, respecti\ ely In females, relative liver weights were increased at 3 mg/kg/day
and 3d mu kg'day by 32".. and I <>3% compared to controls, respectively. Absolute liver weights
also increased at these doses in both sexes and to similar extents. Increases in liver weight
correlated with microscopic liver findings (including single-cell necrosis, increased mitosis, and
hepatocellular hypertrophy). Single-cell necrosis was observed in 40% (4/10) and 100% (10/10)
of the male mice at 3 mg/kg/day and 30 mg/kg/day, respectively, while no liver necrosis was
observed in the control mice As noted above, serum liver enzymes were significantly increased
from control at the 30-mu kg day dose: AST (478%), ALT (1,254%), ALP (1,222%), and ADH
(1,800%)). Single-cell necrosis was also detected in 40% (4/10) of female mice at 30 mg/kg/day
compared to zero in the control. This was associated with an increase in serum SDH (186%) at
30 mg/kg/day. Hepatic peroxisomal P-oxidation activity was induced in both sexes. Specifically,
P-oxidation activity was determined using [14C] palmitoyl CoA as the substrate and total
cytochrome P-450 content as markers of peroxisome proliferation. In the male mice, P-oxidation
activity significantly increased compared to control at doses of 0.1 mg/kg/day, 3 mg/kg/day, and
30 mg/kg/day HFPO dimer acid ammonium salt by 57%, 744%), and 648%), respectively, yet
total cytochrome P-450 content significantly decreased at 3 mg/kg/day and 30 mg/kg/day by
26%) and 53%, respectively (DuPont-24459, 2008). P-oxidation activity significantly increased
relative to control in female mice at 3 mg/kg/day and 30 mg/kg/day by 495%) and 823%),
31
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
respectively, with no alterations in total cytochrome P-450 content. The EPA identified the
NOAEL for this study as 0.1 mg/kg/day and the LOAEL as 3 mg/kg/day based on increase in
single-cell necrosis in males, which was accompanied by increased relative liver weight and
hepatocellular hypertrophy, hematological, and immune effects.
4.3 Subchronic Toxicity Studies
DuPont-17751-1026 (2009)
In a repeated-dose study with rats, HFPO dimer acid ammonium salt (purity 84%) was
administered to 8-week-old Crl:CD(SD) rats (10-20/sex/dose) on 90 consecutive days via oral
gavage (vehicle was deionized water) in accordance wiih OECD TG 408 (DuPont-17751-1026,
2009; OECD, 1998). Male rats were administered the icsl substance at doses of 0, 0.1, 10, or 100
mg/kg/day while females received 0, 10, 100, or l,0On mg kg day [n this study, three high-dose
females died before dosing was complete (two deaths considered as treatment-related; one death
of undetermined cause).
Hematological evaluations revealed decreased hemoglobin, erythrocyte counts, and hematocrit in
males administered greater than or equal to 1" mu kg/day. The decreases in all three parameters
for males were significantly different from control at I <> and I <~>0 mg/kg/day and decreased in a
dose-dependent manner at 90 days (study week 13). The maximum decreases from control in
males were observed at the highest dose and were 1 I"... 13"... and 12% for RBC count,
hemoglobin, and hematocrit, respecti\ ely Likewise, female rats exhibited significant and dose-
dependent decreases in RBC count (2S".>). hemoglobin C21°..), and hematocrit (18%), but only at
the 1,000-mg/kg/day dose. In males, absolute (52"..) and percent ((->7%) reticulocytes and platelet
count (17%>) were significantly increased from control at the highest dose and exhibited a dose-
response. Additionally, both the absolute and percent of basophils (a type of white blood cell)
were significantly decreased relati\e to control at l<> mg/kg/day (25%) and 100 mg/kg/day (50%)
in males. Finally, female rats saw significant increases from control in mean corpuscular volume
(15%>), mean corpuscular hemoglobin (II"..), mean corpuscular hemoglobin concentration (4%),
platelet count (30%.). and absolute (212%) and percent (392%) reticulocytes and a decrease
relative to control in the percent of basophils (33%) at the high dose (1,000 mg/kg/day) (DuPont-
17751-H)2(v 2009).
There were alterations in the clinical chemistry values in both sexes. Males exhibited a dose-
dependent increase in total albumin and the A/G ratio and a decrease in total globulin compared
to control. These changes were statistically significant at 10 mg/kg/day and 100 mg/kg/day. The
maximum increases compared to control observed at the highest dose in total albumin, total
globulin, and A/G ratio were 12".., 15%, and 35%, respectively. As in the 28-day study, females
exhibited a dose-dependent decrease in globulin (33%) and an increase in A/G ratio (58%) that
was significantly different from control for both effects at the highest dose only. Males and
females also showed dose-dependent decreases in serum cholesterol that were statistically
significantly different from control at 100 mg/kg/day (31%) in males and at both 100 mg/kg/day
(20%) and 1,000 mg/kg/day (31%) in females. BUN was significantly increased relative to
control in males at 100 mg/kg/day (38%). The trend for BUN was dose-related and positive in
both sexes. ALP levels were significantly increased from control in a dose-dependent manner at
10 mg/kg/day (48%) and 100 mg/kg/day (106%) in the males and at 1,000 mg/kg/day (66%) in
the females. Serum phosphorus levels increased dose-dependently in males and females and
were significantly different from control at 10 mg/kg/day (10%) and 100 mg/kg/day (11%>) in
32
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
males and at 1,000 mg/kg/day (18%) in females. Total bilirubin was significantly decreased from
control in a dose-dependent manner at the mid-dose (25%) and high dose (50%) only in females.
Total protein and y-glutamyl transferase decreased 10% and 69%, respectively, at the high dose
in females. Finally, a slight but significant and dose-dependent decrease compared to controls in
urine pH (8%) and a large increase in total urine volume (252%) were observed in female rats at
1,000 mg/kg/day (DuPont-17751-1026, 2009).
Kidney weight relative to BW was significantly and dose-dependently increased from control at
10 mg/kg/day (13%) and 100 mg/kg/day (16%) in male rats. Likewise, kidney weight relative to
BW was significantly increased at all dose levels in females and reached a maximum increase of
23% from control; however, microscopic damage of the kidney (tubular and papillary necrosis)
was observed in only one of the rats at the highest dose Additionally, one of the females that
died prior to study termination exhibited tubular and pupillary necrosis of the kidney.
Transitional cell hyperplasia and mild acute inflammation were ohser\ ed in the kidney of 1/10
male rats at the 100-mg/kg/day dose. Statistical analyses were not completed for the microscopic
renal findings.
Liver weight relative to BW was significantly and dose-dependently increased from control at
10 mg/kg/day (31%) and 100 mg/kg/day {61%) in male rats I 'einales exhibited an 85% increase
from control in liver weight at the high dose (1,000 mg kg day) I lepatocellular hypertrophy was
observed in 3/10 and 10/10 males at the I n-mg/kg/day dose and I 00-mg/kg/day dose,
respectively, and in 10/10 females at the I .<><><>-mg/kg,day dose Statistical analyses were not
conducted for hepatocellular hypertrophy 1'iirtherniore. it is not documented in the data tables
whether other histological effects such as li\ er necrosis were detected in the 90-day study,
although the pathology report states that the hypertrophy was not associated with microscopic
changes indicative of li \ er injury such as necrosis (DuPont-1775 1-1026, 2009). The EPA has
determined the study NOALL to be 0.1 mg/kg day and the LOAEL to be 10 mg/kg/day based on
blood effects (i.e., decreased RBC count, hemoglobin, and hematocrit) in males.
DuPom- 1S405-1M)" (2010)
In a repeated-dose subchronic study with 7-week-old Crl:CDl(ICR) mice, the HFPO dimer acid
ammonium salt (purity 84°.7) was administered to 10 animals/sex/group for 95 days (males) or 96
days (females) \ ia gavage (\ chicle was deionized water) at doses of 0, 0.1, 0.5, and 5 mg/kg/day
in accordance with OECD TG 408 (DuPont-18405-1307, 2010; OECD, 1998). A statistically
significant increase in male BW and overall BW gain was observed at the high dose only. Mean
daily food consumption was statistically increased in males between days 0 and 91 in a dose-
related manner.
A small decrease compared to control in mean corpuscular hemoglobin concentration (3%) in
males and increased bilirubin (14%) in females were reported at 5 mg/kg/day. Clinical chemistry
changes were more evident among male mice than female mice. Specifically, AST, ALT, and
ALP were statistically increased from control 106%, 420%), and 1,134%), respectively, at the
5-mg/kg/day dose in males. Comparatively, female mice saw significant increases relative to
control in ALT (42%) and ALP (143%). SDH levels significantly increase compared to control
in both males (308%) and females (32%) at 5 mg/kg/day. Albumin levels were increased relative
to control in the 5-mg/kg/day dose group in both males (14%) and females (4%), but total serum
protein was significantly increased (14%) only in males at this dose (DuPont-18405-1307, 2010).
33
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Macroscopic and microscopic tissue pathology evaluations were conducted for all dose groups.
Male mice exhibited kidney tubular epithelial hypertrophy (9/10 treated mice compared to 0 in
control) while females exhibited dilated kidney tubules (4/10 in treated compared to 2/10 in
control) in the 5-mg/kg/day dose group. Both effects were classified as minimal by the study
authors. Female mice exhibited a decrease in relative spleen weight (10%, 21% and 18% at 0.1
mg/kg/day, 0.5 mg/kg/day, and 5 mg/kg/day, respectively). No effects on the spleen were
observed in male mice in any dose group. The study authors reported that changes in female
spleen weight did not occur in a dose-related manner and were not associated with changes in
absolute spleen weights or histological abnormalities in the spleen (DuPont-18405-1307, 2010).
Increased relative liver weights compared to control in both mule mice (130%) and female mice
(69%) were accompanied by minimal-to-mild hepatocellular hypertrophy at 5 mg/kg/day in all
dosed mice. Minimal hepatocellular hypertrophy was also ohser\ ed at the 0.5-mg/kg/day dose as
well in males (8/10 mice). No hepatocellular hypertrophy was ohser\ ed in the control group.
Large and discolored livers were observed at doses urcater than or ecmal to 0.5 mg/kg/day in
males, but only in the 5-mg/kg/day dose group in females. Key treatment-related findings
considered as adverse at 5 mg/kg/day included increased enzymes indieati\ e of liver injury
(i.e., AST, ALT, ALP, and SDH) and increased total bile acids that co-occurred with
histopathological findings in the liver. Histopatholouical findings in male mice included an
increase in the incidence of single-cell necrosis (10/10 treated mice versus 0 in control), Kupffer
cell pigments (10/10 treated mice versus 0 in control), and mitotic figures (9/10 treated mice
versus 0 in control). Females also exhibited histopathological li \ er findings, but to a lesser
degree. For example, 3/10 female mice exhibited focal necrosis and only 1/10 mice presented
single-cell necrosis at 5 mu kg day (DuPont-l S4<)5-l 3<)7. 2<)|0)
The EPA concluded that the \OALI. in this study is 0.5 mg/kg/day based on the histological
findings for the liver (i e . necrosis and mitotic figures) accompanied by the clinical chemistry
changes (i.e., AST, Al.'l'. ALP. and Sl)l I) at a I.OALL of 5 mg/kg/day.
4.4 ¦ ¦ . ¦ jdies
DuPojh-i *-105-1:3* cum
In a combined chronic toxicity carcinogenicity study in 7-week old Crl:CD(SD) rats (DuPont-
18405-1238, 2013), LLLPO dimer acid ammonium salt (purity 84%) was administered by oral
gavage (vehicle was deionized water) for up to 104 weeks (80/sex/group, of which 10/sex/group
were designated lor a 12-month interim necropsy in accordance with OECD TG 453) (DuPont-
18405-1238, 2013; OLCI). 2<)oi). Kae et al., 2015). Dose levels administered were 0,0.1, 1, and
50 mg/kg/day for males and <>. 1, 50, and 500 mg/kg/day for females.
Mean survival was unaffected by treatment. All females were sacrificed before study termination
at 101 weeks, however, because of decreased survival across all groups, including the control.
There were no statistically significant differences in survival across groups. The females in the
high-dose group were observed to have papillary necrosis and inflammation of the kidneys that
were deemed by the authors to be related to treatment. BW and BW gain were unaffected in
males but reduced compared to control (13% and 20%, respectively) in high-dose females at
52 weeks. The incidence of alopecia and hypotrichosis (abnormal patterns of hair growth) was
statistically significantly increased in females at 500 mg/kg/day.
34
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Statistically significant hematological effects were observed in this study, primarily in female
rats. Blood samples were taken at 3, 6, and 12 months. At 3 months, RBC count, hemoglobin,
and hematocrit were significantly decreased at the highest dose in males and females, although
these decreases did not occur in a dose-dependent manner. Similarly, at 6 months, hemoglobin
and hematocrit were significantly decreased at the highest dose in males, yet these decreases did
not occur in a dose-dependent manner. There were no significant differences in any of these
parameters in male rats at the 12-month time point. At 6 and 12 months, female rats exhibited a
significant decrease in RBC count, hemoglobin, and hematocrit at 500 mg/kg/day and in a dose-
dependent manner. The RBC count was also significantly decreased at 50 mg/kg/day in females
at the 12-month time point; however, hemoglobin and hematocrit were not. The largest decreases
compared to control in RBC count, hemoglobin, and hematocrit in female rats were 28%, 24%,
and 20%), respectively, which were observed at 12 months. Additionally, the percent change
from control of these effects increased over time (i.e., 3 months (•> months < 12 months). At
12 months, serum albumin levels increased in males at 1 nig/kg/day and 50 mg/kg/day by 8%>
and 16% from control, respectively, which led to a concomitant increase in the A/G ratio by 16%>
and 28%o, respectively.
Statistically significant changes from control were ohser\ ed in the kidneys of females, but only
at the highest dose (500 mg/kg/day). For example, there were increased incidences of tubular
dilatation (increased by 34%> compared to control), edema of the renal papilla (increased by 56%>
compared to control), transitional cell hyperplasia (increased In 39% compared to control),
tubular and pelvic mineralization (increased In I 5".. and 24".. compared to control, respectively),
renal papillary necrosis (increased by 23".. compared to control). and chronic progressive
nephropathy (increased In 3(¦>".. compared to control). all statistically significant from control.
These microscopic indications of kidney damage were also associated with a 15% increase in
relative kidney weight compared to control in females administered 500 mg/kg/day of HFPO
dimer acid ammonium salt.
Liver enzyme le\ els also were affected In exposure to IIFPO dimer acid ammonium salt at
12 months in the chronic study. In males, statistically significant increases in ALP (180%), ALT
(228"..), and SDN (141"..) were obser\ed at 50 mg/kg/day. These enzyme changes were
correlated with microscopic I Hidings in the liver, including focal necrosis. Relative liver weights
were increased in high-dose males (1 (•>".. compared to controls) and females (69% compared to
controls) at the 12-month sacrifice. The change in liver weight in females corresponded to
centrilobular hypertrophy in the high-dose females at the interim sacrifice. Females exposed to
500 mg/kg/day of ill PO dimer acid ammonium salt for 2 years also had significantly increased
relative liver weights (43°.. compared to control) at terminal sacrifice. There was no difference in
organ weights in males at any dose at terminal sacrifice despite the changes observed at 12
months. Male and female rats exposed to 50 mg/kg/day and 500 mg/kg/day, respectively, had
statistically significantly increased centrilobular hepatocellular hypertrophy compared to control
rats (7/70 in treated males compared to 0/70 in control; 65/70 in treated females compared to
0/70 in control) and centrilobular hepatocellular necrosis (5/70 in treated males compared to 1/70
in control; 7/70 in treated females compared to 1/70 in control). Male rats also saw a decrease in
incidence from control of 16% and 10% in focal and periportal vacuolization, respectively, at
50 mg/kg/day, and female rats had a 4% decrease from control in centrilobular vacuolation at
500 mg/kg/day. Finally, in females, panlobular hepatocellular hypertrophy (increase in incidence
compared to control of 4%), individual cell hepatocellular necrosis (increase in incidence
35
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
compared to control of 4%), and angiectasis (i.e., dilation of a blood or lymphatic vessel)
(increase in incidence compared to control of 6%), were reported at the high dose.
Nonneoplastic effects also were observed in the stomach and tongue of females exposed to the
high dose. Specifically, there were increased incidences of hyperplasia of the limiting ridge of
the nonglandular stomach (increased by 13% compared to control; incidence was 9/70 for treated
females and 0/70 in control) and of the squamous cell in the tongue (increased 16% from control;
incidence was 13/70 in treated females and 2/70 in control). The tongue also exhibited an
increased incidence of inflammation (increased 14% from control; incidence was 13/70 in treated
and 3/70 in control). The EPA concluded that the NOAEL for chronic toxicity in this study was 1
mg/kg/day and the LOAEL was 50 mg/kg/day for the liver effects in males.
Statistically significant increases in the incidence of li\ er minors in females at 500 mg/kg/day
and pancreatic acinar cell tumors in males at 50 mg/kg day were reported. An increase in
testicular interstitial (Leydig) cell tumors was noted at the high dose hut was not statistically
significant. Because of the observed early deaths in both control and treated animals, the EPA
recommended that the submitter (a) reexamine their test data, (b) identify the animals that died
without tumor within the first year, (c) exclude the animals identified in the pre\ ious step
(i.e., those that died within the first year and had no tumors) from consideration for cancer data
analysis, (d) recalculate tumor incidences, and (e) perform statistical analyses. Because the initial
results indicated that the increased incidences of liver tumors in female rats (500 mg/kg-d) and
combined pancreatic acinar tumors in male rats (5<~> mg/kg-d) were significantly increased from
control despite the inclusion of early deaths, the I-PA agreed to limit the reanalysis to testicular
hyperplasia and tumors in male rats only Additional discussion of tumor findings for the liver,
pancreas, and testes is presented below
Females. There were increases in the incidence of li\ er tumors at the high dose only (500
mg/kg/day), where degenerative and necrotic changes were also observed. The tumor incidences
were 0/70 (0%), 0/70 (0%), 0/70 (<)"..), and II 7<) (I 5 7%) for hepatocellular adenomas and 0/70
(0%). i) 7<) (H%), 0/70 (0%), and 4 7<> (5 7'\>) for hepatocellular carcinomas at the doses of 0, 1,
50, and 5<)i) mg/kg/day, respectively The increased incidences of hepatocellular adenomas were
statistically significant by the Cochran-Armitage trend test, the Peto test, and the pairwise Fisher
Exact test and the increased incidences of hepatocellular carcinomas were statistically significant
by the Cochran-Armitage trend test and the Peto test. The incidences of adenomas and
carcinomas observed at 500 mg/kg/day also exceeded the test laboratory historical control ranges
of 0%—5% and 0%— 1.7%, respectively.
Males: A statistically significant increase was reported in the incidence of pancreatic acinar cell
adenomas/carcinomas combined (but not adenomas or carcinomas alone) at 50 mg/kg/day.
Incidences of pancreatic acinar cell adenomas were 0/70, 1/70 (1.4%), 0/70 (0%), and 3/70
(4.3%) at 0 mg/kg/day, 0.1 mg/kg/day, 1 mg/kg/day, and 50 mg/kg/day, respectively. The
increased incidence at the high dose was not statistically significant and was within the test
laboratory historical control range (0%—5%). The incidence of pancreatic acinar cell carcinomas
was 0/70 (0%>) in all groups other than the high-dose group, in which 2/70 (2.9%) were observed.
The incidence of carcinomas at 50 mg/kg/day was not statistically significant but was slightly
higher than the upper end of the laboratory's historical control range (0%—1.7%). When these
two tumor types were combined, the incidences of adenoma/carcinoma were 0/70 (0%), 1/70
(1.4%>), 0/70 (0%>), and 5/70 (7.1%) at 0 mg/kg/day, 0.1 mg/kg/day, 1 mg/kg/day, and
36
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
50 mg/kg/day, respectively, with the increased incidence at the high dose significant by the
Cochran-Armitage trend test and the Peto test. For reference, the incidences of pancreatic acinar
cell hyperplasia were 16/70 (22.9%), 18/70 (25.7%), 7/70 (10%), and 21/70 (30%) at
0 mg/kg/day, 0.1 mg/kg/day, 1 mg/kg/day, and 50 mg/kg/day, respectively, indicating a lack of
dose-response relationship for this finding. Furthermore, the increased incidence of hyperplasia
at the high dose was not statistically significant (compared to control).
In the testes, the incidences of interstitial cell adenomas were 4/70 (5.7%), 4/70 (5.7%), 1/70
(1.4%), and 8/70 (11.4%) at 0 mg/kg/day, 0.1 mg/kg/day, 1 mg/kg/day, and 50 mg/kg/day,
respectively at 2 years. An interstitial cell adenoma was also present in 1/10 high-dose males at
the interim sacrifice (12 months). The increased adenoma incidence at 50 mg/kg/day (11.4%)
was not statistically significant but was slightly higher than the upper end of the testing
laboratory's historical control range (0%—8.3%). For reference. the incidences of interstitial cell
hyperplasia were 7/70 (10%), 7/70 (10%), 3/70 (4 3"n), and 15 7<) (21 4%) at 0 mg/kg/day,
0.1 mg/kg/day, 1 mg/kg/day, and 50 mg/kg/day. respectively. The increased incidence of
hyperplasia at the high dose was not statistical I \ significant (compared lo control), although the
incidence of hyperplasia at 50 mg/kg/day exceeded the historical control range (0%—8.3%). The
observed incidences in the control and low-dose groups (both 10%) were also slightly above the
upper end of historical controls. DuPont's reanalysis of these findings in the testes indicated that
the number of male rats that died before I \ ear was 4. '¦>. S. and 3 in the 0 mg/kg/day (control),
0.1 mg/kg/day, 1 mg/kg/day, and 5<> mg nig day groups, respecli \ ely. The causes of death were
generally dosing injury or undeternii ned causes, and there \\ ere no testicular lesions or tumors in
the testicular tissues of these animals Including these early deaths, the incidences of testicular
interstitial cell hyperplasia were 7 66 (in (•>"<>). 7 (•>I (II 5"..), 3 (->2 (4.8%), and 15/67 (22.4%) in
the 0 mg/kg/day (control). n I nig kg day. I nig kg day. and 5<> nig kg/d groups, respectively.
The corresponding incidences of testicular interstitial cell adenomas were 4/66 (6.0%), 4/61
(6.6%>), 1/62 (1.6%), and X (->7 (I I l>"n). Thus, there were no statistically significant differences
for either hyperplasia or adenoma, consistent with results from the original report in which all
early deaths were included
Based upon the EP.Vs re\ iew of the study, the increased incidence of liver tumors in females at
500 nig,kg day and combined pancreatic acinar adenomas and carcinomas in males at
50 mg/kg day are treatment-related. The increased incidence of testicular interstitial cell
adenoma was not statistically significant, and the EPA accepted the results of the reanalysis that
excluded the early deaths
4.5 Reproduct amenta! Toxicity Studies
DuPout-18405-103' (2010)
In a combined oral gavage reproductive/developmental toxicity study in mice with HFPO dimer
acid ammonium salt, the test compound (purity 84%) was administered by oral gavage (vehicle
was deionized water) to Crl:CDl(ICR) mice (25/sex/group) at doses of 0, 0.1, 0.5, or 5
mg/kg/day, according to a modified OECD TG421 (DuPont-18405-1037, 2010; OECD, 2016).
The male mice were approximately 6 weeks old and the female mice were approximately 10
weeks old. Parental Fo males were dosed 70 days prior to mating and throughout mating through
1 day prior to scheduled termination, for a total of 84 to 85 total doses. Parental Fo females were
dosed for 2 weeks prior to pairing and were dosed through LD 20 for a total of 53 to 65 doses
(exceptions include females with no evidence of mating or those that failed to deliver yet were
37
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
administered a total of 37 to 50 doses). Fi animals (offspring) were dosed daily beginning on
PND 21 through PND 40.
In this study, increases in BWs and food consumption were observed at 5 mg/kg/day in Fo
animals. In Fo males, increased mean BW gains were reported in the 5 mg/kg/day group during
study days 0-49; differences from the control group achieved significance during study days 0-
7, 14-21, and 21-28. Significantly higher mean BW gains were observed in this high-dose male
group when the overall premating period (study days 0-69) and treatment period (study days 0-
84) were evaluated. Mean BW gains were statistically significantly increased in females during
both the premating period and throughout gestation at 0.5 and 5 mg/kg/day. At the high dose,
mean BW gains were increased (5.1 %— 14.0%) compared to controls throughout lactation; the
differences were significant on LDs 1, 4, and 21. BWs were una fleeted at 0.1 and 0.5 mg/kg/day
during lactation. Overall, final BW was significantly increased lYom control by 9% and 14% in
males and females administered 5 mg/kg/day, respecti\ ely
An increase in relative kidney weight compared lo control by 6.5% was observed only in Fo
females at the 5-mg/kg/day dose. Mild increases i n tubular cell hypertrophy u ere observed in the
kidneys of males at greater than or equal to 0 5 mg kg/day (6 24 mice or 25" <> and 18/24 mice or
75% of male mice at 0.5 mg/kg/day and 5 mg/kg day. rcspecti\ ely, compared lo 1/25 mice or 4%
in the control). Chronic progressive nephropathy was also noted in males at 0 5 mg/kg/day (4/24
mice or 17%) and 5 mg/kg/day (5/24 mice or 21%). This effect was not associated with any
evidence of tubular cell degeneration.
Liver effects also were reported in both males and females in this study. In males, mean absolute
liver weights were increased 2(->"<> and 142".. at <> 5 mu kg day and 5 mg/kg/day, respectively, as
compared to control \ alues. Mean relative li\ er weights were increased by 26% and 121%,
respectively, at the <> 5- mg'kg'day and 5-mg kg day doses. In females, mean absolute liver
weights were increased by 2(->"<> and l<)|"n at <> 5 mg kg day and 5 mg/kg/day, respectively, as
compared to control \ allies Mean relati\ e (".. Ii\\ ) li\ er weights were increased by 17% and
80%, respecti\ ely Microscopic lindings obser\ ed in the liver of Fo males and females
administered 0.5 5 mg,kg,day included increases in hepatocellular hypertrophy, single-cell
necrosis, mitotic figures, and lipofuscin pigment. Fo females exhibited an increase in the
incidence of gross white areas in the li \ er at 5 mg/kg/day, which correlated with microscopic
focal and single-cell necrosis At doses greater than or equal to 0.5 mg/kg/day, minimal-to-
moderate hepatocellular hypertrophy was observed in both sexes, along with the corresponding
increases in relative li\ er weight outlined above. Specifically, male mice exhibited a 50% and
100%) increase in the incidence of hepatocellular hypertrophy compared to control at 0.5
mg/kg/day and 5 mg/kg day. respectively, and similar increases in incidence was also observed
in female mice (58% and 100% at 0.5 mg/kg/day and 5 mg/kg/day, respectively, compared to
control). At greater than or equal to 0.5 mg/kg/day, single-cell necrosis of hepatocytes was
observed in males. Specifically, single-cell necrosis was observed in 5/24 mice at 0.5 mg/kg/day
and 24/24 mice at 5 mg/kg/day compared to 1/25 mice in the control. Female mice exhibited an
increase compared to control in both focal/multifocal necrosis and single-cell necrosis at 5
mg/kg/day. Specifically, 5/24 mice had focal/multifocal necrosis compared to 1/24 in the control
and 21/24 mice had single-cell necrosis compared to 1/24 mice in the control. Finally, the
incidence of mitotic figures increased in males and females administered 5 mg/kg/day by 75%
38
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
and 21% compared to control, respectively, while the incidence of lipofuscin pigment increased
by 88% and 21% compared to control, respectively.
No treatment-related effects were identified for reproductive parameters (mating, fertility, and
copulation indices; mean days between pairing and coitus), although male epidydimal weight
relative to final BW was statistically decreased at 5 mg/kg/day in both the left and right testes
(12%) decrease relative to control). No treatment-related effects were observed for mean
gestation length, mean numbers of implantation sites, mean numbers of pups born, live litter size,
percentage of males at birth, postnatal survival, or general condition of pups. At 5 mg/kg/day,
however, male and female Fi pups exhibited lower mean BWs at PNDs 4, 7, 14, 21, and 28.
Male Fi pups continued to exhibit lower mean BWs at PNDs 35 and 40. Although values for the
attainment of balanopreputial separation and vaginal patency were within the range of historical
control values, the pups showed statistically significant delays in these endpoints at 5 mg/kg/day
(a finding that may be related to the observed effects on li\\ during the preweaning period).
Additionally, the day for attainment of vaginal patency did not exhibit a dose-response. The
NOAEL (Fo) is 0.1 mg/kg/day, and the LOAI-I. is n 5 mg/kg/day based on liver effects (single-
cell necrosis in males). The NOAEL (Fi) is 0 5 mu kg/day based on decreased pup BW and
delays in attainment of balanopreputial separation and \ aginal patency at the high dose.
DuPont-18405-841 (2010)
In a prenatal and developmental toxicity study in 12-week old female Crl:CD(SD) rats, HFPO
dimer acid ammonium salt (purity S4"..) was administered \ ia oral gavage (vehicle was
deionized water) once daily from (il) (¦> through (il) 2<~) at doses of<~). 10, 100, and 1,000
mg/kg/day (22 females group), according to OF.('I) "IXi 414 (I)uPont-18405-841, 2010; OECD,
2001b). The parental males and females were not dosed prior to or during mating and dosing for
the dams was not initialed until GD 6. Lack of dosing for males and females prior to and during
mating and failure lo dose the dams during the (il) n to GD 6 period are limitations when
evaluating this study to fully reflect the ability of the HFPO dimer acid ammonium salt to cause
reproducti\e de\ elopmental toxicity
The dams" liYV decreased at all doses, but significantly decreased (22% compared to control) at
1,000 mu kg day. Moreo\ er. gravid uterine weight was significantly decreased by 10% and 25%
compared to control at I'm mg kg'day and 1,000 mg/kg/day, respectively. Food consumption in
the dams was significantly decreased by 9% over the dosing period (GD 6-GD 21) at the highest
dose. Early deli\ cry on GD 21 was observed in 18% and 41% of the dams at 100 mg/kg/day and
1,000 mg/kg/day, respect i\ el y Importantly, the authors noted in the available historical controls
data for early deliveries in this rat strain (17 datasets), no females showed early deliveries (i.e.,
before GD 21).
Statistically significant increases relative to control in absolute liver weight (12% and 34%) were
observed at 100 mg/kg/day and 1,000 mg/kg/day, respectively. Changes in liver weight relative
to BW were not documented. This increase in liver weight was associated with hepatocellular
hypertrophy at the high dose (19/22 rats, or 86%) and focal necrosis was observed in 9% and
23% of the dams dosed with 100 mg/kg/day and 1,000 mg/kg/day, respectively. Additionally,
absolute kidney weight increased dose-dependently in the dams and was significantly increased
compared to control (10%) at the highest dose. Changes in kidney weight relative to BW were
not documented, and there were no notable microscopic changes in the kidney tissues for the
dams. Of note is that a 1,000 mg/kg/day dam that died on GD 20 had moderate multifocal/focal
39
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
necrosis of the liver and disseminated intravascular coagulation in the kidney glomerular
capillaries.
The pups experienced a 9% and 28% decrease compared to control in fetal weight at doses of
100 mg/kg/day and 1,000 mg/kg/day, respectively. The percentage of male (47%) and female
(53%>) pups born were significantly altered from control (55% male; 45% female) at 1,000
mg/kg/day. Additionally, a 14th rudimentary rib developed in 9% of the control fetuses, 10% of
fetuses in the 10-mg/kg/day dose group, 12% of fetuses in the 100-mg/kg/day dose group, and
27% of the fetuses in the 1,000-mg/kg/day dose group. Statistical analyses were not completed
for the development of the 14th rudimentary rib in individual pups, but a statistically significant
increase in the number of litters developing a 14th rudimentary ri h was observed for those
receiving the high dose.
The NOAEL for this prenatal and developmental toxicity stuck is I n mg/kg/day based on an
increase in early deliveries, decreases in gravid uterine weight, and decreased fetal weights for
both sexes, all occurring at the LOAEL of 100 mg kg/day.
4.6 Other Studies
4.6.1 Immunotoxicity Studies
Rushing et al., 2017
Male and female C57BL/6 mice ((¦> 12 sex group) were administered HFPO dimer acid by
gavage at doses of 0, 1, 10, or 100 mg/kg day for 28 days (Rushing et al., 2017). The animals
were immunized with sheep RIJC antigen on day 24 and. 5 days later, were evaluated for TDARs
and splenic lymphocyte suhpopulations Organs were collected 1 day after the final gavage
exposure.
T lymphocyte numbers were significantly increased (the average increase of CD8+, CD4+/CD8+,
and CD4VCD8" T cells were 74"..) in males at l<>() mg kg/day, yet suppression of TDAR was
obsened in female mice only at l<>() mg kg day TDAR suppression was measured through IgM
antibody production, which decreased In 7 .V\, in females at the high dose. Liver weight relative
to BAV signillcantly increased (4<)"() l(-><)'\>) in both sexes at 10 mg/kg/day in a dose-dependent
manner. Relati\ e spleen weights significantly decreased by 11% in females treated with 100
mg/kg/day, and there were no significant changes in thymus weight.
Peroxisomal fatty acid oxidation was measured using hepatic acyl-CoA oxidase activity as a
readout. In male mice, hepatic acyl-CoA oxidase activity increased 122%) and 222% at 10
mg/kg/day and 100 mg kg day. respectively. Female mice had a 100% increase in acyl-CoA
oxidase activity at the highest dose tested. The NOAEL for immune effects that include TDAR
suppression in females and increased T cells in males is 10 mg/kg/day.
4.6.2 Mechanistic Studies
Overall, there are limited data providing mechanistic insight into the effects of HFPO dimer acid
and/or its ammonium salt.
Wang et al., 2016
In one study investigating changes in gene expression, male ICR mice (n=12/group) were
administered 1 mg/kg/day HFPO dimer acid ammonium salt via oral gavage for 28 days (Wang
40
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
et al., 2016). Although the authors state that HFPO dimer acid was tested and its chemical
structure is presented, the CASRN is listed as 62037-80-3, which is the HFPO dimer acid
ammonium salt. At termination, blood was collected and liver weights were determined. Liver
samples were processed for histopathological analysis or frozen for high-throughput RNA-
sequencing analysis (3/group).
Statistically significant treatment-related findings reported include increased absolute liver
weight (31%) and relative liver weight (28%), ALP (150%), low-density lipoprotein cholesterol
(50%>), decreased total bilirubin (37%), and decreased direct bilirubin (45%) compared to
control. Qualitative histopathological findings were also reported in the liver and included lipid
droplet accumulation, hepatocellular hypertrophy, mild steatosis, and karyolysis. To potentially
gain mechanistic insight into the causes of these liver effects, high-throughput RNA-sequencing
was conducted and 146 hepatic transcripts (101 upreuulated and 45 downregulated) were
statistically significantly changed from control following treatment with HFPO dimer acid
ammonium salt. These changes in hepatic transcripts indicated differential gene expression of
four cell signaling pathways associated with lipid metabolism: the PIWRu signaling pathway,
and the pathways for retinol metabolism and fully acid degradation, as well as the pathway for
the catabolism of the polyunsaturated fatty acid araehidonic acid
Sheng et al., 2018
Sheng et al. (2018) used in vitro experiments to investigate periluoroalkyl cytotoxicity and
binding to proteins. The study tested multiple perfluoroalkyl substances, including the HFPO
dimer acid ammonium salt (CASN 62037-SO-3). The study authors state that they used the
HFPO dimer acid; howe\ er. the CASN listed in the study is for the ammonium salt. The study
authors used a cell \ iability assay in a human liver cell line (HL-7702) to determine the
cytotoxicity of the various perfluoralkyl substances and used flow cytometry to investigate
effects on cell proliferation. The authors noted, however, that the effects of HFPO dimer acid
ammonium salt on cytotoxicity and cell proliferation were undeterminable through these assays
because of the chemical's low boiling point and high volatility. Therefore, the only data on
HFPO dimer acid ammonium salt generated In this study pertain to its ability to bind to human
liver fatty aeid-binding protein (hi ,-l 'ABP). This binding affinity was explored because previous
PFAS ha\ e exhibited effecti\ e binding to hL-FABP and this binding might explain how PFAS
can enter into hepatocytes- a target cell for the HFPO dimer acid and/or its ammonium salt.
Ultimately, this study found that IIFPO dimer acid ammonium salt exhibited a weaker binding
affinity and bound differently to hL-FABP than PFOA and perfluorooctane sulfonate (PFOS)
(Sheng et al., 2018) These results were replicated using a predictive model of binding affinity to
hL-FABP (Cheng and \ u. 2' > IX)
4,6,3 Genotoxicity Studies
HFPO dimer acid ammonium salt was not observed to induce genetic mutations both with and
without metabolic activation of the test substance by rat liver S9 fraction in two species of
prokaryotes: Escherichia coli (strain WP2uvrA) and Salmonella typhimurium (strains TA98,
TA100, TA1535, and TA 1537) (DuPont-19713 RV1, 2008; DuPont-22734 RV1, 2008). An in
vitro gene mutation test of the HFPO dimer acid ammonium salt in mouse lymphoma cells
(strain L5178Y/TK+/-) was negative in the presence and absence of rat liver S9 fraction
(DuPont-26129, 2008). HFPO dimer acid ammonium salt was observed to induce chromosomal
aberrations in Chinese hamster ovary cells in vitro in the presence and absence of S9 activation
41
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
(DuPont-19714 RV1, 2008; DuPont-22620 RV1, 2009). In in vivo mammalian studies, exposure
to HFPO dimer acid ammonium salt by the oral route did not induce chromosomal mutations in
the form of structural aberrations, numerical aberrations, or micronuclei nor DNA effects in the
form of unscheduled DNA synthesis (DuPont-23219, 2007; DuPont-23220, 2007). A table
summarizing the findings of the available genotoxicity studies is provided in appendix D.
4.7 Mode of Action
The MOA of HFPO dimer acid and/or ammonium salt toxicity is not clearly understood.
Additionally, the mode of carcinogenic action of HFPO dimer acid and/or ammonium salt is not
clearly understood. For some PFAS (e.g., PFOA), PPARa agonism has been proposed as a
potential MOA for the liver tumors (Klaunig et al., 2003, 2012; Maloney and Waxman, 1999). In
this MOA, binding of PFOA to the PPARa receptor results in increased peroxisome proliferation
and cell replication. PPARa is primarily expressed in the liver, but also is present in the kidney,
intestines, heart, and brown adipose tissue (Hall el al., 2012). There are four key events in the
PPARa-agonist MOA for liver tumors (Klaunig el al , 2003, 2012). The first key event is
activation of PPARa. Increased palmitoyl-Co.\ oxidase activity is used in many studies as a
biomarker for PPARa activation. Other associated indicators are hepatocellular hypertrophy and
increased liver weight. These indicators alone, ho\\e\ er. are not sufficient to establish a PPARa
MOA because they also are caused by chemicals that ha\ e no inlluence on PPARa. Additional
key events outlined by Klaunig et al (2<)i)3. 2012) include cell proliferation and decreased
apoptosis, development of preneoplastic foci, and their clonal expansion.
For HFPO dimer acid and/or ammonium salt, there are some data that demonstrate activation of
peroxisome proliferation. Activation of peroxisome proliferation was demonstrated in multiple
28-day studies (DuPont-24447, 2008; DuPont-24451). 2<)i)K; Rushing et al., 2017; Wang et al.,
2016). Using acyl-CoA oxidase activity as a measure, Rushing et al. (2017), showed increased
activity compared to control in male C57lil. (•> mice administered 10 mg/kg/day and 100
mg/kg/day of HFPO dimer acid (122% and 222".., respectively) and a 100% increase compared
to control in C57BL/6 female mice at 100 mg, kg day. There were no significant increases in
acyl-Co.\ oxidase activity at 1 mg/kg/day. Rushing et al. (2017) concluded that the HFPO dimer
acid appears to be less potent than PFOA in inducing hepatic peroxisome proliferation. The
DuPont studies used P-oxidation activity and total cytochrome P-450 content as markers of
peroxisome proliferation. In Crl ('l)-l male mice, P-oxidation activity significantly increased
compared to control at doses of n I mg/kg/day, 3 mg/kg/day, and 30 mg/kg/day HFPO dimer
acid ammonium salt In 57" n. 744".), and 648%, respectively, and total cytochrome P-450 content
significantly decreased at 3 mu kg day and 30 mg/kg/day by 26% and 53%, respectively
(DuPont-24459, 20<>S) [>-oxidation activity significantly increased compared to control in female
Crl:CD-l mice at 3 mg/kg/day and 30 mg/kg/day by 495%) and 823%), respectively, with no
alterations in total cytochrome P-450 content (DuPont-24459, 2008). In male Crl:CD(SD) rats,
P-oxidation activity was significantly increased relative to control at dosages of 0.3 mg/kg/day,
3 mg/kg/day, and 30 mg/kg/day by 42%>, 274%), and 772%>, respectively, and total cytochrome P-
450 content was significantly increased by 23%> at 30 mg/kg/day (DuPont-24447, 2008). In
female rats dosed with 30 mg/kg/day and 300 mg/kg/day, P-oxidation activity was significantly
increased compared to control to 49%> and 198%>, respectively, while total cytochrome P-450
content remained unaltered (DuPont-24447, 2008). Finally, Wang et al. (2016) demonstrates
significant increases in hepatic mRNA levels of many PPAR targets (e.g., CD36 antigen, acyl-
42
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
CoA oxidase 1, cytochrome P450 family members) after administration of 1 mg/kg/day HFPO
dimer acid ammonium salt for 28 days.
Although findings consistent with PPARa agonists were observed (e.g., increases in liver weight,
hepatocellular hypertrophy, and increased P-oxidation activity), data gaps exist for key events,
like apoptosis. Other indicators such as steatosis were not assessed in any of the
DuPont/Chemours studies. Wang et al. (2016) is the only publicly available study to qualitatively
mention observing steatosis in mouse liver samples, but does not provide quantitative
measurements. Additionally, liver necrosis was consistently observed in rodent toxicity studies
with HFPO dimer acid ammonium salt, which suggests that cytotoxicity is a possible MOA for
the observed liver tumors. Overall, the findings are not adequate to definitively conclude that a
PPARa MOA is operative for HFPO dimer acid and/or ammonium salt. Additionally, no data
support identification of a potential MOA for the pancreatic and testicular tumors as being
related to PPARa or any of the proposed alternate e MO As for the tumor development in either
organ.
5.0 Summary of Hazard
5.1 Hepatic
The liver is a target organ for toxicity from oral exposure to I II PO dimer acid and its ammonium
salt. Liver effects are observed in both male and female mice and rats at varying durations of
exposures and doses of GenX chemicals I ,i\ er effects are also the endpoints that are observed at
the lowest doses for these chemicals. I lepatocclliilar hypertrophy and an increased liver-to-BW
ratio are common find mills in rodents, hut are considered iioiukK erse and less relevant to humans
when there is evidence for PPARa acti\ ation. The increased relati\ e liver weight and
hepatocellular hypertrophy are only considered ad\ erse when they are accompanied by effects
such as necrosis, fibrosis, innammation, steatosis, and significantly increased serum levels for
enzymes indicative of li\er tissue damage (I lall et al , 2012).
Signillcant increases in li\ er weight relative to IJW were observed in male and female
Crl:CI)(SD) rats and se\ eral strains of male and female mice treated with 0.5 mg/kg/day-1,000
mg/kg day of HFPO dimer acid ammoniiiin salt for up to 28 days or up to 90 days (DuPont-
17751-1026, 2009; DuPont-18405-1037, 2010; DuPont-18405-1307, 2010; DuPont-24447,
2008; DuPont-24459, 2008; Rushing et al., 2017; Wang et al., 2016). These increases were
observed in doses as low as 0.5 mg/kg/day in male Crl:CD-l mice (26% increase over 84-85
days) (DuPont-184D5-11).i7. 2<)|0). and the greatest increases were observed when male (163%)
and female (102.7%.) Crl ('l)-l mice were administered 30 mg/kg/day for 28 days. Likewise,
male Crl:CD(SD) rats exhibited increased relative liver weights of 19%>-61%> compared to
control when administered 3 mg/kg/day-100 mg/kg/day for 28-90 days, while female rats'
relative liver weights compared to control did not increase until much higher doses (12%> at 300
mg/kg/day for 28 days and 85%> at 1,000 mg/kg/day for 90 days). Comparatively, the one
available chronic study in rats indicates that liver weight may increase and return to control
levels after a time. For example, relative liver weights in male rats increased only 15%> when
administered 50 mg/kg/day for 1 year and did not exhibit a significant increase from control at 2
years. Likewise, female rat relative liver weights increased 67%> and 42%> after administration of
500 mg/kg/day for 1 and 2 years, respectively (DuPont-18405-1238, 2013).
43
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Indications of liver damage were also reflected through increases in serum liver enzymes of
Crl:CD-l mice, particularly males, and Crl:CD(SD) rats administered HFPO dimer acid
ammonium salt. For example, significant increases in ALT (420%-l,254%), AST (106%-
478%), ALP (1,134%-1,221%), and SDH (1,134%—1,221%) were observed in male mice
administered the ammonium salt at 5-30 mg/kg/day for 28-90 days. Female mice saw smaller
increases in ALP (140%—143%) and SDH (32%—186%) as compared to male mice administered
the same dose. Overall, rats exhibited far fewer and smaller increases in serum liver enzyme
levels following subchronic exposure compared to the mouse, with increases in AST (106%) and
ALP (52%) at 100 mg/kg/day in male rats and AST (66%) in female rats at 1,000 mg/kg/day. In
the chronic study, however, ALT (228%), ALP (180%). and SDN (140%) significantly increased
in male rats only when administered 50 mg/kg/day for I \ ear
Finally, liver damage was confirmed microscopically in mule and female mice and rats in several
less than chronic studies (15-90 day) and one chronic study (DuPont-l 7751-1026, 2009;
DuPont-18405-841, 2010; DuPont-18405-1037. DuPon1-IS4<)5-l238, 2013; DuPont-
18405-1307, 2010; DuPont-24447, 2008; Dul\>nt-24459, 2008; Wanu el al . 2016). The most
prevalent liver effects following both subchronic and chronic exposure lo I II PO dimer acid
and/or ammonium salt were hepatocellular hypertrophy and single-cell and or focal necrosis. In
both sexes of mice exposed subchronically, hepatocellular hypertrophy was observed at 0.5
mg/kg/day, while male and female rats showed these effects at 3 mg/kg/day and 30 mg/kg/day,
respectively. Interestingly, in the chronic study, male rats did not show any significant increases
in hepatocellular hypertrophy when administered 1-50 mu ku day of HFPO dimer acid
ammonium salt for 1 year, and only 10% of the rats exhibited minimal hypertrophy with
50 mg/kg/day for 2 years (l)ul,ont-18405-1238, 2<)|3) Conversely, female rats had significant
hepatocellular hypertrophy at 5<)i) mg/kg/day after I year (I no",,) and 2 years (93%).
Hepatocellular necrosis was detected in nearly all the available studies and at the lowest doses
tested. In the oral/reproductive subchronic study, male and female mice presented single-cell
necrosis in doses as low as n 5 mu ku day (21°.. and 8%, respectively), which significantly
increased at 5 mg/ku day (|oo% and SS%. respecti\ely). Male and female rats exhibited
subchronic hepatocellular necrosis at much higher doses, with males showing general necrosis
(30%>) at 30 mg/kg/day and females presenting focal liver necrosis at 100 mg/kg/day (9%) and
1,000 mg/kg/day (23%). This finding indicates that mice are more sensitive to liver necrosis than
rats in subchronic exposure scenarios. In the 2-year chronic rat study, centrilobular necrosis
increased at 50 mg/kg/day and 500 mg/kg/day for males (7%) and females (4%), respectively,
while single-cell necrosis was observed only in females (4%) at 500 mg/kg/day.
5.2 Hematoiogica
The hematologic system could be a target of HFPO dimer acid ammonium salt toxicity as effects
have been observed across studies of varying durations of oral exposure to HFPO dimer acid
ammonium salt. The primary effects observed are decreases in RBC number, hemoglobin, and
percentage of RBCs in the blood, indicating that oral exposure to HFPO dimer acid ammonium
salt might promote anemic conditions. In male mice and rats, the percent change in these effects
from the controls was relatively small. For example, male Crl:CD-l mice and Crl:CD(SD) rats
treated with 3 mg/kg/day-100 mg/kg/day of HFPO dimer acid ammonium salt for 28-180 days
had maximum decreases of 12%, 11%, and 12% in hemoglobin, erythrocyte count, and
hematocrit, respectively (DuPont-17751-1026, 2009; DuPont-18405-1238, 2013; DuPont-18405-
1307, 2010; DuPont-24447, 2008; DuPont-24459, 2008). Interestingly, in the available chronic
44
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
study, no hematological effects were observed at the 12-month time point in male rats (DuPont-
18405-1238, 2013). Female Crl:CD-l mice and Crl:CD(SD) rats presented hematological effects
at greater than 90 days and typically at higher doses than males, with one exception. Hemoglobin
alone significantly decreased by 4% when female Crl:CD(SD) rats were administered 1
mg/kg/day HFPO dimer acid ammonium salt for 90 days (DuPont-18405-1238, 2013).
Otherwise, hematological effects occurred at doses greater than or equal to 50 mg/kg/day and the
maximum decreases from control were 24%, 28%, and 20% for hemoglobin, erythrocyte count,
and hematocrit, respectively (DuPont-18405-1238, 2013; DuPont-24447, 2008).
5.3 Renal
The kidney could also be a target organ for toxicity from oral exposure to HFPO dimer acid
and/or ammonium salt; however, kidney effects typical I \ picscnlcd at higher doses than the liver
effects.
Significant increases in kidney weight relative lo li\\ were ohscr\ ed in several less than chronic
studies in Crl:CD-l mice and Crl:CD(SD) rats H ealed with 0.1 mg/kg/day-1 .<">00 mg/kg/day
(DuPont-17751-1026, 2009; DuPont-18405-l<)37. 2on mu kg/day o\ er l><) days (DuPont-17751-1026,
2009). Interestingly, increases in relati\ e kidney weights were not observed in the same type of
male rat when administered HFPO dimer acid ammonium salt lor I or 2 years (DuPont-18405-
1238, 2013). Relati\ e kidney weight did increase in female Crl ('D(SD) rats by 25% and 14%
when administered 5<)i) mu ku day I II PO dimer acid ammonium salt for 1 and 2 years,
respectively (DuPont-1S4<)5-l23S. 2<)|3)
These increases in kidney weight were often associated with increases in BUN, which can be
used as an indicator of renal damage In se\ eral studies, urea nitrogen levels were significantly
increased (lo".. 3S'\.) in male mice and rats administered doses greater than or equal to 30
mg/ku day of III PO dimer acid ammonium salt for 2S 180 days (DuPont-17751-1026, 2009;
DuPont-1 X4< 15-1238, 2013. l)uPonl-24447, 2008, DuPont-24459, 2008). Female rats exhibited
an increase in urea nitrogen le\ els (35%) only when administered 500 mg/kg/day of HFPO dimer
acid ammonium salt for 1 year (DuPont-18405-1238, 2013). Kidney damage was equivocal
microscopically in the less than chronic studies (28-90 days), and typically presented as
increases in basophilic tubulin' cells and tubular epithelial hypertrophy or dilation without tubular
degeneration and/or necrosis (I)uPont-17751-1026, 2009; DuPont-18405-1037, 2010; DuPont-
24459, 2008; DuPont-24447, 2008).
In the chronic study, the increases in BUN and relative kidney weight noted above for female
rats were associated with multiple microscopic observations of kidney damage when female rats
were treated with HFPO dimer acid ammonium salt for 2 years. For example, at 50 mg/kg/day-
500 mg/kg/day, female rats exhibited transitional cell hyperplasia, tubular dilation, pelvic and
tubular mineralization, and papillary edema, which ultimately resulted in papillary necrosis at
500 mg/kg/day (DuPont-18405-1238, 2013).
45
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
To summarize, significant and dose-dependent increases in relative kidney weight occurred in
rats at lower doses (e.g., 10 mg/kg/day) in a subchronic study (DuPont-18405-1307, 2010).
Kidney hypertrophy, however, was not associated with microscopic damage of the kidney such
as necrosis in this study. Additionally, there are instances in which kidney hypertrophy occurred
at low doses in female mice (e.g., 0.1 mg/kg/day (DuPont-24459, 2008) or 5 mg/kg/day
(DuPont-18405-1037, 2010)), but there was not a dose-response in these datasets, and
microscopic damage to the kidney tissues was not reported. Of the available studies, kidney
hypertrophy was associated with significant microscopic damage only in female rats treated with
500 mg/kg/day of HFPO dimer acid ammonium salt for 2 years (DuPont-18405-1238, 2013).
Thus, the observed kidney effects are potentially of concern The biological significance,
however, of the observed hypertrophy and increases in lil \ without microscopic evidence of
kidney damage is not clear.
5.4 Developmental/Reproductive
Evidence suggests HFPO dimer acid and/or ammonium salt could target the reproductive system
and the developing fetus/child. Currently, there are two available studies in\ estigating the
developmental and reproductive effects of HI PO dimer acid and/or ammonium salt (DuPont-
18405-841, 2010; DuPont-18405-1037, 2010). In both studies, there was a decrease in rodent
pup weight that ranged from 9%-24°.. u hen the pups uere exposed to 5 mg/kg/day-1,000
mg/kg/day in utero. The mouse pups showed delays in attaining balanopreputial separation and
vaginal patency at 5 mg/kg/day of 2 (•> days and 3 4 days, respectively (a finding that might be
related to the observed effects on B\V during the preweaning period) (DuPont-18405-1037,
2010). Additionally, the attainment of vaginal patency did not exhibit a dose-response
relationship. The decrease in pup weight also was associated with a decrease in gravid uterine
weight by 10% and 25".. at lOUmu kg'day and 1 .<>(>() nig, kg, day, respectively, in the rat prenatal
developmental toxicity study (DuPont-18405-841. 2<»10). In the mouse study, Fo males exhibited
a decrease of 12% in the relati\ e epididymis weight (DuPont-18405-1037, 2010). Overall, no
treatment-related effects were identified for many reproductive parameters in the mouse study
(mating, fertility, and copulation indices, mean days between pairing and coitus; mean gestation
length: mean numbers of implantation sites; mean numbers of pups born; live litter size;
percentage of males at birth, postnatal survival; or the general condition of pups) (DuPont-
18405-1037. 2<)|0). In the rat de\ elopmental toxicity study, however, early delivery on GD 21
was observed in 18% and 41".. of the dams at 100 mg/kg/day and 1,000 mg/kg/day, respectively,
and the percentage of male (47°..) and female (53%) pups born were significantly altered from
control at 1,000 mg kg day (DuPont-18405-841, 2010). Moreover, in this study, a 14th
rudimentary rib de\ eloped in l->".. of the control fetuses, 10% of fetuses in the 10-mg/kg/day
dose, 12% of fetuses in the I <)()-mg/kg/day dose, and 27% of the fetuses in the 1,000-mg/kg/day
dose (DuPont-18405-841, 2010). Statistical analyses were not completed on the development of
the 14th rudimentary rib in individual fetuses, but a statistically significant increase in the
number of litters developing a 14th rudimentary rib was observed at the high dose.
5.5 Immune System
In the one available study specifically addressing immunotoxicity, suppression of TDARs was
measured through IgM antibody production in mice (Rushing et al., 2017). IgM antibody
production was decreased by 7.3% in female C57BL/6 mice treated with 100 mg/kg/day of
46
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
HFPO dimer acid. In male mice treated with the same dose of HFPO dimer acid, significant
increases in the number of T lymphocytes were observed, but no suppression of TDARs.
In two studies of less than chronic duration (28-90 days), decreases in spleen weight relative to
BW were observed in female mice and rats (DuPont-18405-1307, 2010; Rushing et al., 2017).
For example, in C57BL/6 mice, relative spleen weights significantly decreased by 11% in
females treated with 100 mg/kg/day of HFPO dimer acid for 28 days (Rushing et al., 2017).
Changes in early markers of potential immunotoxic effects were observed in multiple studies
examining the oral toxicity of HFPO dimer acid and/or ammonium salt. The most prevalent
indications were statistically significant decreases from control in serum globulin levels (6%-
22%), which resulted in an increase in the serum A/G ratio (7%-58%) from the controls when
both sexes of Crl:CD-l mice and Crl:CD(SD) rats were iron led with 1 mg/kg/day-500
mg/kg/day of HFPO dimer acid ammonium salt for 12 months or less (DuPont-17751-1026,
2009; DuPont-18405-1238, 2013; DuPont-18405-1307, 2010; DiiPoni-24447, 2008; DuPont-
24459, 2008). Alterations in the serum levels of globulin can be associated with decreases in
antibody production (USFDA, 2002). To delemiine the biological significance of the apparent
decrease in globulin production, however, immune function tests (such as TDAR) need to be
conducted. Finally, female Crl:CD-l mice exhibited a 21°<> and 18% decrease in spleen weight
relative to BW when administered 0 5 mu kg/dav and 5 mu kg day of HFPO dimer acid
ammonium salt for 90 days, respecti\ely (l)ul,oni-IK4<)5-l3<)7. 2010). For HFPO dimer acid
and/or ammonium salt, there were also two local lymph node assays (LLNAs) conducted in mice
that showed equivocal results (DuPont-1lMl)7. 2<)i)(-.. DuPont-22M6 RV1, 2007).
In summation, the results of the Rushing et al. (2d I 7) TDAR assay in combination with the
supportive findings of decreased globulin le\ els and spleen weight provide some evidence that
GenX chemicals can induce immune suppression in female mice.
5f ¦»"*
.b
The single cancer hioassay lor I II PO dimer acid ammonium salt showed increased incidence of
liver minors (females) and combined pancreatic acinar adenomas and carcinomas (males) in rats
at the high doses only. Additionally, a statistically insignificant increase in the incidence of
testicular interstitial cell adenoma was noted at the high dose. Given uncertainties, the existing
evidence from this single chronic study is considered inadequate to justify a quantitative
assessment. Further, the available data for HFPO dimer acid ammonium salt suggest that mice
might be more sensiti\ e to exposure to GenX chemicals than rats. The available study (DuPont-
18405-1238 2013) only e\ aluated rats; there are no studies measuring cancer endpoints in mice.
Given the evidence that the li \ er is the target organ for toxicity and primary organ for tumor
development, the lack of data evaluating cancer in mice is a database deficiency. Thus, under the
EPA's Guidelines for Carcinogen Risk Assessment (USEPA, 2005), there is Suggestive Evidence
of Carcinogenic Potential of oral exposure to GenX chemicals in humans, based on the female
hepatocellular adenomas and hepatocellular carcinomas and male combined pancreatic acinar
adenomas and carcinomas. No data are available to evaluate cancer risk via dermal or inhalation
exposure.
47
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
6.0 Dose Response Assessment
6.1 Study and Endpoint Selection
Multiple animal studies were available for consideration for the development of the subchronic
and chronic RfDs for HFPO dimer acid and its ammonium salt. These studies included short-
term, subchronic, and chronic exposures, including developmental and reproductive toxicity
studies. These studies and their NOAELs and/or LOAELs are presented in Table 7.
48
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
NOVEMBER 2018
Table 7. Summary of Study NOAELS/LOAELS
Study
Overall Study
Quality (See
Appendix B)
Doses
(mg/kg/day)
NOAEL or LOAEL
(mg/kg/day)
Effects at the LOAEL
28-Day Oral (Gavage)
Toxicity Study in Rats
(OECD TG 407)
DuPont-24447 (2008)
High (> 1 and < 1.7)
Males: 0, 0.3, 3,
and 30
Females: 0, 3, 30,
and 300
NOAEL"
LOAEl.
ii ;
1 lenialnlogical effects (J, RBC count, hemoglobin, and hematocrit in
males)
1 iiiiiiiiiic elleels (J, globulin and t A/G ratio in males)
28-Day Oral (Gavage)
Toxicity Study in Mice
(OECD TG 407)
DuPont-24459 (2008)
High (> 1 and < 1.7)
0,0.1, 3, and 30
no \i:i.
LOALI.
o.l
Liver effects (snidc-cell necrosis in males, t relative liver weight in
in males, and ' hepatocellular hypertrophy in males)
1 leinatological elleels (J, hemoglobin and hematocrit in males)
1 minune effects (J, globulin in females, and t A/G ratio in both
SCXCS1
28-Day Oral (Gavage)
Immunotoxicity Study
in Mice
High (> 1 and < 1.7)
0, 1, 10, and luu
\nle 1II lJ() dimer
acid
\o \i:i.
I.OMI.
in
lull
1 iiiiiiiiiic effects (TDAR suppression in females, and t lymphocytes
mi males)
Rushing et al. (2017)
90-Day Oral (Gavage)
Toxicity Study in Rats
(OECD TG 408)
High (> 1 and 1.";
Males (i. (i 1. Hi.
and Inn
Females u. |u.
\o \i:i.
i.o\i:i.
II1
III
Hematological effects (J, RBC count, hemoglobin, and hematocrit in
males)
DuPont-17751-1026
(2009)
luu. and
90-Day Oral (Gavage)
Toxicity Study in Mice
(OECD TG 408)
High ( 1 and " 1 7)
0. ii 1. 11.5. and 5
\< )AEL =
LOAEL=
0.5
5
Liver effects (t AST, ALT, and ALP in males; t relative liver
weight in males; and t hepatocellular hypertrophy and single-cell
necrosis in males)
DuPont-18405-1307
(2010)
49
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
NOVEMBER 2018
Study
Overall Study
Quality (See
Appendix B)
Doses
(mg/kg/day)
NOAEL or LOAEL
(mg/kg/day)
Effects at the LOAEL
Combined Chronic
Toxicity/ Oncogenicity
Study in Rats
(OECD TG 453)
DuPont-18405-1238
(2013)
High (> 1 and < 1.7)
Males: 0,0.1, 1,
and 50
Females: 0, 1, 50,
and 500
NOAEL =1
LOAEL = 5( i
Liver effects (centrilobular necrosis in both sexes; t ALP, ALT, and
SDH in males; and t centrilobular hepatocellular hypertrophy and
cystic focal degeneration in males)
Oral (Gavage)
Reproduction/
Developmental Toxicity
Study in Mice (OECD
TG 421; modified
according to the
Consent Order)
High (> 1 and < 1.7)
0,0.1,0.5, and 5
n< ) \i :i. < i > = o.i
LOALI. d ' ) o 5
\()\i:i,(]i (15
i.<) \i :i. rp ) - 5
Liver effects (snide-ccll necrosis in males, and t relative liver
weight in both sc.ncm
1 )cvelopmental effects (J, pup weights, and delays in the attainment
of halanopreputial separation and vaginal patency)
DuPont-18405-1037
(2010)
Prenatal and
Developmental Toxicity
Study in Rats (OECD
TG 414)
DuPont-18405-841
(2010)
High (> 1 and < 1 ")
U. 1 <). |()(). Mild
1.000
\()\i:i.(|- and l; )
lo
l.()\i:i. d' and 1' )
loo
1 )e\ clop me ntal effects (t early deliveries, j fetal weights in both
se\es. and j gravid uterine weight)
50
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
As summarized in section 5 and in Table 7, adverse effects, including liver toxicity, BW
changes, and hematological and immune effects were observed in the range of 0.5-1,000
mg/kg/day. These studies were evaluated further based on duration of exposure, use of a control
and two or more doses, and provision of NOAEL and/or LOAEL values.
Data from these available studies indicate that the liver is the most sensitive target of GenX
chemicals toxicity. Liver effects were observed in both male and female mice and rats at varying
durations of exposures and doses. These effects occurred at the lowest doses of exposure to
GenX chemicals. As highlighted above, immune and hematological effects were also observed at
low doses; however, these endpoints are not as consistently observed compared to liver effects.
Additionally, there is some uncertainty regarding the biological significance of both the
hematological and immune endpoints. For example, the observed changes in albumin and A/G
ratio at 3 mg/kg/day (DuPont-24447, 2008; DuPont-24451). 2<)OK) are considered early markers
of potential immunotoxic effects. Evaluation of additional immune function assays,
histopathology, and immune endpoints such as antibody levels. ho\\e\ er. are not available.
Currently little or no data on the potential for GenX chemicals to impact aspects of immune
function beyond the immunosuppression (e.g . allergic responses and autoimmunity) exist.
Furthermore, while considered adverse, the hematological effects were inconsistently observed,
especially as study duration increased. For example, the hematological effects observed in the
28-day mouse study at 3 mg/kg/day were not observed in the ^'i-day subchronic study in mice,
except for a 3% decrease in hemoglobin concentration at 5 mg kg/day. No hematological
changes were observed at the 0.1 or n 5 mg/kg/day dose in the subchronic mouse study (DuPont-
18405-1307, 2010). Likewise, the hematological effects obser\ ed in the subchronic rat study at
low doses are not obser\ ed in the chronic rat study (DuPont-1775 1-1026, 2009; DuPont-18405-
1238, 2013). Specifically: decreases in hemoglobin, hematocrit, and RBC count that are
observed at 10 mg/kg day in the subchronic study are not observed after 12 months of dosing,
which adds additional uncertainty to the significance of these effects (DuPont-18405-1238,
2013).
Therefore, the N\\ considered studies that obser\ ed adverse liver effects at the lowest dose
tested in the selection of the critical study for derivation of the RfDs. Because liver effects such
as increases in li\ er weight and hypertrophy can be associated with activation of cellular PPARa
receptors, the N\\ evaluated observed li\ er effects resulting from HFPO dimer acid ammonium
salt exposure against the Flail criteria (Hall et al., 2012). These criteria indicate that increased
liver weight and hepatocellular hypertrophy must be accompanied by histologic or clinical
pathology indicative of li\ er toxicity to be considered adverse. Histologic or clinical pathology
indicative of liver toxicity can include changes in liver enzyme concentrations in the serum,
necrosis, inflammation, and degeneration. Only the doses associated with the effects classified as
adverse were used for the quantification. With these criteria in mind, it can be concluded that
some of the observed liver effects indicate toxicity of relevance to humans as opposed to PPARa
activation unique to rodents.
For the GenX chemicals database, the studies that identified adverse liver effects at the lowest
doses are the 28-day oral (gavage) toxicity study in mice (DuPont-24459, 2008) and the oral
(gavage) reproduction/developmental toxicity study in mice (DuPont-18405-1037, 2010). In
these studies, increases in relative liver weight were accompanied by increases in hepatocellular
hypertrophy and single-cell necrosis at doses as low as 0.5 mg/kg/day. These studies were
51
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
completed according to OECD TGs and followed GLP. Both studies used a sufficient number of
mice per dose group, 10/dose group in the 28-day oral (gavage) toxicity study and 22 -25/ dose
group in the oral (gavage) reproduction/developmental toxicity study. Additionally, both studies
were completed on the mouse, which the data indicate is more sensitive than the rat to liver
effects. Thus, the liver effects noted in these studies (DuPont-24459, 2008 and DuPont-18405-
1037, 2010), specifically single-cell necrosis in males, were selected for BMD modeling.
Longer term studies identifying liver effects were also considered, but not selected for BMD
modeling. Liver effects observed in the 90-day study in mice (DuPont-18405-1307, 2010) were
observed at higher doses (greater than or equal to 5 mg/kg/day) than in the oral
reproductive/developmental toxicity study in mice (0.5 mg/kg/day). Although these studies used
the same strain of mice (Crl:CDl(ICR)), the modified developmental/reproductive study
(DuPont-18405-1037, 2010) evaluated 22-25 mice/dose group while the 90-day study in mice
only used 10 mice/dose group for liver endpoints The difference in the number of mice per
dosing group might have an impact on observed effect levels in these studies. For example, in the
90-day study, statistically significant adverse effects in the liver were not observed until
5 mg/kg/day, yet there are indications of liver damage at 0.5 mg/kg/day, although these effects
did not reach a statistically significant difference lYom Ihe control group. Specifically, absolute
and relative liver weight increased in males by 12".. and I I"... respectively, relati\ e to control
mice at 0.5 mg/kg/day. In males dosed with 0.5 mg/kg/day. 4 I n livers were observed to be
discolored, compared to 0/10 for control mice There were also increases in serum liver proteins
at 0.5 mg/kg/day in males, although they did nol differ significantly from control. AST, ALP,
and ALT increased 35%, 40%, and 35%, respecti\ ely. compared to control. Finally, these
increases in serum liver protein levels were associated with hepatocellular hypertrophy in males
at 0.5 mg/kg/day (8/10), compared to 0 in control Thus, the difference in the NOAEL for liver
effects between the 90-day study and the reproductive/developmental study might reflect on the
difference in animal number per dose group.
Additionally, the chronic 2-year cancer bioassay was not selected as the critical study in the
derivation of the Rfl) for sc\ eral reasons. Effects observed at low doses in this study include
changes in serum albumin levels and the A/G ratio in male rats. An increase in A/G ratio at 1
mg/kg,day at the 3-month time point and increases in both albumin and A/G ratio at the 12-
month time point were observed, but these changes were not seen at 6 months. These changes,
while indicative of an immune system effect, were deemed of unclear biological significance
especially given these inconsistencies. Liver effects were also observed in this study, but did not
occur at comparable doses to the oral reproductive/developmental toxicity study in mice. Also,
the available chronic study only evaluated rats, and data indicate that mice appear to be more
sensitive. For example, male mice presented with single-cell necrosis in doses as low as 0.5
mg/kg/day (5/24 mice compared to 1/25 in the control), which significantly increased at 5
mg/kg/day (24/24 mice compared to 1/25 in the control). Female mice also had a significant
increase in incidence compared to control at 5 mg/kg/day for both focal/multifocal (5/24 mice
compared to 1/24 in the control) and single-cell necrosis (21/24 mice compared to 1/24 in the
control). Conversely, male and female rats exhibited no subchronic hepatocellular necrosis in the
90-day study (DuPont-17751-1026, 2009), yet hepatocellular necrosis is observed in the chronic
study at much higher doses (DuPont-18405-1238, 2013). Specifically, rats have significant
increases in incidence compared to control in focal liver degeneration and centrilobular necrosis
at 50 mg/kg/day (male) and 500 mg/kg/day (female), respectively. While typically a chronic
52
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
study is the preferred duration for development of lifetime RfD, in this case, the oral
reproductive/developmental toxicity study indicates that adverse effects on the liver are observed
in the parental mice at lower doses than those reported in the chronic study in rats.
6.2 BMD Modeling
There are no biologically based dose-response (BBDR) models available for HFPO dimer acid
and its ammonium salt. Thus, using its Benchmark Dose Software (version 2.6), the EPA
evaluated a range of dose-response models thought to be consistent with underlying biological
processes to determine how best to empirically model the dose-response relationship in the range
of observed data.
Consistent with the EPA's Benchmark Dose Technical (liin/aiice (IJSEPA, 2012), the BMD and
the BMDL were estimated using a BMR of 10% extra risk for dichotomous data, in the absence
of information regarding the level of change considered biologically important, and to facilitate a
consistent basis of comparison across endpoinls. studies, and assessments Candidate PODs were
estimated from all three doses (plus control) for I)liPont-18405-10.^7 (201 (V) and DuPont-24459
(2008). Results of the analyses are available in Table 8. Further details, including the modeling
output and graphical results for the selected models, are |iro\ ided in appendix I- of this report.
6.3 Dosimetric Adjustment of tno Fxpo^rrvr.-nwi Animal-Based POO to PODhed
EPA guidance was followed to calculate a candidate PODiii n li'oin the BMDLio using a BW3/4
allometricscaling approach (USEPA, 2<)| 1 h) The allometric scaling approach is derived from
the relationship between body surface area and basal metabolic rate in adults. With infants and
children, surface area and basal metabolic rates are \ cry different than for adults with a slower
metabolic rate. While this BW' 1 allometric scaling is not appropriate for infants and children
because of the limited toxicokinetic data available, the critical effect of liver single-cell necrosis
observed in adult mice is not a de\ elopniental end point nor is it specific to early life stages. The
HFPO dimer acid ammonium salt Ii\lI)I.i..s from the experimental animal studies (DuPont-
18405-|d.i7. 20l<). l)nl,ont-2445^. 2<)i)K) were adjusted via the dosimetric adjustment factor
(DAI') equation below
l).\l (UWa1* '/UW"h1/4),
where:
Ii\\ animal BW
liW'h human BW
For the chronic reproductive/developmental toxicity study (DuPont-18405-1037, 2010), a BWa
value of 0.0372 kg was identified as the mean BW of the Fo male mouse controls on study day
84 (the final day of animal dosing). For the short-term 28-day oral study in mice (DuPont-24459,
2008), a BWa value of 0.0347 kg was identified as the mean BW of the male mouse controls on
study day 28 (the final day of animal dosing). A BWh of 80 kg for humans was selected based on
National Health and Nutrition Examination Survey (NHANES) sampling data (USEPA, 201 la).
For adults (more than 21 years of age), the EPA updated the default BW assumption from 70 kg
to 80 kg based on NHANES data from 1999 to 2006 as reported in Table 8.1 of the EPA's
Exposure Factors Handbook (USEPA, 201 la). The updated BW represents the mean weight for
53
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
adults ages 21 and older. The resulting DAF for the allometric scaling of doses from mice to
humans is 0.15 for DuPont-18405-1037 (2010) and 0.14 for DuPont-24459 (2008). Applying the
DAF to the two BMDLios identified for liver effects in adult mice yields a PODhed as follows:
PODhed = BMDLio animal dose (mg/kg/day) x DAF
Using the BMDLio of 0.15 mg/kg/day to complete the calculation results in a PODhed for single-
cell necrosis of the liver from DuPont-18405-1037 (2010) of 0.023 mg/kg/day. Using the
BMDLio of 0.3 mg/kg/day to complete the calculation results in a PODhed for single-cell
necrosis of the liver from DuPont-24459 (2008) of 0.042 mg'kg'dav. Table 8 presents a summary
of the determination of the PODheds.
Table 8. Summary of Determination of PODhed
Endpoint
Species/
Sex
NOAEL
(mg/kg/day)
Model
l!MR
ISM Dm
(ing/kg/daj)
BMDLio
(mg/kg/day)
DAF
PODhed3
(mg/kg/day)
Liver single-
cell necrosis
in parental
males
(DuPont-
18405-1037
2010)
Crl:CDl(I
CR) mice
F0 parental
male
0.1
Benchmark
Dose
Multistage 2
|()"„
0.37
(I 15
0.15
0.023
Liver single-
cell necrosis
in males
(DuPont-
24459 2008)
Crl:CDl(I
CR) male
mice
0.1
Benchmark
Dose
Muliisiaue 2
and niianlal-
l.iiiear
In" „
0 <>o
(Muliislaue 2)
and
1 4
iQuantal-
Linear)
0.3
0.14
0.042
Note:
aCalculated using UW-1 1 scaling (USMPA, 201 lb).
6.4
6.4.1 i nticol Study ana irjject
The oral i cpmckicti\^/developmental toxicity study in mice and liver effects (single-cell necrosis
in males) were selected as the critical study and the effect for deriving the subchronic and
chronic RfDs for I II PO dimer acid and its ammonium salt (DuPont-18405-1037, 2010). This
study used a larger sample size (n=24 / dose versus n=10 / dose in 28 day study) to provide the
most health-protective I'ODin n. Additionally, this study was of a longer duration (84/85 days
versus 28 days) in the mouse, which is the more sensitive species.
Several of the other studies provide support for the selection of this study as the critical study
and liver necrosis as the critical effect (DuPont-24459, 2008; DuPont-24447, 2008; DuPont-
18405-1307, 2010; DuPont-18405-1238, 2013; DuPont-18405-841, 2010). The liver is the
primary target organ for toxicity from oral exposure to HFPO dimer acid and its ammonium salt.
Liver effects are observed in both male and female mice and rats at varying durations of
exposures and doses of GenX chemicals. Specifically, changes in liver enzyme levels,
histopathological lesions, and tumors are observed in both male and female mice and rats at
54
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
varying durations of exposures (15 days to 2 years) and doses of these GenX chemicals (0.5-
1,000 mg/kg/day).
6,4,2 Uncertainty Factors
An interspecies uncertainty factor (UFa) of 3 (101/2= 3.16, rounded to 3) was applied to account
for uncertainty in extrapolating from laboratory animals to humans. The UFa is generally
presumed to include both toxicokinetic (i.e., absorption, distribution, metabolism, and
elimination) and toxicodynamic (i.e., MOA) aspects. A PODhed was derived from the BMDL
using the EP A's Recommended Use of Body Weight4 as the Default Method in Derivation of the
Oral Reference Dose (USEPA, 201 lb). This guidance describes approaches for deriving
PODheds from data from laboratory animals, with the preferred approach being physiologically
based pharmacokinetic (PBPK) modeling. For HFPO dimer acid and ammonium salt, no PBPK
models have been developed and published. Other approaches described by the guidance include
the use of chemical-specific data to inform the deri\ ation of human equivalent oral exposures. In
the absence of either PBPK models or chemical-specific information, a liW scaling to the 3/4
power approach is applied to extrapolate toxicologically equivalent doses of orally administered
agents from adult laboratory animals to adult luimans Although this scaling addresses some
aspects of cross-species extrapolation of toxicokinetic and toxicodynamic processes, some
residual uncertainty remains (i.e., MOA) (USEPA. 2<)| lb) Thus, in the absence of chemical-
specific data to quantify this uncertainty, the LPA's guidance recommends use of a UF of 3.
An intraspecies uncertainty factor (UI 'm) of I" is assigned to account for variability in the
responses within the human populations because of both intrinsic (toxicokinetic, toxicodynamic,
genetic, life stage, and health status) and extrinsic (life style) factors that can influence the
response to dose. ISo information to support a I I'm other than in was available to characterize
interindividual and age-related \ ariahiIit\ in the toxicokinetics or toxicodynamics.
A LOAEL to NOAEL extrapolation uncertainty factor (IJFl) of 1 is applied because a BMDL is
used as the basis for the PODm n deri\ation When the POD type is a BMDL, the current
approach is to address this factor as one of the considerations in selecting aBMRforBMD
modeling In this case, the BMR of a 10% change for the modeled liver endpoints (single-cell
necrosis in male mice) was selected under the assumption that it represents a minimal, but
biologically significant change lor these effects.
A UF for extrapolation from a subchronic to a chronic exposure duration (UFs) of 3 (101/2= 3.16,
rounded to 3) was applied for the derivation of the chronic RfD, but not of the subchronic RfD.
The study selected as the critical study was the oral reproductive/developmental toxicity study of
HFPO dimer acid ammonium salt in mice (DuPont-18405-1037, 2010). The critical effect
selected is single-cell necrosis in parental (Fo) male mice. The duration of dosing of these Fo
males was 84-85 days, a study duration that falls short of a standard subchronic study and below
the duration of a chronic study. Chronic studies typically employ repeated dosing for longer than
90 days or more than 10% of the human life span (USEPA, 2002). Chronic data in another
species are available to inform whether effects would be expected to occur with longer exposure
durations. A 2-year combined chronic toxicity/oncogenicity study is available in rats (DuPont-
18405-1238, 2013), which identified a NOAEL of 1 mg/kg/day for liver effects (increased liver
enzyme levels and centrilobular hepatocellular hypertrophy and cystic focal degeneration in
males and centrilobular necrosis in both sexes). These effects are consistent with those observed
55
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
in the oral reproductive/developmental toxicity screen in mice. The NOAELs for the oral
reproductive/developmental toxicity study and the chronic study are within one order of
magnitude of each other, suggesting consistency in dose-response relationships between these
studies. The combined chronic toxicity/oncogenicity study was conducted, however, in rats that
appear to be less sensitive than mice. For these reasons, a UF of 3 was used to account for
extrapolation from subchronic to chronic exposure duration for the chronic RfD. For the
subchronic RfD, a UF was not applied to account for duration as the study is of subchronic
duration.
A database uncertainty factor (UFd) of 3 (101/2= 3.16, rounded to 3) was applied to account for
database deficiencies. There are no human data from epidemiological studies in the general
population or worker cohorts evaluating the effects of exposure to these GenX chemicals. The
database for the HFPO dimer acid and its ammonium sail includes data submitted to the EPA
under TSCA assessing numerous endpoints such as acute toxicity, metabolism and
toxicokinetics, genotoxicity, and systemic toxicity in mice and rats with dosing durations of up to
2 years. Many of the studies were conducted according to OECD TGs and principles of GLP,
and full study reports were submitted for Agency rex iew. For the HFPO dimer acid ammonium
salt specifically, there are two short-term and two subchronic studies in both rats and mice (28
and 90 days) and a chronic carcinogen icily study (2 years) in rats One reproductive and
developmental toxicity study in mice and one prenatal and de\ elopmental toxicity study in rats
are also available. In addition to studies submitted under TSC A. a small number of studies
presenting other health data are a\ailahle in the pnhlic literature, including a single
immunotoxicity study. In some cases, the identified publications in the open literature represent
the studies previously submitted to the EPA and do not constitute additional information.
There are several deficiencies in the database for I II PO dimer acid and its ammonium salt,
however, including limited testing of developmental toxicity and immunological responses.
Other PFAS, particularly PI OA and PFOS. have been shown to lead to developmental effects,
including skeletal variations, decreased neonatal survival, altered fetal BW, and developmental
alterations such as delayed vaginal opening, accelerated preputial separation, and delayed
mammary gland development (USI-IW, 2016a, 2016b). For HFPO dimer acid and/or ammonium
salt, there is e\ idence of de\ elopmental toxicity, although at higher doses than effects on the
liver (DuPont-1X405-1037. 2<)|0. DuPont-18405-841, 2010). Significant adverse effects
observed following exposure to I ll'PO dimer acid ammonium salt include early delivery of pups,
decreased pup BW. and delays in the attainment of balanopreputial separation and vaginal
patency. The lack of a full two-generation reproductive toxicity study evaluating exposures
during early organogenesis (i e , GD 0 to GD 6) and studies evaluating additional developmental
endpoints that have been observed following exposure to other PFAS (i.e., skeletal development
in mice and altered puberty in mice (USEPA, 2016a, 2016b)) is a database deficiency.
PFAS chemicals, including PFOS and PFOA, interact with the immune system in studies of both
humans and animals (USEPA, 2016a, 2016b). The GenX chemicals database includes two
LLNAs and a 28-day immunotoxicity study (Rushing et al., 2017). Rushing et al. (2017)
identified suppression of TDAR by a reduction in antigen-specific IgM antibody production in
females and increased T cell numbers in males at the high dose only (100 mg/kg/day). The
LLNA is typically used to identify potential skin-sensitizing chemicals through their ability to
induce allergic immune response (OECD, 2010). The LLNAs were conducted with HFPO dimer
56
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
acid ammonium salt preparations of varied purity and yielded equivocal results (one positive and
one negative). Evaluation of additional immune function assays, histopathology, and immune
endpoints such as antibody levels are not available. The combined dataset was found to be weak
as it did not include sufficient measures of immunopathology, humoral immunity, cell-mediated
immunity, nonspecific immunity, or host resistance. Data on the potential for these GenX
chemicals to impact aspects of immune function beyond immunosuppression are lacking.
Additional studies, therefore, would be useful to support a more conclusive determination of
immunotoxic potential.
Finally, there are some additional research needs that factor into the database uncertainty. First,
the lack of a chronic study in the mouse, which appears to be more sensitive than the rat to GenX
chemicals exposure, is a data gap. This uncertainty, however, is also addressed in the subchronic-
to-chronic UF. Second, additional research is needed lo help determine if the inconsistent
hematological effects observed in many of the studies are ad\ eise and should be considered
critical.
57
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
NOVEMBER 2018
6.5 Subchronic RfD
The subchronic RfD is calculated as follows:
Subchronic RfD
_ PODhED
Total UF
0.023^/ day
kg
100
0.0002 mg/kg/day or 0.2 iig/kg/dav
where:
PODhed = 0.023 mg/kg/day, the HED based on the IJMDI.i.. Ibr liver effects (single-cell
necrosis) in parental male mice exposed to HFPO dimer acid ammonium salt by
gavage for 84-85 days (DuPont-184<)5-l<)37, 2010).
Total UF = 100, including a 10 for UFh, a 3 Ibr I I ' ... and a 3 for T I 'd
6.6 Chronic RfD
The chronic RfD is calculated as follows
where:
PODiin) i) (123 mu kg day, llie HED based on the BMDLio for liver effects (single-cell
necrosis) in parental male mice exposed to HFPO dimer acid ammonium salt by
gavage Ibr 84 85 days (DuPont-18405-1037, 2010).
Total UF = 300. including a 10 for UFh, a 3 for UFa, a 3 for UFs, and a 3 for UFd.
Chronic RfD
PODheD
Total UF
0.023^/day
300
0.0000S mg/kg/day or 0.08 jug/kg/day
58
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
7.0 Discussion of Uncertainties
7.1 Uncertainty and Variability
The variability and uncertainty in an RfD is a function of both intrinsic and extrinsic factors. The
EPA has identified eight short-term, subchronic, and chronic studies that provided dose-response
information and were considered during the quantitative assessment of risk. The range of
external dose NOAELs among these studies is 0.1 mg/kg/day-10 mg/kg/day. The LOAELs
range from 0.5 mg/kg/day to 100 mg/kg/day. The EPA selected studies with the lowest NOAELs
for BMD modeling and determination of the POD. The Agency believes the uncertainty in the
chosen POD is minimized because of the available data following various durations of exposure
that support the liver as the primary target of toxicity.
The intrinsic uncertainties in the assessment reflect the lacl thai the NOAELs and LOAELs are
derived using central tendency estimates for variables such as IJW. food and drinking water
intakes, and dose. The central tendency estimates are derived from small numbers of genetically,
relatively similar animals representing one or more strains of rats or mice living in controlled
environments. The animals lack the heterogeneous genetic complexity. hcha\ ioral diversity, and
complex habitats experienced by humans. These differences, lo some extent, are minimized
using the modeled outcomes and use of allometric scaling to help inform the application of the
UF.
While the EPA has routinely used li\\ to allometrically scale toxicity data from animal test
species to HEDs during the development of human health risk assessments, the applied
methodology is not without limitation (USEPA, 2<>l lb) Allometric scaling using BW scaled to
the 3/4 power primarily addresses uncertainty associated with toxicokinetics, although the exact
amount of uncertainty addressed In this method for any specific chemical is often not
quantifiable. In following the recommended method to apply BW14 scaling, it remains possible
that the toxicokinetic uncertainty associated with (ienX chemicals might be more or less than
what is accounted for using this scaling methodology. BWV4 scaling is appropriate in this
scenario because GenX chemicals are not metabolized and have relatively short clearance times,
especially compared to other longer chain PFAS chemicals such as PFOA (USEPA, 201 lb;
DuPont-1 X4< 15-1017 RV1, 2< • I I; Gannon et al., 2016). The BW3/4 scaling methodology is not
appropriate. howe\ er, when using children's BWs. This limitation exists due to the absence of
quantitative information describing the toxicokinetic and toxicodynamic differences between test
animals and early life stage humans (USEPA, 201 lb). Because the liver effects observed
following exposure to GenX chemicals were in adult animals, the allometric scaling
methodology was scaled to the a\ erage adult human BW.
Variability in the study outcomes is extrinsically a function of study design and the endpoints
monitored. Studies of systemic toxicity monitor an array of endpoints that are not evaluated in
studies of reproductive, developmental, neurological, and immunological toxicity. The reverse is
true for the other types of toxicity studies compared to standard short-term and long-term
systemic studies. Studies of systemic toxicity do not often examine neurological or
immunological endpoints. Increases in liver weight were seen in many of the studies with dose-
response information, and the histological evaluation of the liver supported a determination that
the increase in liver weight when it is accompanied by necrosis can be considered as adverse
rather than adaptive, according to the Hall et al. (2012) criteria. Increases in relative liver weight
59
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
with confirmed liver necrosis were observed in DuPont-24447 (2008), DuPont-24459 (2008),
DuPont-18405-1307 (2010), DuPont-18405-1238 (2013), and DuPont-18405-1037 (2010).
The chronic RfD is based on the PODhed derived from the parental males from the oral
reproductive/developmental toxicity study in mice with application of a total UF of 300 to
account for variability in the human population, database uncertainties, and possible differences
in the ways in which humans and rodents respond to the HFPO dimer acid and/or its ammonium
salt that reaches their tissues (DuPont-18405-1037, 2010). The selected RfD is based on the
adverse liver effects observed in the parental male animals. Selection of this endpoint is expected
to provide protection to both the sensitive life stages and the general population. The RfD is
supported by the outcomes from other studies (DuPont-1 775 l-l<)2(\ 2009; DuPont-18405-1037,
2010; DuPont-24459, 2008) based on different endpoinls. including hematological, immune, and
developmental effects. These supporting data from the I II PO dimer acid and its ammonium salt
database increase confidence in the RfD.
7.2 Composition of Test Substance
Most of the available data for HFPO dimer acid and or its ammonium sail were submitted to the
EPA by DuPont/Chemours, the manufacturer of GenX chemicals, under TSC.V including with
PMNs, as required pursuant to a consent order for these chemicals (USEPA, 2009) or as required
under TSCA reporting requirements (e g . section 8(e)) In these submissions, DuPont/Chemours
provided information on the purity of the test substance used in each of the studies. Purity ranged
from 84 to 88% across the toxicity studies considered in this assessment. DuPont/Chemours
provided a certificate of these analyses and noted that they were conducted under EPA GLP
standards (40 CFR part 792). The major impurity identified is water (12.7%—13.3%). Trace
amounts of PFOA were also identified in the test substance (3 4 I 50 parts per million).
DuPont/Chemours noted that test results were adjusted for purity based on the reported test
article formulations. Based on the information pro\ ided, administered doses of PFOA present as
a contaminant in the formulations used by DuPont Chemours are low. For example, in the
critical study chosen for the dei i \ ation of the RfDs. the dose of administered PFOA is 0.000075
mg/kg'day at the GenX chemicals NOAEI. (<) I mg/kg/day) (DuPont-18405-1037, 2010). For
PFOA. \().\F.I.s ranging from to 30 mg/kg-d have been identified for effects including
developmental, liver, and immune endpoints (US EPA, 2016a). Despite trace amounts of PFOA
that might be present as an i in purity, the EPA recognizes the potential for this impurity to
contribute to the observed toxicity at very high doses of GenX chemicals. At present, however,
discerning the contribution of this low level of PFOA to observed toxicity is not possible. Thus,
the EPA concluded that the presence of PFOA at these low levels is not the primary driver of
toxicity observed in the studies. Of note is that the same test substance (Lot H-28548) was used
in the 90-day mouse and rat studies, the chronic rat study, and the oral reproductive and
developmental toxicity and prenatal developmental toxicity studies (DuPont-17751-1026, 2009;
DuPont-18405-1307, 2010; DuPont-18405-841, 2010; DuPont-18405-1238, 2013). Additionally,
the same test substance (Lot H-28397) was used in both the mouse and rat 28-day studies
(DuPont-24459, 2008; DuPont-24447, 2008). Despite differences in test substance purity,
adverse effects were observed consistently across the DuPont/Chemours studies. The available
published literature did not report purity in their methods and formulations of HFPO dimer acid
and ammonium salt (Rushing et al., 2017; Sheng et al., 2018; Wang et al., 2016).
60
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Given the database for GenX chemicals, the quality of these studies—including adequacy of
reporting of methods and results—and the weight of evidence for effects on the liver,
hematological and immune systems, and reproductive and developmental endpoints, the EPA
concluded that the DuPont/Chemours studies demonstrated adverse effects as a result of
exposure to the HFPO dimer acid ammonium salt formulations and were appropriate for
derivation of toxicity values for these chemicals.
7.3 Use of Data-Derived Extrapolation Factors
For HFPO dimer acid and/or ammonium salt, there are no human data and no BBDR or PBPK
models available to evaluate toxicokinetic and toxicodynamic differences between humans and
animals. Additionally, only a few studies are available in rats and mice that evaluate
toxicokinetics. These studies indicate that there is little lo 110 metabolism and that clearance is
relatively rapid compared to other longer chain PFAS \l().\ (holli in vivo and in vitro) data are
also inadequate. The EPA considered the 2014 Guidance for Applying Quantitative Data to
Develop Data-Derived Extrapolation Factors for Interspecies and Iniraspecies Extrapolation in
determining UFa and UFh (USEPA, 2014c). I sing the decision process described in Figure 2,
the EPA concluded that data are not adequate lo support derivation of data-derived extrapolation
factors. Specifically, given the lack of available models and data to address external dose and
clearance in humans, default approaches to the application of I Ta and UFh were employed,
including BW scaling for oral exposure (I SI - PA, 2< > I I It) These approaches are described
further below.
7.4 Limited Da
One study is available on e\ aliiatinu carcinogenicity of I II PO dimer acid and its ammonium salt
in rats (DuPont-184<)5-l 23S. 2<)|3) I11 this study. Ii \ er and pancreatic tumors were noted at the
highest doses tested The a\ ailahle data for I II PO diiner acid ammonium salt suggest that mice
might be more sensitix e to exposure to these (ien.X chemicals than rats. Given the evidence that
the liver is the target organ for toxicity and the primary organ for tumor development, there is a
need lor additional research using chronic duration exposures in mice. This uncertainty is not
considered in the application of the lTi>gi\en that a noncancer toxicity value was developed for
this assessment
7.5 Effects on Bilirubin
A decrease in serum bilirubin is a consistent effect observed across multiple studies, especially in
female rodents (DuPont-17751-1026, 2009; DuPont-18405-1238, 2013; DuPont-18405-1307,
2010; Wang et al., 20l(->) This finding was surprising given that increased serum bilirubin levels
rather than decreased le\ els are typically indicative of liver damage, and multiple studies
outlined above have confirmed microscopic liver damage (DuPont-18405-841, 2010; DuPont-
18405-1238, 2013; DuPont-18405-1037, 2010; DuPont-18405-1307, 2010; Tietze, 2012). In
female mice and rats, however, serum bilirubin levels were significantly decreased by 14%-50%
relative to controls when the females were administered 5 mg/kg/day-1,000 mg/kg/day of HFPO
dimer acid ammonium salt for 3-12 months (DuPont-17751-1026, 2009; DuPont-18405-1307,
2010; DuPont-18405-1238, 2013). Additionally, male ICR mice treated with 1 mg/kg/day of
HFPO dimer acid ammonium salt exhibited a significant 37% and 45% decrease in total and
direct bilirubin, respectively, when compared to controls (Wang et al., 2016); this finding was
not replicated in the other 28-day studies (DuPont-24459, 2008; DuPont-24447, 2008). The
61
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
biological or mechanistic significance of this effect is unknown, yet the consistency of this effect
across multiple studies is noteworthy.
7.6 Susceptible Populations and Life Stages
Data for the elucidation of differential susceptibility dependent on life stage (e.g., developing
fetus, women of reproductive age, or pregnant women) are not available. Children are frequently
more vulnerable to pollutants than the average adult because of the differences in their behaviors
and biology. These differences can result in greater exposures and/or unique windows of
developmental susceptibility during the prenatal and postnatal periods for both pregnant mothers
and the developing fetus. No human toxicity or epidemiological studies are available in the
literature that address early developmental or reproductive life stages. DuPont submitted data
examining reproductive and developmental endpoints in both mice and rats (DuPont-18405-
1037, 2010; DuPont-18405-841, 2010), and summaries of these studies can be found in section
5.4 (Developmental/Reproductive). HFPO dimer acid ammonium salt can be transferred from a
pregnant animal to the fetus, although with the a\ ailable data, it cannot be determined if this
transfer occurs during gestation or during laclation (DuPont- l 8405-I <)37. 2<~>10). When present,
developmental and reproductive effects were found at doses higher than those associated with
the selected critical effect: single-cell necrosis in the liver of male mice. The UFn of 10 accounts
for variability in the responses within human populations because of both intrinsic (including life
stage) and extrinsic (life style) factors that can influence the response to dose. No information to
characterize interindividual and age-related \ariability in the toxicokinetics or toxicodynamics is
available. Thus, the RfDs provided in sections 6.5 and 6 6 (Subchronic RfD and Chronic RfD)
are applicable to all life stages. While this document is not itself an assessment of risk, when
reviewing data pertinent to the hazard potential of (ien.X chemicals, the EPA adhered to the
requirements of its 2013 reaffirmation of the Policy on l '.\aluating Health Risks to Children
(USEPA, 2013).
Sex-specilic \ ariation in the toxicokinetics of these two GenX chemicals is pronounced in
rodents Toxicokinetic data show the HFPO dimer acid and its ammonium salt clearance times to
be considerably faster for females than for males (see the summary in section 2.3
(Toxicokinetics). For example. DuPont-24281 (2008) identified 12 hours as the clearance time
for HFPO dimer acid ammonium salt in male rats at the low dose and 22 hours for the high dose.
In female rats, the clearance \ allies were 4 and 8 hours for the low dose and the high dose,
respectively. A difference in toxicokinetics was not observed in primates where beta phase T1/2S
for male and female monkeys were 64.1 and 79.6 hours, respectively. The observed sex-specific
toxicokinetic differences in rodents likely contribute to the observed sex-specific differences in
toxic response.
Toxicity also varied by sex. Most of the statistically significant changes in clinical chemistries of
rats were observed in the males. Decreases in total globulin and increases in the A/G ratio were
observed in males and female rats; however, male rats exhibited a total globulin decrease at the
3- and 30-mg/kg/day doses while females responded only at 300 mg/kg/day (DuPont-24447,
2008). The most prevalent liver effects following both subchronic and chronic exposure to HFPO
dimer ammonium salt were hepatocellular hypertrophy and single-cell and/or focal necrosis. In
both sexes of mice exposed subchronically, hepatocellular hypertrophy was observed at
0.5 mg/kg/day, while male and female rats showed these effects at 3 mg/kg/day and
30 mg/kg/day, respectively (DuPont-17751-1026, 2009; DuPont-18405-841, 2010; DuPont-
62
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
18405-1238, 2013; DuPont-18405-1037, 2010; DuPont-18405-1307, 2010; DuPont-24447,
2008; DuPont-24459, 2008; Wang et al., 2016). In the oral/reproductive subchronic study, male
and female mice presented single-cell necrosis in doses as low as 0.5 mg/kg/day (5/24 mice, or
21% in males and 2/24 mice, or 8% in females (not a significant increase from control in
females), which significantly increased at 5 mg/kg/day (24/24 male mice and 21/24 female
mice). While these findings indicate that mice are more sensitive to liver necrosis than rats in
subchronic exposure scenarios, they also show that adverse impacts on the clinical chemistry and
liver toxicity manifest at lower doses in males than females. The critical effect of liver single-cell
necrosis in male mice was found at the lowest identified NOAEL. Its use in the derivation of the
subchronic and chronic RfDs is assumed to be protects e of females as well.
63
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
8.0 References
Beekman, M., P. Zweers, A. Muller, W. de Vries, P. Janssen, and M. Zeilmaker. 2016.
Evaluation of Substances Used in the GenX Technology by Chemours, Dordrecht. RIVM
Letter Report 2016-0174. The Netherlands, National Institute for Public Health and the
Environment, Ministry of Health, Welfare and Sport. Accessed May 2018.
https://www.rivm.nl/dsresource?obiectid=3186e480-7d d-acS9-
acd2e804d3b5&type=org&disposition=inline.
Chemours. 2018. Sustainability-GenX. The Chemours Company. Accessed May 2018.
https://www.chem.ours.com/Industrial Bake GB/sustain ability/dib s gemx
.html.
Cheng, W., and C.A. Ng. 2018. Predicting relative prolan affinity of novel per- and
polyfluoroalkyl substances (PFAS) by an efficient molecular dynamics approach.
Environmental Science Technology 52:7972-7980.
DuPont CCAS (DuPont Corporate Center for Analylical Sciences). 2UIN Sublimation of
Processing Aids FRD-903K and PRD-'JO 2. ed A D English. CCAS. Wilmington, DE.
DuPont-2-63: E.I. du Pont de Nenionrs and Company I lH-<3 .1 cute Oral Test. Test guideline not
identified. Study conducted by I laskell Laboratory for Toxicology and Industrial
Medicine (Study Completion Dale January 3. 1963) Testing laboratory location not
identified.
DuPont-770-95 LI du Pont de Nemours and Company llW Approximate Lethal Dose (ALD)
ofH-212K> in Rats Test guideline not identified. Study conducted by E.I. du Pont de
Nemours and Company (Study Completion Date: February 26, 1996), Newark, DE.
DuPoni-839-95 L I du Pont de Nemours and Company. 1996. Approximate Lethal Dose (ALD)
by Skm. \bs(>/¦/>/ioii <>j 11-212 K> in Rabbits. Test guideline not identified. Study conducted
by L I du Pont de Nemours and Company (Study Completion Date: April 1, 1996),
Newark. I)L
DuPont-1 750S-I (->75: E.I. du Pont de Nemours and Company. 2008. Estimation of the Adsorption
Coefficient fK ) of the Hll'O Dimer Acid Ammonium Salt on Soil and Sludge. OECD
Test Guideline 121 Study conducted by DuPont Haskell Global Centers for Health and
Environmental Sciences (Study Completion Date: September 11, 2008), Newark, DE.
DuPont-17751-723: E.I. du Pont de Nemours and Company. 2009. H-28548: Inhalation Acute
Exposure with Anatomic Pathology Evaluation in Rats. Test guideline not identified.
Study conducted by E.I. du Pont de Nemours and Company (Study Completion Date:
May 11, 2009), Newark, DE.
DuPont-17751-1026: E.I. du Pont de Nemours and Company. 2009. A 90-Day Oral (Gavage)
Toxicity Study of H-28548 in Rats with a 28-Day Recovery. OECD Test Guideline 408.
Study conducted by WIL Research Laboratories, LLC (Study Completion Date: October
5, 2009), Ashland, OH.
64
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
DuPont-17751-1579 RV1: E.I. duPont de Nemours and Company. 2009. Cross-species
Comparison of FRD-902 Plasma Pharmacokinetics in the Rat and Primate Following
Intravenous Dosing. Test guideline not identified. Study conducted by E.I. du Pont de
Nemours and Company (Original Report Completed: December 8, 2008; Report Revision
1 Completed: February 2, 2009), Newark, DE.
DuPont-18405-841: E.I. du Pont de Nemours and Company. 2010. An Oral (Gavage) Prenatal
Developmental Toxicity Study ofH-28548 in Rats. U.S. EPA OPPTS 850.3700; OECD
Test Guideline 414. Study conducted by WIL Research Laboratories, LLC (Study
Completion Date: July 2, 2010), Ashland, OH.
DuPont-18405-849 RV1: E.I. duPont de Nemours and Company 2011. H-28548: Toxicokinetic
Study in Pregnant Rats. Test guideline not idenli lied Study conducted by E.I. du Pont de
Nemours and Company (Original Report Completed March 29, 2011; Report Revision 1
Completed: April 11, 2011), Newark, DM
DuPont-18405-1017 RV1: E.I. du Pont de Nemours and Company. 201 I H-28548: Absorption,
Distribution, Metabolism, and Elimination in the Rat. U.S. EPA OPPTS 870.7485. Study
conducted by E.I. du Pont de Nemours and Company (Original Report Completed:
November 3, 2010; Report Re\ ision 1 Completed April 21, 2011), Newark, DE, and
Wilmington, DE.
DuPont-18405-1037: E.I. duPont de Nemours and Company 2010. An Oral (Gavage)
Reproduction/Developmental Toxicity Screening Snii/y <>J H-28548 in Mice. U.S. EPA
OPPTS 870.3550; OECD Test Guideline 421 Study conducted by WIL Research
Laboratories, LLC (Study Completion Date December 2^. 2010), Ashland, OH.
DuPont-18405-1238: E.I. du Pont de Nemours and Company. 2013. H-28548: Combined
Chrome Toxicity/Oncogeniciiy Study 2-) ear Oral Gavage Study in Rats. U.S. EPA
OPPTS S70.4300; OI-CI) Test Guideline 453 Study conducted by MPI Research, Inc.
(Study Completion Date: March 28, 2d 13). Mattawan, MI.
DuPont-1 S4< 15-1307: E.I. du Pont de Nemours and Company. 2010. H-28548: Subchronic
toxicity Vfl-Pay (lavage Study in Mice. OECD Test Guideline 408. Study conducted by
E.I. du Pont de Nemours and Company (Study Completion Date: February 19, 2010),
Newark. DI-.
DuPont-18647-1017 RVI I- I du Pont de Nemours and Company. 2011. H-28548: Absorption,
Distribution. Metabolism, and Elimination in the Mouse. U.S. EPA OPPTS 870.7485.
Study conducted by E.I. du Pont de Nemours and Company (Original Report Completed:
November 3, 2010; Report Revision 1 Completed: April 21, 2011), Newark, DE, and
Wilmington, DE.
DuPont-19713 RV1: E.I. du Pont de Nemours and Company. 2008. H-27529: Bacterial Reverse
Mutation Test. U.S. EPA OPPTS 870.5100; OECD Test Guideline 471. Study conducted
by E.I. du Pont de Nemours and Company (Original Report Completed: May 31, 2006;
Report Revision 1 Completed: February 22, 2008), Newark, DE.
65
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
DuPont-19714 RV1: E.I. du Pont de Nemours and Company. 2008. H-27529: In Vitro
Mammalian Chromosome Aberration Test in Chinese Hamster Ovary Cells. U.S. EPA
OPPTS 870.5375; OECD Test Guideline 473. Study conducted by E.I. du Pont de
Nemours and Company (Original Report Completed: June 27, 2006; Report Revision 1
Completed: February 25, 2008), Newark, DE.
DuPont-19897: E.I. du Pont de Nemours and Company. 2006. H-27529: Local Lymph Node
Assay (LLNA) in Mice. U.S. EPA OPPTS 870.2600; OECD Test Guideline 429. Study
conducted by E.I. du Pont de Nemours and Company (Study Completion Date: June 9,
2006), Newark, DE.
DuPont-22616 RV1: E.I. du Pont de Nemours and Company. 2<)< '7 H-28072: Local Lymph
Node Assay (LLNA) in Mice. U.S. EPA OPPTS X7t) 2(->oo. OECD Test Guideline 429.
Study conducted by E.I. du Pont de Nemours and Company (Original Report Completed:
July 2, 2007; Report Revision 1 Completed October 1, 2<)()7). Newark, DE.
DuPont-22620 RV1: E.I. du Pont de Nemours and ('ompany. 2009. H-2X0 ~2: In Vitro
Mammalian Chromosome Aberration I est in Chinese Hamster ()vury Cells. U.S. EPA
OPPTS 870.5375; OECD Test Guideline 473 Study conducted by L I du Pont de
Nemours and Company (Original Report Completed July 25, 2007; Report Revision 1
Completed: September 23, 2')'»*¦•>). Newark. 1)1-
DuPont-22734 RV1: E.I. du Pont de Ncmouis and Company 2< >08. H-28072: Bacterial Reverse
Mutation lest. U.S. EPA OPPTS X7t) 5 I no. OECD l est Guideline 471. Study conducted
by E.I. du Pont de Nemours and Company (Original Report Completed: July 26, 2007;
Report Revision I Completed: August 13. 2<)08), Newark, DE.
DuPont-22932: E.l. du Pont de Nemours and Company. 2007. H-28072: Acute Oral Toxicity
Smdy m Rats (Ip-and-l)own Procedure I S. EPA OPPTS 870.1100; OECD Test
Guideline 425. Study conducted by E.l. du Pont de Nemours and Company (Study
Completion Date: July 25, 2007), Newark, DE.
DuPont-2321 ^ I !.I. du Pont de Nemours and Company. 2007. H-28072: UnscheduledDNA
Synthesis (USD) Test with Mammalian Cells In Vivo. OECD Test Guideline 486. Study
conducted by BioReliance (Study Completion Date: August 14, 2007), Rockville, MD.
DuPont-23220: E.I du Pont de Ncinours and Company. 2007. H-28072: In VivoMicronucleus
and Chromosome . \berraiion Assay in Mouse Bone Marrow Cells. U.S. EPA OPPTS
870.5395; OECD l est Guidelines 474 and 475. Study conducted by BioReliance (Study
Completion Date: October 10, 2007), Rockville, MD.
DuPont-23459: Study Sponsor not identified. 2007. In Vitro Trout Hepatocyte Bioaccumulation
Screen. Test guideline not identified. Study conducted by Haskell Laboratory Discovery
Toxicology Group (Study Completion Date: June 15, 2007). Testing laboratory location
not identified.
DuPont-23460: Haskell Laboratory Discovery Toxicology Group. 2007. In Vitro Rat Hepatocyte
Screen. Test guideline not identified. (Study Completion Date: June 12, 2007). Testing
laboratory location not identified.
66
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
DuPont-24009: Dupont Haskell Global Centers for Health and Environmental Sciences. 2008.
Repeated Dose Oral Toxicity 7-Day Gavage Study in Rats. Test guideline not identified.
(Report Issue Date: February 14, 2008). Testing laboratory location not identified.
DuPont-24010: Dupont Haskell Global Centers for Health and Environmental Sciences. 2008.
Repeated Dose Oral Toxicity 7-Day Gavage Study in Mice. Test guideline not identified.
(Report Issue Date: February 14, 2008). Testing laboratory location not identified.
DuPont-24030: E.I. du Pont de Nemours and Company. 2007. FRD-902: Acute Dermal
Irritation Study in Rabbits. U.S. EPA OPPTS 870.2500; OECD Test Guideline 404.
Study conducted by E.I. du Pont de Nemours and Com puny (Study Completion Date:
November 21, 2007), Newark, DE.
DuPont-24113: E.I. du Pont de Nemours and Company 2')'>7 IRI >-902: Acute Dermal Toxicity
Study in Rats. U.S. EPA OPPTS 870.120<\ OIX'D Test Guideline 402. Study conducted
by E.I. du Pont de Nemours and Company (Study Completion Dale: November 28,
2007), Newark, DE.
DuPont-24114: E.I. du Pont de Nemours and Company 2<)<>7 FRD-902:. \ciile lye Irritation
Study in Rabbits. U.S. EPA OPPTS 870.24<)(). OI-CI) Test Guideline 405. Study
conducted by E.I. du Pont de Nemours and Company (Study Completion Date: December
14, 2007), Newark, DE.
DuPont-24116: Dupont Haskell Global Centers for I leallh and I ji\ironmental Sciences. 2008.
Repeated I)nsc ()ral toxicity 7-l)ay (lavage Sim/y in Rats l est guideline not identified.
(Report Issue Date I'ehruary 14, 2008). Testing laboratory location not identified.
DuPont-24126: E.I. du Pont de Nemours and Company. 2007. FRD-902: Acute Oral Toxicity
Study in Mice I p-and-Pown Procedure I S. EPA OPPTS 870.1100; OECD Test
Guideline 425 Study conducted by E.I. du Pont de Nemours and Company (Study
Completion Dale: No\ ember 29, 2007), Newark, DE.
DuPont-24128 E.I. du Pont de Nemours and Company. 2008. Determination of the Water
Solubility and Vapor I'ressnre oj 11-28307. U.S. EPA OPPTS 830.7840 and 830.7950;
OECD Test Guidelines 104 and 105. Study conducted by Wildlife International, Ltd.
(Study Completion Date: March 27, 2008), Easton, MD.
DuPont-24129: E.I du Pont de Nemours and Company. 2008. Determination of the Water
Solubility and \ 'apor Pressure ofH-28308. U.S. EPA OPPTS 830.7840 and 830.7950;
OECD Test Guidelines 104 and 105. Study conducted by Wildlife International, Ltd.
(Study Completion Date: March 27, 2008), Easton, MD.
DuPont-24198: E.I. du Pont de Nemours and Company. 2008. Determination of Dissociation
Constant ofH-28308: Revision 1. U.S. EPA OPPTS 830.7370; OECD Test Guideline
112. Study conducted by Wildlife International, Ltd. (Study Completion Date: April 1,
2008), Easton, MD.
67
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
DuPont-24199: E.I. du Pont de Nemours and Company. 2008. H-28308: An Evaluation of
Hydrolysis as a Function ofpH. U.S. EPA OPPTS 835.2110; OECD Test Guideline 111.
Wildlife International, Ltd. (Study Completion Date: March 27, 2008), Easton, MD.
DuPont-24281: Dupont Haskell Global Centers for Health and Environmental Sciences. 2008.
Biopersistence and Pharmacokinetic Screen in the Rat. Test guideline not identified.
(Report Issue Date: February 13, 2008). Testing laboratory location not identified.
DuPont-24286: Dupont Haskell Global Centers for Health and Environmental. 2008.
Biopersistence and Pharmacokinetic Screen in the Rat. Test guideline not identified.
Study conducted by Critical Path Services Sciences (Suidy Completion Date: October 10,
2007). Testing laboratory location not identified
DuPont-24447: E.I. du Pont de Nemours and Company Zoos . 1 2X-I)ay Oral (Gavage) Toxicity
Study ofH-28397 in Rats with a 28-Day Recovery Ol.(l) Test Guideline 407. Study
conducted by WIL Research Laboratories. LLC (Study Com pi el i on Date: August 22,
2008), Ashland, OH.
DuPont-24459: E.I. du Pont de Nemours and Com puny 2<)()X . 1 28-Day ()ral f( lavage) Toxicity
Study ofH-28397 in Mice with a 28-Day Recovery () 1.(1) Test Guideline 407. Study
conducted by WIL Research Laboratories, LLC (Siudv Completion Date: August 29,
2008), Ashland, OH.
DuPont-24637: DuPont Haskell Global Centers for I leallh and Ln\ironmental Sciences. 2008.
Physical am! ('liemical (Iniracteristics of J'RI >-VU2: State of the Substance,
Melting/ freezing I'oini. Boiling Point. Relative Pensiiy. Surface Tension, Flash Point,
Auto-fgnition lemperaiiire am! Viscosity OECD Test Guidelines 102 and 115; ASTM
Methods D 92, 445. 44(\ Sl) I. and 1 120, ASTM Method E 659-78. Study conducted by
Case Consulting Laboratories. Inc (Study Completion Date: May 5, 2008), Whippany, NJ.
DuPonl-24(->l)S DuPont I laskell Global Centers for Health and Environmental Sciences. 2008.
J'/iysical and {'liemical Characteristics oj FRD-903: State of the Substance,
Melung/Freezing Ponu. Boiling Roint, Relative Density, Surface Tension, Flash Point,
Auto-Ignition Temperature and I iscosity. OECD Test Guidelines 102 and 115; ASTM
Methods D 92. 445, 44(\ SlH, and 1120; ASTM Method E 659-78. Study conducted by
Case Consulting Laboratories, Inc. (Study Completion Date: May 5, 2008), Whippany, NJ.
DuPont-25281: Dupont I laskell Global Centers for Health and Environmental Sciences. 2008.
Repeated Dose ()ral toxicity 7-Day Gavage Study in Male Mice. Test guideline not
identified. (Report Issue Date: February 14, 2008). Testing laboratory location not
identified.
DuPont-25292: E.I. duPont de Nemours and Company. 2008. Determination of a Permeability
Coefficient (Kp) for H-28308 Using Human and Rat Skin Mounted in an fn Vitro Static
Diffusion Cell. Test guideline not identified. Testing laboratory not identified (Study
Completion Date: February 27, 2008). Testing laboratory location not identified.
68
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
DuPont-25300: Dupont Haskell Global Centers for Health and Environmental Sciences. 2008.
Biopersistence and Pharmacokinetic Screen in the Mouse. Test guideline not identified.
(Report Issue Date: July 31, 2008). Testing laboratory location not identified.
DuPont-25438 RV1: E.I. du Pont de Nemours and Company. 2008. H-28308: Acute Oral
Toxicity Study in Rats—Up-and-Down Procedure. U.S. EPA OPPTS 870.1100; OECD
Test Guideline 425. Study conducted by E.I. du Pont de Nemours and Company (Original
Report Completed: May 28, 2008; Report Revision 1 Completed: July 23, 2008),
Newark, DE.
DuPont-25875: E.I. du Pont de Nemours and Company 2<)i)X IRl )-903: Acute Oral Toxicity
Study in Rats—Up-and-Down Procedure. U.S I-PA OPPTS 870.1100; OECD Test
Guideline 425. Study conducted by E.I du Ponl de Nemouis and Company (Study
Completion Date: October 13, 2008), Newark. 1)1-
DuPont-25938 RV1: E.I. du Pont de Nemours and Company. 200S /I-2S397: Activated Sludge
Respiration Inhibition Test. OECD Test Guideline 209. Study conducted by DuPont
Haskell Global Centers for Health and I -n\ iron mental Sciences (Study Completion Date:
September 5, 2008; Revision Date: October 21. 2<)0S). Newark, Dl-
DuPont-26129: E.I. du Pont de Nemouis and Company 2<)()X 11-28548: In Vitro Mammalian
Cell Gene Mutation Test (L? I ~s) I A - Mouse I ympl ion la Assay). U.S. EPA OPPTS
870.5300; OECD Test Guideline 47(-> Study conducted by BioReliance (Study
Completion Date June 25. 2<)()X). Rock\ille. Ml)
DuPont-26349: E.I du Pont de \emouis and Company 2<)()X. Determination of the Dissociation
Constant and I I -I IS. \hsorpnon Spectra of 11-28307. U.S. EPA OPPTS 830.7370;
OECD Test Guidelines |n| and 112 Study conducted by Wildlife International, Ltd.
(Study Completion Date September 17. 2<)()X). l-aston, MD.
DuPont-138X23 1-R2<)i)i)\('<>3 l(a)-<>2 I- I du Pont de Nemours and Company. 2010. Report for
Inherent Biodegradauon of IRD903 ('ModifiedMITI (II) Test). OECD Test Guideline
3<)2C Study conducted In Key l.ab of Pesticide Environmental Assessment and
Pollution Control, Ml-P (Study Completion Date: January 20, 2009), Nanjing, China.
DuPont-138823 I-R2<~><~>9NC<>3 l(s)-02: E.I. du Pont de Nemours and Company. 2010. Report for
Inherent Btodegradai/on of FRD902 (ModifiedMITI (II) Test). OECD Test Guideline
302C. Study conducted by Key Lab of Pesticide Environmental Assessment and
Pollution Control. Ml-P (Study Completion Date: January 20, 2009), Nanjing, China.
DuPont-A080558: Du Pont-Mitsui Fluorochemicals Company, Ltd. 2009. Ready
Biodegradability Test of FRD903. Test guideline not identified. Study conducted by
Mitsubishi Chemical Medience Corporation, Yokohama Laboratory (Study Completion
Date: May 25, 2009), Yokohama, Japan.
DuPont-A080560: Du Pont-Mitsui Fluorochemicals Company, Ltd. 2009. Bioconcentration
Study ofFRD903 with Carp. Test guideline not identified. Study conducted by Mitsubishi
Chemical Medience Corporation, Yokohama Laboratory (Study Completion Date: June
26, 2009), Yokohama, Japan.
69
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Dupont-PMN Attachment 119. 2008. Rates of Thermal Transformation of 4(A) & 4(B), pp. 1647-
1648. Premanufacture Notice. Received in non-CBI form on April 4, 2018.
DuPont. 2013. Fluoropolymers. E.I. du Pont de Nemours and Company. Accessed May 2018.
http://www2.diipont.com/Directories/en CA/Products Services Index/Plastics Polymers
El astom er s/F tuoropol y m ers.html.
Erol, E., L.S. Kumar, G.W. Cline, G.I. Shulman, D.P. Kelly, and B. Binas. 2003. Liver fatty
acid-binding protein is required for high rates of hepatic fatty acid oxidation but not for
the action of PPARa in fasting mice. The FASEB Journal 18(2):347-349.
Gannon, S.A., W.J. Fasano, M.P. Mawn, D.L. Nabb. R (' linck. I. W. Buxton, G.W. Jepson, and
S.R. Frame. 2016. Absorption, distribution, metabolism. excretion, and kinetics of
2,3,3,3-tetrafluoro-2-(heptafluoropropoxy)pi'opanoic acid ammonium salt following a
single dose in rat, mouse, and cynomolgus monkey, loxico/ogy 340(18): 1-9.
Gebbink, W.A., L. van Asseldonk, and S.P.J \ an I .eeuwen. 2017. Presence of emerging per- and
polyfluoroalkyl substances (PFASs) in ri\ er and drinking water near a fluorochemical
production plant in the Netherlands. Environmental Science and lechnology 51:11057—
11065.
Gomis, M.I., R. Vestergren, D. Boru. and IT Cousins 2<)|x Comparing the toxic potency in
vivo of long-chain perfluoroalkyl acids and lliiorinaled alternatives. Environment
International 113:1-9.
Hall, A.P., C.R. El com he. .1 R I-osier, T I larada. \V kauliiiann. A Knippel, K. Kiittler, D.E.
Malarkey, R R Maronpol. A. Nishikaua. T Nolle. A Scluilte, V. Strauss, and M.J.
York. 2012. Li\ er hypertrophy A review of adaptive (adverse and non-adverse)
changes—Conclusions from the 3rd International ESTP Expert Workshop. Toxicologic
Pathology 4'i(7)-971
Heydehreck. I '. L. Tanu. \. Zhiyong, and R. Ebinghaus. 2015. Alternative and legacy
peril uoroalky1 substances: Differences between European and Chinese river/estuary
systems Environmental Science and Technology 49:8386-8395.
Hoke, R.A., B.D. I'enell, T.L. Sloman, R.C. Buck, and L.W. Buxton. 2016. Aquatic hazard,
bioaccumulalion and screening risk assessment for ammonium 2,3,3,3-tetrafluoro-2-
(heptafluoropropoxy )-propanoate. Chemosphere 149:336-342.
Klaassen, C.D., ed. 199(v Casarett andDoull's Toxicology. McGraw and Hill, New York.
Klaunig, J.E., M.A. Babich, L.P. Baetcke, J.C. Cook, J.C. Corton, R.M. David, J.G. DeLuca,
D.Y. Lai, R.H. McKee, J.M. Peters, R.A. Roberts, and P A. Fenner-Crisp. 2003. PPARa
agonist-induced rodent tumors: Modes of action and human relevance. Critical Reviews
in Toxicology 33(6):655-780.
Klaunig, J.E., B.A. Hocevar, and L.M. Kamendulis. 2012. Mode of action analysis of
perfluorooctanoic acid (PFOA) turnorigenicity and human relevance. Reproductive
Toxicology 3 3 (4): 410-418.
70
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Maloney, E.K., and D.J. Waxman. 1999. Trans-activation of PPARa and PPARy by structurally
diverse environmental chemicals. Toxicology & Applied Pharmacology 161(2):209-218.
NCDEQ (North Carolina Department of Environmental Quality). 2017. 60-Day Notice of Intent
to Suspend Permit NC00035 73: The Chemours Company, Fayetteville Works. NCDEQ.
Accessed May 2018. https://files.nc.gov/ncdeq/GenX/60-
Dav%20Notice%20of%20Intent%20-%20Chemours.pdf.
NCDEQ (North Carolina Department of Environmental Quality). 2018a. 60-Day Notice of Intent
to Modify Air Quality Permit No. 03735T43: The Chemours Company, Fayetteville
Works. NCDEQ. Accessed June 2018.
https://files.nc.gov/ncdeq/GenX/201S \pril em qui Jj \0 J_l _ signe
d.pdf.
NCDEQ (North Carolina Department of Environmental Quality) 2<)|xb. Groundwater and
Rainwater Sample Testing Results. KCDI.Q Accessed \lav 2< > IS
https://files.nc.gov/ncdeq/GenX/Rair !0and%20Gr ';r%2QT est%20Results
%2020180405.pdf.
National Research Council. 1983. Risk . \ssessmeni in the leileral Government: Managing the
Process. National Research Council National Acadeim Press, Washington, DC.
Accessed May 2018. httgs id/ ;r/l.
OECD (Organization for Economic Cooperation and l)e\ elopmenl). 1998. Test No. 408:
Repeated Dose 90-I)ay Oral Toxicity Study in Rodents In OECD Test Guidelines for the
Testing of Chemicals, Section 4, No 4<)X OIX'I) Publishing, Paris. Accessed May 2018.
http://dx.doi.ore/1 u, i. /87/9789264070707-en.
OECD (Organization for Economic Cooperation and Development). 2001a. Test No. 121:
r.siumiiion <>J ihe . hlsorpnon ('oejficient (Kf>c) on Soil and on Sewage Sludge using High
I'erjormance I n/md Chromatography (HPI.C). In OECD Test Guidelines for the Testing
of Chemicals, Section 1, No. 121. OECD Publishing, Paris. Accessed May 2018.
http://dx.doi.Org/l0.1787/978°2n iOo^iFj-em.
OECD (Organization for Economic Cooperation and Development). 2001b. Test No. 414:
Prenatal / development Toxicity Study. In OECD Test Guidelines for the Testing of
Chemicals, Section 4, No. 414. OECD Publishing, Paris. Accessed May 2018.
http://dx.doi.or; "" 1 ""7/9789264070820-en.
OECD (Organization for Economic Cooperation and Development). 2008. Test No. 407:
Repeated Dose 28-day Oral Toxicity Study in Rodents. In OECD Test Guidelines for the
Testing of Chemicals, Section 4, No. 407. OECD Publishing, Paris. Accessed May 2018.
http://dx.doi.on 87/9789264070684-en.
OECD (Organization for Economic Cooperation and Development). 2009. Test No. 453:
Combined Chronic Toxicity/Carcinogenicity Studies. In OECD Test Guidelines for the
Testing of Chemicals, Section 4, No. 453. OECD Publishing, Paris. Accessed May 2018.
http://dx.doi.orp. 1" l 87/978? l. _'3-en.
71
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
OECD (Organization for Economic Cooperation and Development). 2010. Test No. 429: Skin
Sensitisation: Local Lymph Node Assay. In OECD Test Guidelines for the Testing of
Chemicals, Section 4, No. 429. OECD Publishing, Paris. Accessed May 2018.
http://dx.doi.on >0 I 87/97892640' I 100 en.
OECD (Organization for Economic Cooperation and Development). 2016. Test No. 421:
Reproduction/Developmental Toxicity Screening Test. In OECD Test Guidelines for the
Testing of Chemicals, Section 4. OECD Publishing, Paris. Accessed May 2018.
http://dx.doi.on 87/9789264264380-en.
OECD (Organization for Economic Cooperation and De\ clopmcnl). 2018. OECD Guidelines for
the Testing of Chemicals. OECD. Accessed Max 2<)|S
http://www.oecd.org/chemicalsafety/testing/oec brthetestingofchemicals.htm.
Pan, Y., H. Zhang, Q. Cui, N. Sheng, L.W.Y. Yeung. Y Gno. Y Sun. and J. Dai. 2017. First
report of occurrence and bioaccumulalion of hexalluoropropylene oxide trimer acid: An
emerging concern. Environmental Science and Technology 51(17) 9553-9560.
Pan, Y., H. Zhang, Q. Cui, N. Sheng, L.W.Y. Yainu. Y Sun. Y. Guo, and .1 Dili 2018.
Worldwide distribution of novel perfluoroclhcr caihowlic and sulfonic acids in surface
water. Environmental Science and /echnology 52 7(->21 7629.
Rae, J.M., L. Craig, T.W. Slone, S.R. hnme. I. \V Buxton, nnd G L Kennedy. 2015. Evaluation
of chronic toxicity and carcinogen icily of ammonium 2.3.3.3-tetrafluoro-2-
(heptafluoropi'opow )- propanoalc in Sprimiie-Dnw lev nils toxicology Reports 2:939-
949.
Rushing, B., Q. Hu, J. I'liinklin. R \lc\lnheii. S Dnunio, C. Higgins, M. Strynar, and J. DeWitt.
2017. Evaluation of the immunomodulnloiy effects of 2,3,3,3-tetrafluoro-2-
(he|")liilluoi'o|")ropo\\ )-|")i"o|")iinonle in ("57151. (•> mice. ToxicologicalSciences 156(1): 179—
IS1)
Sheng. \ . R Cui, .1 \\ nnu. Y Guo, J. Wang, and J. Dai. 2018. Cytotoxicity of novel fluorinated
alleinnli\ es lo long-chain perfluoroalkyl substances to human liver cell line and their
binding enpnei l\ lo human liver fatty acid binding protein. Archives of Toxicology
92(1 ).35l) M*)
Song, X., R. Vestergien. Y Shi. .1 Huang, and Y. Cai. 2018. Emissions, Transport, and Fate of
Emerging Per- nnd Polylluoroalkyl Substances from One of the Major Fluoropolymer
Manufacturing Facilities in China. Environmental Science and Technology 52(17):9694-
9703.
Spahr, P.F., and J.T. Edsall. 1964. Amino Acid Composition of Human and Bovine Serum
Mercaptalbumins. Journal of Biological Chemistry 239(3):850-854.
Storch, J., and L. McDermott. 2009. Structural and functional analysis of fatty acid-binding
proteins. Journal of Lipid Research 50(Suppl):S126-S131.
72
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
StrynarM., S. Dagnino, R. McMahen, S. Liang, A. Lindstrom, E. Andersen, L. McMillian, M.
Thurman, I. Ferrer, and C. Ball. 2015. Identification of novel perfluoroalkyl ether
carboxylic acids (PFECAs) and sulfonic acids (PFESAs) in natural waters using accurate
mass time-of-flight mass spectrometry (TOFMS). Environmental Science and
Technology 49(19): 11622-11630.
Sun, M., E. Arevalo, M. Strynar, A. Lindstrom, M. Richardson, B. Kearns, A. Pickett, C. Smith,
and D.R.U. Knappe. 2016. Legacy and emerging perfluoroalkyl substances are important
drinking water contaminants in the Cape Fear River watershed of North Carolina.
Environmental Science & Technology Letters 3 415 419
Tietze, K.J. 2012. Review of Laboratory and Diagnostic Tests Chapter 5 in Clinical Skills for
Pharmacists. 3rd ed. pp. 86-122. Elsevier Mosby. Inc . Saint Louis, MO.
USEPA (U.S. Environmental Protection Agenc) ) II (innle/ines for Developmental Toxicity
Risk Assessment. USEP A. Federal Register. Dec 5, 1991, 5(->(234) 63798-63826.
Accessed May 2018.
fattps://ofm.i)ub. epa.gov/elms/elmscoi • ^—load .
USEPA (U.S. Environmental Protection Agency). 1')% (iiiidehnes for Reproductive Toxicity
Risk Assessment. USEPA. Federal Register, Oct 3 I 61(212):56274-56322.
Accessed May 2018. ht£Es://ww\y — eov/fdsv : i996-10-3 l/pdf/96-27473.pdf.
USEPA (U.S. Environmental Protection Agency). llWK (iuidelines for Neurotoxicity Risk
Assessment. I SI-PA. / eileral'Register. May 14 ll)l)S (•>3(l'3):26926-26954. Accessed
May 2018, -- ¦ ¦ ¦• ¦¦¦ "ov/: V-:. - - < .-i cuments/neuro tox.pdf.
USEPA (U.S. Environmental Protection Agency) 2< >< »2 A Review of the Reference Dose and
Reference Concentration I'rocesses I-PA > P-02/0002F. USEPA, Risk Assessment
I'oinni. Washington. DC Accessed May 2d IS
;um ents/ rfd-fin al .pdf.
USEPA (IS I-n\ iionmental Protection Agency). 2005a. Guidelines for Carcinogen Risk
. \ssessment EPA 63d |>-i )3 001IJ USEPA, Risk Assessment Forum, Washington, DC.
Accessed May 2018. 7www.epa.eov/risk/euidelines-carcinoeen-risk-assessment.
USEPA (U.S. En\ ironmental Protection Agency). 2005b. Supplemental Guidance for Assessing
Susceptibility from I .at ly-life Exposure to Carcinogens. EPA/630/R-03/003F. USEPA,
Risk Assessment I 'oinni, Washington, DC. Accessed May 2018.
https://www3.eua.gvjv/airtoxics/childrens supplement final.pdf.
USEPA (U.S. Environmental Protection Agency). 2006. A Framework for Assessing Health Risk
of Environmental Exposures to Children. EPA/600/R-05/093F. USEPA, Office of
Research and Development, National Center for Environmental Assessment,
Washington, DC. Accessed May 2018.
http://ofmpub.epa.eov/eims/eimscomm.eetfile7p download id=459047.
USEPA (U.S. Environmental Protection Agency). 2008. Certain New Chemicals; Receipt and
Status Information. USEPA. Federal Register, Aug. 8, 2008, 73(154):46263-46268.
73
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Accessed May 2018. https://www\ regulations. gov/doeiiment?D=EPA.~H.Q~QPPT-2008-
0598-0001.
USEPA (U.S. Environmental Protection Agency). 2009. Consent Order and Determinations
Supporting Consent Order for Premanufacture Notice Numbers: P-08-508 and P-08-509.
USEPA, Office of Pollution Prevention and Toxics, Washington, DC. Accessed May
2018.
https://chemview.epa.gov/chemview/proxy?filename=sanitized consent order p 08 050
8c.pdf.
USEPA (U.S. Environmental Protection Agency). 2011 a Exposure Factors Handbook: 2011
Edition. EPA/600/R-09/052F. USEPA, Office of Research and Development,
Washington, DC. Accessed May 2018.
http://ofmpiib.epa. gov/eim. s/eim scomm. : J~=52.2996.
USEPA (U.S. Environmental Protection Agency) 2<) I I h Recommended I Ise of Body Weight4
as the Default Method in Derivation <>J the (hat Reference / h>se I - PA/100/R11/0001.
USEPA, Office of the Science \dvisor. Risk \ssessment Forum, Washington, DC.
Accessed May 2018. https:/A* on/files/2013
09/docum ents/recom m en ded-
USEP A (U.S. Environmental Protection \genc>) 2d 12 I leach mark Dose Technical Guidance.
EPA/100/R-12/001. USEPA, Risk \ssessmcnt I oriini. Washington, DC. Accessed May
2018. https:" 'sii '
01 /docum ei
USEPA (U.S. Environmental Protection Agency) 2< > 13. Reaffirmation of the U.S. Environmental
Protection Agency 's I'J'J 5 I'o/icy on / .valiiaiing Health Risks to Children. USEPA,
Washington. DC \ccessed May 2d 18. . -.\ww.epa.eov/sites/production/files/2014-
P f "* ¦ fi . * Jjj?
USEPA (IS I ji\ ironmental Protection Agency). 2014a. Framework for Human Health Risk
Assessment to Inform Decision Making. EPA/100/R-14/001. USEPA, Office of the
Science Advisor, Risk Assessment Forum, Washington, DC. Accessed May 2018.
htttyq./, wyy-.epa.g;ov/sites/production/files/2 :uments/hhra-framework-final-
USEPA (U.S. Environmental Protection Agency). 2014b. Child-Specific Exposure Scenarios
Examples (Final Report). EPA/600/R-14-217F. USEPA, Office of Research and
Development, National Center for Environmental Assessment, Washington, DC.
Accessed May 2018.
http://ofmpub.epa.gov/eims/eimscomm.eetfile7p download id=520166.
USEPA (U.S. Environmental Protection Agency). 2014c. Guidance for Applying Quantitative
Data to Develop Data-Derived Extrapolation Factors for Interspecies and Intraspecies
Extrapolation. EPA/R-14/002F. USEPA, Office of the Science Advisor, Risk Assessment
Forum, Washington, DC. Accessed May 2018.
https://www.epa.gov/sites/production/files/2015-01/dociiments/ddef-final.pdf.
74
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
USEPA (U.S. Environmental Protection Agency). 2016a. Health Effects Support Document for
Perfluorooctanoic Acid (PFOA). EPA 822-R-16-003. USEPA, Office of Water, Health
and Ecological Criteria Division, Washington, DC. Accessed May 2018.
https://www.epa.gov/sit.es/production/files/2 /documents/pfoa hesd final-
plain.pdf.
USEPA (U.S. Environmental Protection Agency). 2016b. Health Effects Support Document for
Perfluorooctane Sulfonate (PFOS). EPA 822-R-16-002. USEPA, Office of Water, Health
and Ecological Criteria Division, Washington, DC. Accessed May 2018.
https://www.epa.gov/sites/production/files/201o ib/documents/pfos hesd final >08.pdf.
USEPA (U.S. Environmental Protection Agency). 2018a. HFPO-DA Results for Public and
Private Water Supplies in Vicinity of Chemonrs II ashingion Works Facility-February
2018. USEPA. Accessed May 2018. hftp"'"' ^roduetion/files/lOl 8-
04/documents/hfpo chem.ours wash wr • pling * f
USEPA (U. S. Environmental Protection Agenc\) 2<) 18b. Application of Systematic Review in
TSCA Risk Evaluations. EPA 740-Pl-S<)<) I I SEP A, Office of Chemical Safety and
Pollution Prevention, Washington, DC. Accessed June 2018.
https://www.epa. gov/ sites/pre on/files/2
06/docum ents/fin al appli cati c ¦ - - ., ¦: n; • ¦ ' • ;.df.
USFDA (U.S. Food and Drug Administration). 2<)<)2 < imdunce jor Industry: Immunotoxicology
Evaluation of Investigational New Drugs. U S Department of I lealth and Human
Services, USFDA, Center for Drugs Evaluation Research. Accessed May 2018.
https://www.fda.gov/dowi" * ' gutatoryinformation./guidan.
ces/ucm.07923 v .pdf.
Wang. J . X Wanu. N. Sheng, X. /lion. R Cui. 11 /hang, and J. Dai. 2016. RNA-sequencing
analysis ie\eals the hepatoloxic niechanisni of perfluoroalkyl alternatives, HFP02 and
I ll'l,()4. following exposure in mice. Journal of Applied Toxicology 3 7(4):43 6-444.
75
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Appendix A: Literature Search Strategy
This appendix presents the details of the literature search strategy used to identify primary, peer-
reviewed literature pertaining to HFPO dimer acid (Chemical Abstracts Service Registry
Number (CASRN) 13252-13-6) and/or its ammonium salt (CASRN 62037-80-3). The literature
search was conducted using the databases listed in Table A-l. The literature searches for these
GenX chemicals were conducted in July 2017 (acid) and January/February 2018 (ammonium
salt). The searches were conducted using CASRN, synonyms, and additional relevant search
strings (see Table A-2). Because the results of this core search were so limited, additional
databases were searched for physiochemical property information, health effects, toxicokinetics,
and mechanistic information. A list of these additional databases is provided in Table A-3 and
Table A-4. Combined, these database searches returned 27 studies lor HFPO dimer acid and
HFPO dimer acid ammonium salt. The available data lor (ien.X chemicals comes primarily from
studies submitted under Toxic Substances Control Act (TSCA) These studies were combined
with the results of the search of the publicly a\ ailahlc peer-reviewed literature for evaluation for
relevance to the assessment. Potential relevance was hased primarily on a title and abstract
screen. The inclusion/exclusion criteria applied to the literature searches conducted are presented
in Table A-5.
Table A-l. Summary of Core l);il;ihuso Search Results
Search Date
PiihMcd
wos
Toxline
TSCATS via Toxline/NLM
11 IPO dimer acid (I3252-I3-(.| < III.KO pro.jcel id:
7/24/17
i:
0
0
HFPO dimer acid ammonium sail ((i2O3"7-80-3) (HERO project id: 2683)
1/18 and 2/18
<>
i:
0
0
A-l
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
NOVEMBER 2018
Table A-2. Database Search Strings
HFPO Dimer Acid (13252-13-6)
lll l'O Dimer Acid Ammonium Salt (CASRN 62037-80-3)
Pubmed
13252-13-6[rn] OR "2,3,3,3-Tetrafluoro-2-
(heptafluoropropoxy)propionic acid"[tw] OR "2,3,3,3-tetrafluoro-2-
(l,l,2,2,3,3,3-heptafluoropropoxy)-Propanoic acid"[tw] OR
"Perfluoro(2-methyl-3-oxahexanoate) "[tw] OR "Propanoic acid.
2,3,3,3-tetrafluoro-2-(l,l,2,2,3,3,3-heptafluoropropoxy)- "[tw | Ok
"Perfluorinated aliphatic carboxylic acid"[tw] OR "Pcrfluoroi 2-
methyl-3-oxahexanoic) acid"[tw] OR " 2.3.3.3 -1ct rafl uo ro - 2 -
(l,l,2,2,3,3,3-heptafluoropropoxy)propanoic acid"[tw] OR "2. V \3-
tetrafluoro-2-(heptafluoropropoxy)propanoic acid"[tw] OR "perilnoin-
2-(propyloxy)propionic acid"[l\\ | OR "pcrfluoro-2-mcthyl-3-
oxahexanoic acid"[tw] OR "pcrfluoro-2-propoxypropanoic acid"|lw|
OR "perfluoro-2-propoxypropionic acid"|lw| OR "perfluoro-a-
propoxypropionic acid"[tw| OR "propanoic acid. 2.3.3.3-lclrafluoro-2-
(heptafluoropropoxy)-"[tw| OR "propionic acid. 2 3 3 3-tclrafluoro-2-
(heptafluoropropoxy)-"[tw| OR (GcnX AND (Iluoroca.rbon*|lw| OR
fluorotelomer*[tw] ORpolvnuoro*|lw| OR pcrriuoro-*|l\\ | OR
perfluoroa*[tw] ORperfluorob*|tw| OR pcrfluoroc*|tw | ()R
perfluorod*[tw] ORperfluoroc*|lw| OR pcrfluoroh*|lw | ()R
perfluoron*[tw] ORperfluoroo*|lw| OR pcrfluorop:;:|l\\ | ()R
perfluoros*[tw] ORperfluorou*|lw| OR pcrfluorinak'd|l\\ | ()R
nuorinatcd|t\\ |)) OR (("2 3 3 3-Tclrafluoro-''-
(heptafluoropi'iipii\y)propionic"|lw| OR "2. V V "-leti;ifliKni>-2-
(l,l,2,2,3,3.3-licpianuoropropoxy)-Propaiuiic"|l\\ | ()R "IVrllimniKilod
aliphatic carbn\\ 1 ic"|LwJ OR "Pcrfluoro(2-mclhyl-3-
oxahexanoici"|l\\ | OR "2 3 3 3-lclrafluoro-2-( 1.1.2.2.3.3,3-
heptafluoropropow ipropanoic"|lw| OR "2 3 3 3-lctrafluoro-2-
(lieptaflnoropropo\\ )propanoic"|l\v| OR "pcrriuoro-2-
(propyloxy)propionic"|lw| OR "pcrfluoro-2-mclhyl-3-
oxahexanoic"[tw] OR "pcrnuoro-2-propoxypropanoic"[tw] OR
"perfluoro-2-propo\ypropiniiic"|l\\ | ()R "perriuoro-a-
propoxypropionic"LtwJ) A\l) iacid|l\\ | OR acids|lw|))
«.:tr,--xt)-3[rn] OR "62037-80-3"[tw] OR "Ammonium 2,3,3,3-
k,li'alliKH'o-2-(heptafluoropropoxy)propanoate"[tw] OR "Propanoic
acid. 2. \ i.3-tetrafluoro-2-(l,l,2,2,3,3,3-heptafluoropropoxy)-,
amiiKiiiiiim salt"[tw] OR "Perfluorinated aliphatic carboxylic acid,
ammonium sall"|lw] OR "2,3,3,3-Tetrafluoro-2-(l,l,2,2,3,3,3-
hcplariuoropi'opo\y)propanoic acid, ammonium salt"[tw] OR
"Ammonium 2-(peifliioropropoxy)perfluoropropionate"[tw] OR
"Ammonium PcrHiioi'o(2-incthvl-3-o\ahcxanoatc)"|t\\ | OR
" \nimoniumpcrlluoi'n(2-methyl-3-oxahexanoic) acid"[tw] OR
" \minoniuinperfluoi'ii-2-incthyl-3-oxahexanoate"[tw] OR "FRD-
l>u:"|iw| OR "GenX-H3N"[tw] OR "HFPO-DA"[tw] OR "Propanoic
acid. 2 3 3 3-tetrafluoro-2-(heptafluoropropoxy)-, ammonium salt"[tw]
OR "Undccafluoro-2-methyl-3-oxahexanoic acid"[tw] OR ((GenX[tw]
AND (fluorocarbon*[tw] OR fluorotelomer*[tw] ORpolyfluoro*[tw]
OR pcrHi in in-* [tw] ORperfluoroa*[tw] ORperfluorob*[tw] OR
pcrfluoriic:;:|lw] ORperfluorod*[tw] ORperfluoroe*[tw] OR
pcrfluoroh*[tw] ORperfluoron*[tw] ORperfluoroo*[tw] OR
pcrfluorop*[tw] ORperfluoros*[tw] ORperfluorou*[tw] OR
pcrfluorinaled[tw] OR fluorinated[tw])) OR (("Undecafluoro-2-
mclhyl-3-o.\ahexanoic"[tw] OR "Ammonium perfluoro(2-methyl-3-
o.\ahc.\anoic)"[tw] OR "2,3,3,3-Tetrafluoro-2-(l,l,2,2,3,3,3-
heptafluoropropoxy)"[tw] OR "Perfluorinated aliphatic
carboxylic" [tw]) AND (salt[tw] OR salts[tw] OR acid[tw] OR
acids[tw])))) OR (((Undecafluoro AND oxahexanoic) OR (Ammonium
AND perfluoro AND oxahexanoic) OR (Tetrafluoro AND
heptafluoropropoxy) OR "Perfluorinated aliphatic carboxylic "[tw] OR
"Perfluorinated aliphatic carboxylic "[tw]) AND (salt[tw] OR salts [tw]
OR acid[tw] OR acids[tw]))
A-2
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
NOVEMBER 2018
HFPO Dimer Acid (13252-13-6)
HFPO Dimer Acid Ammonium Salt (CASRN 62037-80-3)
wos
TS="2,3,3,3-Tetrafluoro-2-(heptafluoropropoxy)propionic acid" OR
TS="2,3,3,3-tetrafluoro-2-(l,l,2,2,3,3,3-heptafluoropropoxy)-
Propanoic acid" OR TS="Perfluoro(2-methyl-3-oxahexanoate)" ()R
TS="Propanoic acid, 2,3,3,3-tetrafluoro-2-(l,l,2,2,3,3,3-
heptafluoropropoxy)-" OR TS="Perfluorinated aliphatic carboxylic
acid" OR TS="Perfluoro(2-methyl-3-oxahexanoic) acid" OR
TS="2,3,3,3-tetrafluoro-2-(l,l,2,2,3,3,3-heptafluoropropox} ipropanoic
acid" OR TS="2,3,3,3-tetrafluoro-2-(heptafluoropropoxy)propaiioic
acid" OR TS="perfluoro-2-(propyloxy)propionic acid" OR
TS="perfluoro-2-methyl-3-oxahexanoic acid" OR TS=" peril noro-2-
propoxypropanoic acid" OR TS="pcrnuoro-2-propoxy propionic acid"
OR TS="perfluoro-a-propoxypropionic acid" OR TS="prop;inoie acid.
2,3,3,3-tetrafluoro-2-(heptafluoropropoxy)-" OR TS="propionic acid.
2,3,3,3-tetrafluoro-2-(heptafluoropropoxy)-" OR (TS="Gcn\'" \M)
TS=(fluorocarbon* OR fluorotelomer* OR polyfluoro* OR perfluoro-
* OR perfluoroa* OR perfluorob* OR perfluoroc* OR perfluorod* OR
perfluoroe* OR perfluoroh* OR perfluoron* OR perfluoroo* OR
perfluorop* OR perfluoros* OR perfluorou* OR pcrfluorinalcd OR
fluorinated ORPFAS OR PFOS OR PFOA)) OR ((TS="2.3.3.3-
Tetrafluoro-2-(heptafluoropropoxv)propionic" OR TS="2 3 3 3-
tetrafluoro-2-(l,l,2,2,3,3,3-hcplanuoropropo.\y)-Propanoic" OR
TS="Perfluorinated aliphatic carboxylic" OR TS="Pcrfluoro(2-melhvl-
3-oxahexanoic)" OR TS="2 3 3 3-iciranuoro-2-( 1.1.2.2.3.3.3-
heptafluordpinpnw ipinpaiKiic" ORTS="2 3 3 3-lelrafluoro-2-
(heptafluoropropow ipropaiioic" OR TS= "pcrfluoro-2-
(propylo.w ipropionic" OR TS "pernnoro-2-nielhyl-3-o\ahc\anoic"
OR TS="pcrHiioro-2-propo\yprop;inoic" OR 1 S^"pcrlluoro-2-
propoxypnipionic" OR TS="perriiini\>-»-pi\>po\vpi\>piniiic") AND
TS=(acid OR acids n
T S=(" Ammo nium 2,3,3,3-tetrafluo ro-2-
rheptafluoropropoxy)propanoate" OR "Propanoic acid, 2,3,3,3-
lcir;illiioro-2-(l. 1.2.2,3.3.3-hcptafluoropropoxy)-. ammonium salt" OR
"IVi lliKH inated aliphatic carboxylic acid, ammonium salt" OR
"2. V v ^ - Ic t ra fl no ro-2-(1.1.2,2.3.3.3 -lie pta fl no ro p ro po xv) p ro pa no i c
acid, ammonium salt" OR "Ammonium 2-
(perTlikn\ipi'opoxy)perfluoropropionate" OR "Ammonium Perfluoro(2-
mclhyl-3-o.\alic.\anoate)" OR "Ammonium perfluoro(2-methyl-3-
oxaliexanoic) acid" OR "Ammonium perfluoro-2-methyl-3-
oxahexanoalc" OR "FRD-902" OR "GenX-H3N" OR "HFPO-DA" OR
"Propanoic acid. \3-lctrafluoro-2-(heptafluoropropoxy)-,
ammonium salt" OR "I ndccaflnoro-2-nicthy 1-3-oxaliexanoic acid")
()R i(TS=GcnX AND ( I S =(lluorocarbon* OR fluorotelomer* OR
polvfluoro* OR perfluoro-* OR perfluoroa* OR perfluorob* OR
perfluoroc* OR perfluorod* OR perfluoroe* OR perfluoroh* OR
perfluoron* OR perfluoroo* OR perfluorop* OR perfluoros* OR
perfluorou* OR perfluorinated OR fluorinated)))) OR
((TS=("Undccafluoro-2-methyl-3-oxahexanoic" OR "Ammonium
perfluoro(2-inethyl-3-oxahexanoic)" OR "2,3,3,3-Tetrafluoro-2-
(1,1,2,2,3,3,3-heptafluoropropoxy)" OR "Perfluorinated aliphatic
carboxylic" OR "Perfluorinated aliphatic carboxylic")) AND (TS=(salt
OR salts OR acid OR acids)))
Timcspan: All years. Indexes: SCI-EXPANDED, CPCI-S, CPCI-
SSH. BKCI-S, BKCI-SSH, CCR-EXPANDED, IC.
A-3
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
NOVEMBER 2018
HFPO Dimer Acid (13252-13-6)
HFPO Dimer Acid Ammonium Salt (CASRN 62037-80-3)
Toxline
(13252-13-6[rn] OR "2,3,3,3-Tetrafluoro-2-
(heptafluoropropoxy)propionic acid" OR "2,3,3,3-tetrafluoro-2-
(l,l,2,2,3,3,3-heptafluoropropoxy)-Propanoic acid" OR "Perfluoro(2-
methyl-3-oxahexanoate)" OR "Propanoic acid, 2,3,3,3-tetrafluoro-2-
(1,1,2,2,3,3,3-heptafluoropropoxy)-" OR "Perfluorinated aliphatic
carboxylic acid" OR "Perfluoro(2-methyl-3-oxahexanoic) acid" OR
"2,3,3,3 -tetrafluoro-2-( 1,1,2,2,3,3,3 -heptafluoropropoxy)propanoie
acid" OR "2,3,3,3-tetrafluoro-2-(heptafluoropropoxy)propanoic acid"
OR "perfluoro-2-(propyloxy)propionic acid" OR "perfluoro-2-niclhyl-
3-oxahexanoic acid" OR "perfluoro-2-propoxypropanoic acid" <)k
"perfluoro-2-propoxypropionic acid" OR "perfluoro-a-
propoxypropionic acid" OR "propanoic acid, 2,3,3,3-tetrallunm-2-
(heptafluoropropoxy)-" OR "propionic acid, 2 3 3 3 -tctru flik> i\i-
(heptafluoropropoxy)-" OR (GenX AND (fluorocarhmr OR
fluorotelomer* OR polyfluoro* OR pcrfluoro* OR peri lnori lintcd ()R
fluorinated OR PFAS OR PFOS OR PFOA)) OR < l-3-oxahcxaiKiici" ()R
"2,3,3,3-tetrafluoro-2-(l,l. \2. V \"-hepialluoropropoxv ipmpaiinic"
OR "2,3.3.3-tctrafluoro-2-(licpi;i 11 uoi'opiopo.w ipropanoic" ()R
"perfluoro-2-(propyloxy)propionic" OR "pci'lliioro-2-mclh\ 1-"-
oxahexanoic" OR "pcrfluoro-2-propoxyprop;11k>ic" OR "pciTlnoio-2-
propoxypropKinic" OR "peil'likiro-a-pnipow propionic") AND (acid
OR acids))) A\l) K anenpl |(.>ru| OR biosis |oiu| OR cis |org| OR dart
[org] ORpubdart |org| OR cmic |org| ()R epidem |org| OR fedrip
[org] ORhccp |org| OR hmtc |org| OR ipa |org| OR riskline [org] OR
mtgabs [org| OR niosh |org| OR nlis |oiu| OR peslab |org] ORppbib
[org]) AND NOT piibmcd"|org| AND \() \ pubdarl |org])
(62037-80-3[rn] OR "Ammonium 2,3,3,3-tetrafluoro-2-
(heptafluoropropoxy)propanoate" OR "Propanoic acid, 2,3,3,3-
lclraniioro-2-(l,l,2,2,3,3,3-heptafluoropropoxy)-, ammonium salt" OR
"Pcrriuorinated aliphatic carboxylic acid, ammonium salt" OR
"2 3 3 3-Tetrafluoro-2-(l,l,2,2,3,3,3-heptafluoropropoxy)propanoic
acid, ammonium salt" OR "Ammonium 2-
(pcrriuoi\ipi'iipoxy)perfluoropropionate" OR "Ammonium Perfluoro(2-
mclhyl-3-ii.\alic.\anoate)" OR "Ammonium perfluoro(2-methyl-3-
oxaliexanoic) acid" OR "Ammonium perfluoro-2-methyl-3-
oxahexanoatc" OR "FRD-902" OR "GenX-H3N" OR "HFPO-DA" OR
"Propanoic acid. 2. V \3-tctrafluoro-2-(heptafluoropropoxy)-,
ammonium salt" OR "Undccafluoro-2-methyl-3-oxahexanoic acid" OR
"(icn.X" OR (("Undccanuoro-2-methyl-3-oxahexanoic" OR
" \mmoniumperfluoro(2-mcthyl-3-oxahexanoic)" OR "2,3,3,3-
rcliafluoro-2-(l,l,2,2,3,3,3-heptafluoropropoxy)" OR "Perfluorinated
aliphalic carboxylic" OR "Perfluorinated aliphatic carboxylic") AND
(sail ()R salts OR acid OR acids))) AND ((aneupl [org] OR biosis
|org| OR cis [org] OR dart [org] OR pubdart [org] OR emic [org] OR
cpidcin |org| OR fedrip [org] OR heep [org] OR hmtc [org] OR ipa
|oi u| OR riskline [org] OR mtgabs [org] OR niosh [org] OR ntis [org]
()R peslab |org] OR ppbib [org]) AND NOT pubmed [org] AND NOT
pubdart [org])
TSCATS1
13252-13-6|rn| AND (TSCATS |orgJ)
62037-80-3 [rn] AND (TSCATS [org])
Notes'. PFAS = per- and polyfluoroalkyl substances; PFOA = peril norooctanoic acid; PFOS = perfluorooctane sulfonate.
A-4
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Table A-3. Processes Used to Augment the Search of Core Databases for HFPO Dimer Acid (13252-13-6)
System Used
Selected Key Reference(s) or Sources
References
Identified
TSCATS3
TSCA Test Submissions 2.0; website now retired (httDs://vosemite. ts/eDatscat8.nsf/ReDortSearch?ODenForm)
0
Chemical Data Access Tool (CDAT): website now retired (hu^s.//; odd! chemical search/)
0
ChemVie w Chtt ds ://i ava. eoa. gov/che mview)
0
Resources searched for
physiochemical
property information
Agency for Toxic Substances and Disease Registry (ATSDR) ( xlc.gov/)
Australian National Industrial Chemicals Notification and \ssessment Sthenic (\ICNAS)
(httDs://www. nicnas.gov.au/chemical-informatiom
CAMEO Chemicals dittes://cameocliemicals.noa; i
Canada DSL List (http://webnet.oecd.org/CCR'WEB/S asox)
Chemical Risk Information Platform (CHRIP) (http://- :hrip. l/svstemTop)
ChemlDolus (httDs://chem. nlm.nih.gov/chem' " " ")
ChemSmder (httD://www.chemsDider.com/)
CRC Handbook of Chemistry and Physics
rtittD://hbcDonline.com/:faees/conteiits/Co " " ' 875156F724E0E945D3A6D0454891)
EC HA Information on Chemicals (https:/ i
eChemPortal Chtt ds ://www. ecte mtx) rta I. o i g/e i
Hazardous Substances Data Bank (HSDB) !•: ' ,tmlgen?HSDB)
HSNO Chemical Classification and Information Dal.ih.isc (C C ll)i npdalcd link1'
ChttDs://www.eDa. govt.r i/ctemic auun-anu-uuui niation-database-ccid/)
IARC Monographs ditti jrg/page! )
Integrated Risk Information s\ sicm (IRIS) (httos://wv igov/iris)
J-Check i sa: i..actkjM;i«niest locale=en)
Kirk-Olhmci 1 iK_\dopedia ol Chemii_al 1 eJinoloux updated linkb
(httBs:j gitexoin/dgi/t i >61.)
NIEHS i s.// w m w.nielis.nih.gov/")
OSHA (kxnmlionnl Chemical Database d' " /.osha.gov/chemicaldata/)
PubChem i f/Diibctem.ncbi.nlm.nih.gov/search/index.html)
SRC Fate Poinicrs i ;tD://esc.svrres.com/fateDointer/search.asD)
Ullmann's Enc\ elopedia updated link1' d-:;:' sibrarv.wilev.com/doi/book/10.1002/14356007)
USEPA ACToJ< i ctor.eoa.gov/actor/home.xhtml)
USEPA CDAT; wehsiie now reined ( -hw 'a.eDa.sov/oDDt chemical searcli/)
USEPA Chemistrv Dashboaid i -v.-.eDa.gov/dashboard/)
USEPA ChemView dittos:// - sv/che mview)
USEPA Substance Registry Seniles (SRS)
Chtt ds ://ofniDub .eoa. gov/so r interne t/registrv/substreg/searchandretrieve/substancesearch/search. do)
Web-based search for chemical manufacturer documents
3
A-5
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
NOVEMBER 2018
System Used
Selected Key Reference(s) or Sources
References
Identified
Resources searched for
health effects,
toxicokinetics, and
mechanistic
information
ATSDR (htt p ://www. at sdr. cdc. gov/substa nces/i ndex. asp)
CalEPA OEHHA diftD://www.oelilia.ca.gov/r!sk.lifnin
CPSC ChttD://www.cDsc.eov)
ECHA ChUD://echa.euroDa.ei!/information-on-chemicalsN)
eChemPortal0 ditto://www.eclieiiidortal.org/ectenroorta 1/)
EFSA Europe ('http://www.efsa.enropa.en/')
Environment Canada (http://www.ec.ec.ca/default.asD it )
European Union Risk Assessment Reports (https://ec.< n/t i
Federal Docket (http://www.regulations.gov")
Health Canada (tittm^^ . ii)
IARC (htto://monographs.iarc.fr/ENG/ClassificatJ )
ITER (littp://www. tera.org/ite rA
Japan Existing Chemical Data Base ( ijg/ /SearchPageEK _ )
NICNAS (http://www.nicrcis.gov.au/ .on)
NIEHS (http://www.nielis.nili.gov/)
NTP (http://ntpsearch.niehs.nih.gov/] )
OEHHA Toxicity Criteria Database i.)
USEPA NSCEP (Ll!'"'' m)
USFDA (http://ww )
WHO (tifto://www: i/e )
0
a Only relevant TSCATS studies from these interfaces w ere added to the 111 iRO project page.
b The URL has been updated (as listed here) since the literature search; during the search, a previous URL was used.
c eChemPortal includes the following databases: ACToR. AGRITOX. CCR. CCR DATA, CESAR, CHRIP, ECHA CHEM, EnviChem, ESIS, GHS-J, HPVIS, HSDB, HSNO
CCID, INCHEM, J-CHECK, JECI)l.3. NICNAS 1'] !C. ()] !CD-I IPV. ()1 '.CD SIDS IUCI.ID. SIDS UNEP, UK CCRMP Outputs, EPA-IRIS, and EPA-SRS.
A-6
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Table A-4. Processes Used to Augment the Search of Core Databases for HFPO Dimer Acid Ammonium Salt (62037-80-3)
System Used
Selected Key Reference(s) or Sources
References
Identified
TSCATS3
Chemical Data Access Tool (CDAT); website now retired (itiri gov/oDDt chemical search/)
ChemView (httDs://iava.eDa.gov/chemview)
Resources searched for
physiochemical property
information
ATSDR (httDs://www.atsdr. cdc.gov/)
CAMEO Chemicals (lit tos ://ca meoche mica Is. noaa. gov A
Canada DSL List (htto://webnet.oecd.org/CCRWEB/Search.asDx)
ChcinlDolus (httDs://chem. nlm.nih.gov/chemidDlus/)
CRC Handbook of Chemistry and Physics
(httD://hbcDonline.com/faces/contents/ContentsSearch.xhtml:isessionid .. ,, '.4E0E945D3A6D0454891)
ECHA Information on Chemicals (littos://echa.euromeu/)
eChemPortal dittos://www.eclieitidortal.org/echemoortal/index.action)
Hazardous Substances Data Bank (HSDB) (httDs://toxnet.Tilm. nih.gov/cgi-bin/sL.... ?HSDB)
HSNO Chemical Classification and Information Database (CCID) updated linkb
(httDs://www.eDa.govt.nz/database-search/chemical-classificat' ' information-database-ccid/)
1
IARC Monographs ditto://www. i nclieiiiors/oa ees/iarc.ht ml)
Integrated Risk Information System (IRIM < )
J-Check (htto://ww " ' D.io/icliecK " )
Kirk-Othmer Encyc loped la ol ( hemical Techuologs upd.iled link
NICNAS dittps://www.menas.gov.aii/cheinie tion)
NIEHS (https://www.niehs.nih.gov/)
OSHA Occupational Chemical Dalahaso ( : n-»v/clieniicaklata/)
PubChem ( :-S:;il)
SRC Fate Pointers C :; . ... ; j)
Ullmann's Encyclopedia updated linkb (li.irt-s. <•;. -.¦¦¦ ilev.com/doi/book/10.1002/14356007)
USEPA \CToR (https://actor.epa.gO' ]-..'me.xhtml)
USEPA CI) \T: website now retired (i-.¦¦¦¦¦¦ --/a.eDa.gov/oDDt chemical search/)
USEPA ( hemisirv Dashboard (httDs://coniDiox.eDa.gov/dashboard/)
USEPA ('liemYiew (htIds://iava.eoa.gov/chemview)
USEPA Substance Registry Services (SRS)
(httds://ofmoiib.eoa.gov/sor internet/re gistrv/substreg/searchandretrieve/substancesearch/search. do)
Web-based search fur chemical manufacturer documents
A-7
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
NOVEMBER 2018
System Used
Selected Key Reference(s) or Sources
References
Identified
Resources searched for
health effects,
toxicokinetics, and
mechanistic information
ATSDR (littD://www.atsdr. cdc.gov/substances/index.asD)
CalEPA - OEHHA (http://www.oehha.ca.gov/risk.html; hi ta.ca.gov/tcdb/index.asp)
CPSC (http://www.cpsc.gov)
ECHA ditt'D://eclia.eiiroDa.eii/informafion-on-clieniica )
eChemPortal0 Chttd://www.echemtx)rtal.org/ediemtx)i )
EFSA Europe
Environment Canada dittD://www.ec.ec.ca/di )
EPA-NSCEP ChttDs://www.eDa.gov/nsceDN)
European Union Risk Assessment Reports di ropa.eu/irc/en/t - 0
Federal Docket (1ittp://www.regulations.gov)
Google (Quick search onlv www.gooele.com)
Health Canada Chttps://www.canada.ca/en/liea 1th- i
IARC dittD://moiiograDlis.iarc.fr/ENG/CIassificat!on/i )
ITER (TERA database) (http://www.tera.org/iter/)
Japan Existing Chemical Data Base (JECDB) ( = lata/isD/SeareliPaeeENG.isp)
NICNAS (litto://www.nicnas.eov.au/clieiTiic.al-in-f )
NIEHS (http://www.niehs.nih.gov/)
NTP (liM;MEsear ' gov/hom )
USFDA (http://ww )
WHO (http://w t/e )
2
Notes:
a Only relevant TSCATS studies from these interlaces were added to the 111 ¦'R() project page.
b The URL has been updated (as listed here) since the literature search: during the search, a previous URL was used.
c eChemPortal includes the following databases: ACToR, AGRI'IX )X, CCR. CCR DATA, CESAR, CHRIP, ECHA CHEM, EnviChem, ESIS, GHS-J, HPVIS, HSDB, HSNO
CCID, INCHEM, J-CHECK, .IEC])B,NICNAS PEC. OECD-I IPV. OIX'I) SIDS IUCLII). SIDS UNEP, UKCCRMP Outputs, EPA-IRIS, andEPA-SRS.
A-8
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Table A-5. Inclusion-Exclusion Criteria for HFPO Dimer Acid and HFPO Dimer Acid Ammonium Salt
Inclusion Criteria
Exclusion Criteria
Population
• Humans
• Standard mammalian animal models, including rat, mouse, rabbit, guinea
pig, hamster, monkey, dog
• Alternative animal models in standard laboratory conditions < e u .
Xenopus, zebrafish, minipig)
• Human or animal cells, tissues, or organs (not whole animals i. hactena.
nonmammalian eukaryotes; other nonmammalian laboratory species
• Ecological species
Exposure
• Exposure is to HFPO dimer acid and/or its ammonium sail
• Exposure via oral, inhalation, dermal, intraperitoneal, oi" nilra\ eiious
injection routes
• Exposure is measured in air, dust, drinking water, diet, gavage or
injection vehicle, or via a biomarker of exposure (WVS levels in u hole
blood, serum, plasma, or breast milk)
• Exposure is via cells in culture or subcellular matrices
• Study population is not exposed to HFPO dimer acid and/or its
ammonium salt
• 1 Exposure is to a mixture only without evaluating HFPO dimer
acid and/or its ammonium salt individually
Outcome
• Studies that include a measure of one or more health elTecl ciidpoiuts.
including effects on reproduction, development. de\ clopnicutal
neurotoxicity, liver, thyroid, immune s\stem. ner\ ous s\ stem,
genotoxicity, and cancer
• In vivo and/or in vitro studies related to toxicity mechanisms,
physiological effects/adverse outcomes, and studies useful for elucidating
toxic modes of action
• Qualitative or t|ii;inlil;ili\e description of absorption. distribution,
metabolism, elimination, and toxicokiuclic and or to\icod> iiamic models
(e.g., PBPk. PI i 1 EC, PBTK/TI) i
• Studies addressiiiu' risks to infants, children, pregnant women,
occupational workers, the elderly, and any other susceptible or
differentially exposed populations
A-9
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
NOVEMBER 2018
Inclusion Criteria
Exclusion Criteria
Other
• Structure and physiochemical properties
• Reviews and regulatory documents
Not on topic, including3:
• Abstract only, inadequately reported abstract, or no abstract
a nd not considered further because study was not potentially
re leva nl
• 1 iiorcmcdialion, biodegradation, or chemical or physical
irealmcnl of HFPO dimer acid and/or its ammonium salt,
including evaluation of wastewater treatment technologies and
mclliods for remediation or contaminated water and soil
• Ecosystem effects, studies in ecological species that are not
relevant to lieallli effects in humans
• Studies of cn\ ironmental fate and transport of HFPO dimer
acid and/or lis ammonium salt compounds in environmental
media
• Vnalytical methods for detecting/measuring HFPO dimer acid
and/or its ammonium salt compounds in environmental media
and use in sample preparations and assays
• Si ikI ics describing the manufacture and use of HFPO dimer
acid and or its ammonium salt compounds
• \oi chemical-specific (studies that do not involve testing of
11I IJ() dimer acid and/or its ammonium salt compounds)
• Studies iliat describe measures of exposure to HFPO dimer acid
and/or its ammonium salt compounds without data on
associated health effects
Notes'. PBPK = physiologically based pharmacokinetic: Pl'AS = pre- and polvlluoroalkyl substances.
a Although these criteria were used lor the peer-reviewed literature, the current document describes data on environmental fate data submitted by DuPont (now the Chemours
Company). A subsequent targeted search lor bioconcentration and bioaccumulation data was also conducted. In addition, a summary of occurrence data is also described in the
current document to give context to the toxicity values.
A-10
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Appendix B: Systematic Review of DuPont/Chemours Submissions
Background
As mentioned in Section 3.3.2, most of the available data were submitted to the U.S.
Environmental Protection Agency (EPA) under the Toxic Substances Control Act (TSCA) for
hexafluoropropylene oxide (HFPO) dimer acid and/or ammonium salt. Submitted test data on
HFPO dimer acid and/or ammonium salt were available for numerous endpoints such as acute
toxicity, metabolism and toxicokinetics, genotoxicity, and systemic toxicity in mice and rats with
dosing durations of up to 2 years. Most of these submitted studies were conducted according to
Organization for Economic Cooperation and Development (OI '('I)) Test Guidelines (TGs)
and/or EPA Health Effects Test Guidelines for Pesticides and Toxic Substances, which "are
generally intended to meet testing requirements for human health impacts of chemical substances
under (the Federal Insecticide, Fungicide, and Rodenticide Act) I II R.\ and TSCA." The
majority of the studies adhered to the Good Laboratory Practices ((il V) principles, and full study
reports were submitted for Agency review either with the Pre-manufacluie \otice under TSCA
or subsequently as the result of a Consent Order associated with that TSC A new chemical
review.
The EPA Office of Pollution Prevention and Toxics (OPPT) reviewed studies for quality and
adherence to the guidelines when they were recei ved. However, for the purpose of developing
the toxicity values, EPA OPPT evaluated the human health TSCA guideline studies, including
toxicokinetic and repeated-dose studies of 28 days or longer and other studies identified in the
literature search (see section 3 3.1) according to the e\ aluation strategy described in this
appendix. The data quality criteria and scoring tables are provided in the subsequent pages for
the studies supporting the de\ clopment of toxicity \ allies lor HFPO dimer acid and dimer acid
ammonium salt. The data quality criteria used in this document may slightly differ from those
later published in the Applications of Systematic Re\iew in TSCA Risk Evaluations (2018;
https://www "ov/sites/
06/documen applic _ .pelf). These slight modifications are
not expected to change the o\erall confidence of the studies.
Strategy for Assessing the Quality of Data/Information
The strategy for assessing the quality of data/information sources is based on a structured
framework with predefined criteria for each type of data/information source. EPA OPPT
developed a numerical scoring system to inform the characterization of the data/information
sources during the data integration phase. The goal is to provide transparency and consistency to
the evaluation process.
The general structure of the evaluation strategy is composed of evaluation domains, metrics, and
criteria. Evaluation domains represent general categories of attributes that are evaluated in each
data/information source (e.g., test substance, test design). Each domain contains a unique set of
metrics, or sub-categories of attributes, intended to assess an aspect of the methodological
conduct of the data/information source. Each metric specifies criteria expressing the relevant
elements or conditions for assessing confidence that, along with professional judgement, will
guide the identification of study strengths and limitations/deficiencies.
B-l
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Reporting quality is an important aspect of a study that needs to be considered in the evaluation
process. The challenge, in many cases, is to distinguish a deficit in reporting from a problem in
the underlying methodological quality of the data/information source. The evaluation strategy
incorporates reporting criteria within the existing domains, rather than adding a separate
reporting domain as recommended in some evaluation tools/frameworks. Since reporting
contributes to the evaluation of each facet of the data source, EPA OPPT assesses reporting and
methodological quality simultaneously with the idea of untangling reporting from study conduct
while the reviewer is assessing a particular metric for each domain. Developing a reporting
checklist, guidance document, or a separate reporting quality domain may be possible in the near
future as EPA OPPT uses and optimizes the evaluation strategy for animal toxicity and in vitro
studies.
Data/information sources should also be evaluated for their rele\ ance or appropriateness to
support the risk assessment. Specifically, data/information sources should support the assessment
questions, analytical approaches, methods, models, and considerations that are laid out during
problem formulation. EPA OPPT uses a tiered approach to check for relevance starting at the
data search stage and continuing during the title, abstract and full text screening and evaluation
and integration stages. By design, the systematic re\ iew process uses a fit-for-purpose literature
search and relevance-driven eligibility criteria to end up e\ aluating the most relevant
data/information sources for the risk assessment The re\ iew ers also check for relevance while
assessing the quality of the data/information source and are asked to document4 any relevancy
issues during the evaluation process.
The evaluation straleuy in some cases refers to study guidelines along with professional
judgement as a helpful guidance in determining the adequacy or appropriateness of certain study
designs or analytical methods This should not be construed to imply that non-guideline studies
have lower confidence than guideline or (iLP studies. EPA OPPT will consider any and all
available. rele\ant data and information irrespecti\e of whether they were collected in
accordance with standardized methods (eg. Olid) TGs or GLP standards).
EPA OI'I'T will consider data and information from alternative test methods and strategies (or
new approach methodologies or \ AMs). as applicable and available. This is consistent with EPA
OPPT's Strategic Plan to Promote the Development and Implementation of Alternative Test
Methods to reduce, refine, or replace vertebrate animal testing (USEPA, 2018). The
data/information quality criteria may need to be optimized or new criteria may need to be
developed as part of evaluating and integrating NAMs in risk assessments.
B.l.l. Evaluation Method
Based on the strengths, limitations, and deficiencies of each data/information source, the
reviewer assigns a confidence level score of 1 (high confidence), 2 (medium confidence), 3 (low
confidence), or 4 (unacceptable) for each individual metric that is used to evaluate a particular
aspect of the methodological conduct of the data/information source. Although many metrics
have criteria for all four bins (i.e., High, Medium, Low, and Unacceptable), there are some
metrics with dichotomous or trichotomous criteria to fit better the nature of the criteria.
4 Relevancy issues will be documented in the reviewer's comments, when pertinent.
B-2
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
The confidence levels and corresponding scores at the metric level are defined as follows:
• High: No notable deficiencies or concerns are identified in the domain metric that are
likely to influence results (score of 1).
• Medium: Minor uncertainties or limitations are noted in the domain metric that are
unlikely to have a substantial impact on results (score of 2).
• Low: Deficiencies or concerns are noted in the domain metric that are likely to have a
substantial impact on results (score of 3).
• Unacceptable: Serious flaws are noted in the domain metric that consequently make the
data/information source unusable (score of 4).
• Not rated/applicable: Rating of this metric is not applicable to the data/information
source being evaluated (no score). Not rated/applicable will also be used in cases in
which studies cite a literature source for their lesi methodology instead of providing
detailed descriptions. In these circumstances. I -PA will score the metric as Not rated/not
applicable and capture the score in the re\ iew er's notes. If the data'information source is
not classified as "unacceptable" in the initial review, the cited literature source will be
reviewed during a subsequent evaluation step, and the metric \\ i 11 lx- rated at that time.
A numerical scoring method is used to convert the confidence level for each metric into the
overall quality level for the data/information source. The o\ erall study score is equated to an
overall quality level {High, Medium. or Low) using the level definitions and scoring scale shown
in Table B-l. The scoring scale was obtained by calculating the difference between the highest
possible score of 3 and the lowest possible score of 1 (i e . 3 I 2) and dividing into three
equal parts (2-^3 = 67) This results in a range of approximately 0.7 for each overall data
quality level, and this range was used to estimate the transition points (cut-off values) in the scale
between High and Medium scores and Medium and Low scores. These transition points between
the ranges of 1 and 3 were calculated as follows
• Cut-off values between lligli and Medium I + 0.67 = 1.67, rounded up to 1.7 (scores
lower than 1.7 will be assigned an o\ erall quality level of High)
• Cut-olV\ allies between Medium and loir 1.67 + 0.67 = 2.34, rounded up to 2.3 (scores
between I 7 and lower than 2 3 will be assigned an overall quality level of Medium)
A study is disqualified from further consideration if the confidence level of one or more metrics
is rated as Unacceptable (score of 4). EPA OPPT plans to use data with an overall quality level
of High, Medium, or / confidence, but does not plan to use data rated as Unacceptable. Data
or information from I naccepiable studies might be useful qualitatively and such use of
unacceptable studies may be done on a case-by-case basis.
Table B-l. Definition of Overall Quality Levels and Corresponding Quality Scores
(hersill
Oii;ilih l.c\cl
DiTinilion
<>\ Ol'illl
Qu;ili(> Score
High
No notable dcliciciicicb or conceriib are identified and die data therefore could
be used in the assessment with a high degree of confidence.
> 1 and < 1.7
Medium
Possible deficiencies or concerns are noted and the data therefore could be
used in the assessment with a medium degree of confidence.
> 1.7 and <2.3
B-3
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
(hersill
Oii;ilih l.c\cl
DiTinilion
<>\ Ol'illl
Qu;ili(> Score
Low
Deficiencies or concerns are noted and the data therefore could be used in the
assessment with a low degree of confidence.
> 2.3 and < 3
Unacceptable
Serious flaw(s) are identified and therefore, the data cannot be used for the
assessment.
4
After the overall score is applied to determine an overall quality level, professional judgment
may be used to adjust the quality level obtained by the weighted score calculation. The reviewer
must have a compelling reason to invoke the adjustment of the o\ erall score and written
justification must be provided. This approach has been used in oilier established tools such as the
ToxRTool (Toxicological data Reliability Assessment Tool) de\ eloped by the European
Commission ("https://eurl-ecvam.irc.ec.europa.eu/about ive-publications/toxrtool).
B.1.2. Documentation and Instructions for Reviewers
Data evaluation is conducted in Excel to track and record the evaluation for each
data/information source. Refer to Section B.5 for ilie data evaluation and scoring tables
documenting the evaluation.
A confidence level is assigned for each iele\ ant metric within each domain by following the
confidence level specifications pro\ ided in section B.2.2, along with professional judgment to
identify study strengths and limitations The assigned confidence level is indicated by placing a
score between 1 and 4 in the column labeled Metric Score In some cases, reference to study
guidelines (in addition lo professional judgement) might he helpful in determining the adequacy
or appropriateness of certain study designs or analytical methods. This should not be construed
to imply that non-guideline studies necessarily ha\ e lower confidence than guideline studies. If a
publication reports more than one study data type or endpoint, each study/datatype or endpoint
will be e\aluated separately and assigned a separate overall quality level.
Some metrics might not he applicable to all study types. If a metric is not applicable to the study
under re\ iew. zero, which is equi\ alent to NR (not rated), is placed in the Metric Score column
for this metric
After scoring of the individual metrics within each domain, the overall study Weighted Score is
calculated and assigned to the corresponding bin {High, Medium, Low, or Unacceptable).
In the Comments column, the reviewer documents concerns, uncertainties, strengths, limitations,
deficiencies, and any additional comments observed for each metric, when necessary. For
instance, EPA might not always provide a comment for a metric that has been categorized as
High. However, a reviewer is strongly encouraged to provide a comment for metrics categorized
Medium or Low to improve transparency. The reviewer also records any relevance issues with
the data/information source (e.g., study is not useful to answer assessment questions).
B-4
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
B.1.3. Important Caveats about the Evaluation Method
The following is a discussion of important caveats for the data quality evaluation method:
• Although specifications for the data quality evaluation metrics have been developed,
professional judgment is required to assess the metrics.
• Data evaluation is a qualitative assessment of confidence in a study or dataset. In order to
provide consistency and transparency to the evaluation process, a scoring system is being
applied to ascertain a qualitative rating. Scores will be used for the purpose of assigning
the confidence level rating of High, Medium, Low, or Unacceptable, and inform the
characterization of data/information sources during the data integration phase. The
system is not intended to imply precision and/or accuracy of the scoring results.
• Every study or dataset is unique and therefore the indi\ idnal metrics and domains may
have various degrees of importance (e.g., more or less important). The weighting
approach might need to be adjusted as EPA OPPT tests the e\ aluation method with
different types of studies.
• The metrics developed are intended to be indicators of data quality They were selected
because they are generally considered common and important for a broad range of
studies. Other metrics not listed may also be important and added if necessary. Also,
there is the possibility of de\ iatinu from the calculated o\ erall confidence level score in
case the metric criteria are unable to capture professional judgement. A reviewer must
provide a justification for the score adjustment to ensure transparency for the decision.
Data Quality Criteria for Studies on Animal and In Vitro Toxicity
B.1.4. Types of Data Sources
The data quality will be e\ aluated lor a \ ariety of animal and in vitro toxicity studies. Table B-2
pro\ ides examples of types of studies falling into these two broad categories. EPA OPPT may
tailor the criteria to capture the inherent characteristics of particular studies that are not captured
in the current criteria (eg. optimization of criteria to e\ aluate the quality of new approach
methodologies or VWIs)
Table IJ-2. Typos of Animal ami In I itro Toxicity Data
Diilii Csilcgon
Tjpe of Diilii Sources
Animal Toxicity
()ial. dermal, and inhalation routes: lethality, irritation, sensitization, reproduction, fertility.
de\elnpnienial. neurotoxicity, carcinogenicity, systemic toxicity, metabolism,
pharmacokinetics, absorption, immunotoxicity, genotoxicity, mutagenicity, endocrine
disruption
In Vitro Toxicity
Studies
Irritation, corrosion, sensitization, genotoxicity, dermal absorption, phototoxicity, ligand
binding, steroidogenesis, developmental, organ toxicity, mechanisms, high throughput,
immunotoxicity
Mechanistic evidence is highly heterogeneous and may come from human, animal, or in vitro
toxicity studies. Mechanistic evidence may provide support for biological plausibility and help
explain differences in tissue sensitivity, species, gender, life-stage, or other factors (USEPA,
2006). Although highly preferred, the availability of a fully elucidated mode of action (MOA) or
B-5
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
adverse outcome pathway (AOP) is not required to conduct the human health hazard assessment
for a given chemical.
EPA OPPT plans to prioritize the evaluation of mechanistic evidence instead of evaluating all of
the identified evidence upfront. This approach has the advantage of allowing for a focused
review of those mechanistic studies that are most relevant to the hazards under evaluation.
B.1.5. Data Quality Evaluation Domains
The methods for evaluation of study quality were developed after review of selected references
describing existing study quality and risk of bias evaluation tools for toxicity studies (EC, 2018;
Cooper et al., 2016; Lynch et al., 2016; Moermond et ill . 2<> I Samuel et al., 2016; NTP,
2015a; Hooijmans et al., 2014; Koustas et al., 2014; kushman el al , 2013; Hartling et al., 2012;
Hooijmans et al., 2010). These publications, coupled with professional judgment and experience,
informed the identification of domains and metrics for consideration in the evaluation and
scoring of study quality.
The data quality of animal toxicity studies and /// vitro toxicity studies are e\ aluated by assessing
the following seven domains: Test Substance, Test Design, Exposure Characterization, Test
Organisms/Test Model, Outcome Assessment, Confounding Variable Control, and Data
Presentation and Analysis The dala quality within each domain is evaluated by assessing unique
metrics that pertain to each domain The domains are defined in Table B-3 and further
information on evaluation metrics is |iro\ ided in section 1} 3 3
Table B-3. Data Evaluation Domains and Definitions
l-'.\iiliiiiiion Doniiiin
IkTinilion
Test Substance
Metrics in this domain arc used in e\ aluaie u liellier ilie iiiHh iikiikhi pan ided mi ilie siud\
provides ;i reliable coiilirnialiou lh;il ilie lesi suhsiauce used mi ;i siud> h;is ilie s;ime (or
siifficieiuls similar) ideniiis. purii\. ;ind properties ;is ilie suhsiauce of iiiicresi
Tom Dcsium
Metrics mi i Ins domain arc used lo evaluate whether the experimental design enables the
siud> lo disiiiiguish the effect of exposure from other factors. This domain includes
metrics related lo the use of control groups and randomization in allocation to ensure that
the effect of exposure is isolated.
Exposure
Characterization
Metrics in this domain arc used to assess the validity and reliability of methods used to
measure or characterize exposure. These metrics evaluate whether exposure to the test
substance was characterized using a method(s) that provides valid and reliable results,
\\ liellier i lie exposure remained consistent over the duration of the experiment, and
u liellier llie exposure levels were appropriate to the outcome of interest.
Test Organisms/Test
Model
These nieirics are used to assess the appropriateness of the population or organism(s),
group sizes used in the study (i.e., number of organisms and/or number of replicates per
exposure group), and the organism conditions in order to determine the outcome of
interest associated with the exposure of interest.
Outcome Assessment
Metrics in this domain are used to assess the validity and reliability of methods, including
sensitivity of methods, that are used to measure or otherwise characterize the outcome(s)
of interest.
B-6
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
l-'.\iiliiiiiion Doniiiin
Definition
Confounding/Variable
Control
Metrics in this domain are used to assess the potential impact of factors other than
exposure that might affect the risk of outcome. The metrics evaluate whether studies
identify and account for factors that are related to exposure and independently related to
outcome (confounding factors) and whether appropriate experimental or analytical
(statistical) methods are used to control for factors unrelated to exposure that might affect
the risk of outcome (variable control).
Data Presentation and
Analysis
Metrics in this domain are used to assess whether appropriate statistical methods were
used and whether data for all outcomes are presented.
Other
Metrics in this domain are added as needed to incorporate chemical- or study-specific
evaluations.
Note:
a Reliability is defined as "the inherent property of a study or data, which includes the use of well-founded scientific approaches,
the avoidance of bias within the study or data collection design and faithful study or data collection conduct and
documentation" (ECHAJ_2011a).
B.1.6. Data Quality Evaluation Metrics
The data quality evaluation domains are evaluated In assessing unique metrics that have been
developed for animal and in vitro studies. Each metric is binned into a confidence level of High,
Medium, Low, or Unacceptable. Each confidence le\ el is assigned a numerical score (i.e., 1
through 4) that is used in the method of assessing the o\ ei ill I quality of the study.
Table B-4 lists the data evaluation domains and metrics for animal toxicity studies, including
metrics that inform risk of bias and types of bias, and Table li-5 lists the data evaluation domains
and metrics for in vino toxicity studies. Each domain has between 2 and 6 metrics; however,
some metrics mighl not apply to all study types. A general domain for other considerations is
available for metrics that are specific to a given test substance or study type.
EPA may modify the metrics used for animal toxicity and in vitro toxicity studies as the Agency
acquires experience with the e\ aluation tool Any modifications will be documented.
Table 15-4. Evaluation Domains and Metrics for Animal Toxicity Studies
l'.\iiliiiilioii
Number of
M dries
Doniiiin
Metrics
()\cr;ill
(Meli'ie Number ;iihI Description. Tj pc of l$i;is)
• Metric 1. Tesi Substance klenlilN
Tes>l S ub ^ la nee
3
• Metric 2: Test Substance Source
• Metric 3: Test Substance Purity (information bias3) (*detection biasb)
* Metric 4: Negative and Vehicle Controls (*performance biasb)
Test Design
3
• Metric 5: Positive Controls (information bias3)
• Metric 6: Randomized Allocation of Animals (*selection biasa b)
• Metric 7: Preparation and Storage of Test Substance
• Metric 8: Consistency of Exposure Administration
Exposure
Characterization
• Metric 9: Reporting of Doses/Concentrations
6
• Metric 10: Exposure Frequency and Duration
• Metric 11: Number of Exposure Groups and Dose/Concentration
Spacing
• Metric 12: Exposure Route and Method
B-7
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
l'\;i III lit ion
Doniiiin
Number of
Metrics
()\er;ill
Metrics
(Metric Number ;iihI Description. T\pe of
Test Organisms
3
• Metric 13: Test Animal Characteristics
• Metric 14: Adequacy and Consistency of Animal Husbandry Conditions
• Metric 15: Number of Animals per Group (*missing data bias3)
Outcome
Assessment
5
• Metric 16: Outcome Assessment Methodology
(^information bias3) (*detection biasb)
• Metric 17: Consistency of Outcome Assessment
• Metric 18: Sampling Adequacy
• Metric 19: Blinding of Assessors
(*sclcction bias ") (*performancc bias1')
• Metric 20: Negatiw ( oiiiiol Responses
Confounding/
Variable Control
2
• Metric 21: Confounding \ anahles in Test Design and Procedures
(*other biasb)
• Metric 22 1 Icallli (Hilcomcs Unrelated in 1 \posure (*attrition/exclusion
biasb)
Data
Presentation and
Analysis
2
• Metric 23 Sialisiical Moll mils (^information hias 11 :;:other biasb)
• Metric 24: Repoiiiuu of 1 )ala (^selective report11iu bias' i
Notes:
Items marked with an asterisk (*) are examples of metrics thai can be used to assess internal validity/risk of bias.
^National Academies of Sciences, Engineering, and Medicine. 2017.
bNTP, 2015b.
Table B-5. Data Evaluation Domains and Metrics lor In Vitro Toxicity Studies
l-'.\ ;i In ;i t ion
Doniiiin
Number of
Metrics ()\er
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
l-'.\ ;i III ;i t ion
Doniiiin
Number of
Mon ies (hersill
Metrics
(Metric Number ;uul Description. T\|)o of ISi;is)
Confounding/
Variable Control
2
• Metric 20: Confounding Variables in Test Design and Procedures
• Metric 21: Confounding Variables in Outcomes Unrelated to
Exposure
Data Presentation
and Analysis
4
• Metric 22: Data Analysis
• Metric 23: Data Interpretation
• Metric 24: Cytotoxicity Data
• Metric 25: Reporting of Data
Note:
aThese are for the assay performance, not necessarily for the "validation" of extrapolating to a particular apical outcome (i.e.,
assay performance versus assay validation).
B.1.7. Scoring Method and Determination of Overall Data Quality Level
Section B.2.2 provides information about the e\ alualion method thai is used to assess the quality
of the data/information. This section provides details about the scoring system that is applied to
animal and in vitro toxicity studies, including the weighting factors assigned to each metric score
of each domain.
Some metrics are given greater weights than others, if they are regarded as key or critical
metrics. Thus, EPA OPPT uses a weighting approach to reflect that some metrics are more
important than others when assessing the o\ erall quality of the data.
B. 1.7.1. Weighting Factors
Each metric is assigned a weighting factor of I or 2. with the higher weighting factor (2) given to
metrics deemed critical for the c\ alualion The critical metrics are identified based on
professional judgment in conjunction with consideration of the factors that are most frequently
included in other study quality risk of bias tools for animal toxicity studies (reviewed by Lynch
et al. (2<)!(•>). Samuel el al (2<)!(¦>)) In selecting critical metrics, EPA OPPT recognized that the
relevance of an individual study to the risk analysis for a given substance is determined by its
ability to inform hazard identification and or dose-response assessment. Thus, the critical metrics
are those that can be used to determine how well a study answers these key questions:
• Is a change in health outcome demonstrated in the study?
• Is the obsen ed change more likely than not attributable to the substance exposure?
• At what substance dose( s) does the change occur?
EPA OPPT assigns a weighting factor of 2 to each metric considered critical to answering these
questions. Remaining metrics are assigned a weighting factor of 1. Tables B-6 and B-7 identify
the critical metrics (i.e., those assigned a weighting factor of 2) for animal toxicity and in vitro
toxicity studies, respectively, and provide a rationale for selection of each metric. Tables B-8 and
B-9 identify the weighting factors assigned to each metric for animal toxicity and in vitro
toxicity studies, respectively.
B-9
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Table B-6. Animal Toxicity Metrics with Greater Importance in the Evaluation and
Rationale for Selection
Domain
Critical Metrics with
Weighting Factor of 2
(Metric Number)3
Rationale
Test Substance
Test Substance Identity
(Metric 1)
The test substance must be identified and characterized definitively
to ensure that the study is relevant to the substance of interest.
Test Design
Negative and Vehicle
Controls
(Metric 4)
A concurrent negative control and vehicle control (when indicated)
are required to ensure that any observed effects are attributable to
substance exposure. Note that more than one negative control might
be necessary in some studies.
Exposure
Characterization
Reporting of
Doses/Concentrations
(Metric 9)
Dose levels must be defined without ambiguity to allow for
determination of the dose-response relationship and to enable valid
comparisons across studies.
Test Organisms
Test Animal
Characteristics
(Metric 13)
The test animal characteristics must be reported to enable assessment
of (a) whether they are suitable for the endpoint of interest; (b)
whether there are species, strain sex, or age/lifestage differences
within or between different studies; and (c) to enable consideration of
approaches for extrapolation to humans.
Outcome
Assessment
Outcome Assessment
Methodology
(Metric 16)
The methods used for outcome assessment must be fully described,
valid, and sensitive to ensure that effects are detected, that observed
effects are true, and to enable valid comparisons across studies.
Confounding/
Variable Control
Confounding Variables
in Test Design and
Procedures
(Metric 21)
Control for confounding variables in test design and procedures is
necessary to ensure that any observed effects are attributable to
substance exposure and not to other factors.
Data
Presentation and
Analysis
Reporting of Data
(Metric 24)
Detailed results are necessary to determine whether the study
authors' conclusions are valid and to enable dose-response modeling.
Note:
aA weighting factor of 1 is assigned for the remaining metrics.
Table B-7. In Vitro Toxicity Metrics with Greater Importance in the Evaluation and
Rationale for Selection
Domain
Critical Metrics with
Weighting Factor of 2
(Metric Number)3
Rationale
Test Substance
Test Substance Identity
(Metric 1)
The test substance must be identified and characterized definitively
to ensure that the study is relevant to the substance of interest.
Test Design
Negative and Vehicle
Controls
(Metric 4)
A concurrent negative control and vehicle control (when indicated) are
required for comparison of results between exposed and unexposed
models to allow determination of treatment-related effects.
Positive Controls
(Metric 5)
A concurrent positive control or proficiency control (when
applicable) is required to determine whether the chemical of interest
produces the intended outcome for the study type.
Exposure
Characterization
Reporting of
doses/concentrations
(Metric 10)
Dose levels must be defined without ambiguity to allow for
determination of an accurate dose-response relationship or and to
ensure valid comparisons across studies.
Exposure Duration
(Metric 11)
The exposure duration during the study must be defined to accurately
assess potential risk.
B-10
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Domain
Critical Metrics with
Weighting Factor of 2
(Metric Number)3
Rationale
Test Model
Test Model
(Metric 14)
The identity of the test model must be reported and suitable for the
evaluation of outcome(s) of interest.
Outcome
Assessment
Outcome Assessment
Methodology
(Metric 16)
The methods used for outcome assessment must be fully described,
valid, and sensitive to ensure that effects are detected and that
observed effects are true.
Sampling Adequacy
(Metric 18)
The number of samples evaluated must be sufficient to allow data
interpretation and analysis.
Confounding/
Variable Control
Confounding Variables
in Test Design and
Procedures
(Metric 20)
Control for confounding variables in test design and procedures are
necessary to ensure that any observed effects are attributable to
substance exposure and not to other factors.
Data
Presentation and
Analysis
Data Interpretation
(Metric 23)
The criteria for scoring and/or evaluation criteria are necessary so
that the correct categorization (e.g., positive, negative, equivocal)
can be determined for the chemical of interest.
Reporting of Data
(Metric 25)
Detailed results are necessary to determine whether the study
authors' conclusions are valid and to enable dose-response modeling.
Note:
aA weighting factor of 1 is assigned for the remaining metrics.
B. 1.7.2. Calculation of Overall Study Score
A confidence level (1, 2, or 3 for High, Medium, or Low confidence, respectively) is assigned for
each relevant metric within each domain. To determine the overall study score, the first step is to
multiply the score for each metric (1, 2, or 3 for High, Medium, or Low confidence, respectively)
by the appropriate weighting factor (as shown in Tables B-8 and B-9 for animal toxicity and in
vitro studies, respectively) to obtain a weighted metric score. The weighted metric scores are then
summed and divided by the sum of the weighting factors (for all metrics that are scored) to obtain an
overall study score between 1 and 3. The equation for calculating the overall score is shown below:
Overall Score (range of 1 to 3) =^(Metric Score x Weighting Factor)/^(Weighting Factors)
Some metrics might not be applicable to all study types. These metrics will not be included in the
numerator or denominator of the equation above. The overall score will be calculated using only
those metrics that receive a numerical score. Scoring examples for animal toxicity and in vitro
toxicity studies are in tables B-10 through B-13.
Studies with any single metric scored as unacceptable (score = 4) will be automatically assigned
an overall quality score of 4 (Unacceptable). An unacceptable score means that serious flaws are
noted in the domain metric that consequently make the data unusable. If a metric is not
applicable for a study type, the serious flaws would not be applicable for that metric and would
not receive a score. EPA OPPT plans to use data with an overall quality level of High, Medium,
or Low confidence to quantitatively or qualitatively support the risk evaluations, but it does not
plan to use data rated as Unacceptable. An overall study score will not be calculated when a
serious flaw is identified for any metric. If a publication reports more than one study or endpoint,
each study or endpoint will be evaluated separately and given a separate overall quality rating.
B-ll
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Table B-8. Metric Weighting Factors and Range of Weighted Metric Scores for Animal
Toxicity
Domain
Number/
Description
Metric Number/Description
Range
of
Metric
Scores"
Metric
Weighting
Factor
Range of
Weighted
Metric
Scoresb
1. Test Substance
1. Test Substance Identity
1 to 3
2
2 to 6
2. Test Substance Source
1
1 to 3
3. Test Substance Purity
1
1 to 3
2. Test Design
4. Negative and Vehicle Controls
2
2 to 6
5. Positive Controls
1
1 to 3
6. Randomized Allocation
1
1 to 3
3. Exposure
Characterization
7. Preparation and Storage of Test Substance
1
1 to 3
8. Consistency of Exposure Administration
1
1 to 3
9. Reporting of Doses/Concentrations
2
2 to 6
10. Exposure Frequency and Duration
1
1 to 3
11. Number of Exposure Groups and Dose Spacing
1
1 to 3
12. Exposure Route and Method
1
1 to 3
4. Test Organisms
13. Test Animal Characteristics
2
2 to 6
14. Adequacy and Consistency of Animal
Husbandry Conditions
1
1 to 3
15. Number per Group
1
1 to 3
5. Outcome
Assessment
16. Outcome Assessment Methodology
2
2 to 6
17. Consistency of Outcome Assessment
1
1 to 3
18. Sampling Adequacy
1
1 to 3
19. Blinding of Assessors
1
1 to 3
20. Negative Control Responses
1
1 to 3
6. Confounding/
Variable Control
21. Confounding Variables in Test Design and
Procedures
2
2 to 6
22. Health Outcomes Unrelated to Exposure
1
1 to 3
7. Data
Presentation and
Analysis
23. Statistical Methods
1
1 to 3
24. Reporting of Data
2
2 to 6
Sum (if all metrics scored)0
31
31 to 93
Range of Overall S
Overall
cores, where
Score = Sum of We
High
ighted Scores/Sum
Medium
of Metric Weightin
Low
31/31 = 1;
g Factor 93/31 = 3
Range of
overall
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
score = 1 to 3 d
Notes'.
Tor the purposes of calculating an overall study score, the range of possible metric scores is 1 to 3 for each metric,
corresponding to High and Low confidence. No calculations will be conducted if a study receives an "Unacceptable" rating
(score of "4") for any metric.
bThe range of weighted scores for each metric is calculated by multiplying the range of metric scores (1 to 3) by the weighting
factor for that metric.
The sum of weighting factors and the sum of the weighted scores will differ if some metrics are not scored (not applicable).
dThe range of possible overall scores is 1 to 3. If a study receives a score of 1 for every metric, then the overall study score will
be 1. If a study receives a score of 3 for every metric, then the overall study score will be 3.
B-12
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Table B-9. Metric Weighting Factors and Range of Weighted Metric Scores for In Vitro
Toxicity Studies
Domain
Number/
Description
Metric Number/Description
Range
of
Metric
Scores8
Metric
Weighting
Factor
Range of
Weighted
Metric
Scoresb
1. Test Substance
1. Test Substance Identity
1 to 3
2
2 to 6
2. Test Substance Source
1
1 to 3
3. Test Substance Purity
1
1 to 3
2. Test Design
4. Negative and Vehicle Controls
2
2 to 6
5. Positive Controls
2
2 to 6
6. Assay Procedures
1
1 to 3
7. Standards for Test
1
1 to 3
3. Exposure
Characterization
8. Preparation and Storage of Test Substance
1
1 to 3
9. Consistency of Exposure Administration
1
1 to 3
10. Reporting of Doses/Concentrations
2
2 to 6
11. Exposure Duration
2
2 to 6
12. Number of Exposure Groups and Dose Spacing
1
1 to 3
13. Metabolic Activation
1
1 to 3
4. Test Model
14. Test Model
2
2 to 6
15. Number per Group
1
1 to 3
5. Outcome
Assessment
16. Outcome Assessment Methodology
2
2 to 6
17. Consistency of Outcome Assessment
1
1 to 3
18. Sampling Adequacy
2
2 to 6
19. Blinding of Assessors
1
1 to 3
6. Confounding/
Variable Control
20. Confounding Variables in Test Design and
Procedures
2
2 to 6
21. Outcomes Unrelated to Exposure
1
1 to 3
7. Data
Presentation and
Analysis
22. Data Analysis
1
1 to 3
23. Data Interpretation
2
2 to 6
24. Cytotoxicity Data
1
1 to 3
25. Reporting of Data
2
2 to 6
Sum (if all metrics scored)0
36
36 to 108
Range of Overall S
Overall
cores, where
Score = Sum of We
High
ighted Scores/Sum
Medium
of Metric Wei
Low
ghtin
3
g Factor
36/36 = 1;
108/36 = 3
Range of
overall
score = 1 to 3 d
>1 and <1.7
>1.7 and <2.3
>2.3 and <
Notes'.
Tor the purposes of calculating an overall study score, the range of possible metric scores is 1 to 3 for each metric,
corresponding to High and Low confidence. No calculations will be conducted if a study receives an "Unacceptable" rating
(score of "4") for any metric.
bThe range of weighted scores for each metric is calculated by multiplying the range of metric scores (1 to 3) by the weighting
factor for that metric.
The sum of weighting factors and the sum of the weighted scores will differ if some metrics are not scored (not applicable).
dThe range of possible overall scores is 1 to 3. If a study receives a score of 1 for every metric, then the overall study score will
be 1. If a study receives a score of 3 for every metric, then the overall study score will be 3.
B-13
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
Table B-10. Scoring Example for Animal Toxicity Study with all Metrics Scored
NOVEMBER 2018
Domain
Metric
Metric
Score
Metric Weighting
Factor
Test Substance
1. Test Substance Identity
2. Test Substance Source
3. Test Substance Purity
Test Design
4. Negative and Vehicle Controls
5. Positive Controls
6. Randomized Allocation
Exposure Characterization
7. Preparation and Storage of Test Substance
8. Consistency of Exposure Administration
9. Reporting of Doses/Concentrations
10. Exposure Frequency and Duration
11. Number of Exposure Groups and Dose Spacing
12. Exposure Route and Method
Test Organisms
13. Test Animal Characteristics
14. Consistency of Animal Conditions
15. Number per Group
Outcome Assessment
16. Outcome Assessment Methodology
17. Consistency of Outcome Assessment
18. Sampling Adequacy
19. Blinding of Assessors
20. Negative Control Responses
Confounding/Variable Control
21. Confounding Variables in Test Design and Procedures
22. Health Outcomes Unrelated to Exposure
Data Presentation and Analysis
23. Statistical Methods
24. Reporting of Data
NR= not rated/not applicable
Sum of scores
Overall Study Score
1.9
31
: Medium
59
Overall Score = Sum of Weighted Scores/Sum of Metric Weighting Factors
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-14
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Table B-ll. Scoring Example for Animal Toxicity Study with Some Metrics Not Rated/Not Applicable
Metric
Metric Weighting
Weighted
Domain
Metric
Score
Factor
Score
1. Test Substance Identity
2
2
4
Test Substance
2. Test Substance Source
3
1
3
3. Test Substance Purity
2
1
2
4. Negative and Vehicle Controls
1
2
Test Design
5. Positive Controls
NR
6. Randomized Allocation
3
1
3
7. Preparation and Storage of Test Substance
2
1
2
8. Consistency of Exposure Administration
NR
Exposure Characterization
9. Reporting of Doses/Concentrations
10. Exposure Frequency and Duration
1
2
1
2
2
11. Number of Exposure Groups and Dose Spacing
1
1
1
12. Exposure Route and Method
1
1
1
13. Test Animal Characteristics
2
4
Test Organisms
14. Consistency of Animal Conditions
2
1
2
15. Number per Group
1
1
1
16. Outcome Assessment Methodology
2
4
17. Consistency of Outcome Assessment
NR
Outcome Assessment
18. Sampling Adequacy
2
1
2
19. Blinding of Assessors
NR
20. Negative Control Responses
2
1
2
Confounding/Variable Control
21. Confounding Variables in Test Design and Procedures
22. Health Outcomes Unrelated to Exposure
2
2
1
4
2
Data Presentation and Analysis
23. Statistical Methods
24. Reporting of Data
2
2
1
2
2
4
NR = not rated/not applicable
Sum
27
49
Overall Study Score
1.8
= Medium
Overall Score
= Sum of Weighted Scores/Sum of Metric Weighting Factor
High
Medium
Low
| >1 and <1.7
>1.7 arid <2.3
>2.3 and <3
B-15
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
Table B-12. Scoring Example for In Vitro Study with all Metrics Scored
NOVEMBER 2018
Domain
Metric
Metric Score
Metric Weighting
Factor
Test Substance
1. Test Substance Identity
2. Test Substance Source
3. Test Substance Purity
Test Design
4. Negative Controls
5. Positive Controls
6. Assay Procedures
7. Standards for Test
Exposure Characterization
8. Preparation and Storage of Test Substance
9. Consistency of Exposure Administration
10. Reporting of Doses/Concentrations
11. Exposure Duration
12. Number of Exposure Groups and Dose Spacing
13. Metabolic Activation
Test Model
14. Test Model
15. Number per Group
Outcome Assessment
16. Outcome Assessment Methodology
17. Consistency of Outcome Assessment
18. Sampling Adequacy
19. Blinding of Assessors
Confounding/Variable Control
20. Confounding Variables in Test Design and Procedures
21. Outcomes Unrelated to Exposure
Data Presentation and Analysis
22. Data Analysis
23. Data Interpretation
24. Cytotoxicity Data
25. Reporting of Data
NR = not rated/not applicable
Sum
36
Overall Study Score
1.8
= Medium
66
Overall Score = Sum of Weighted Scores/Sum of Metric Weighting Factor
High
Medium
Low
51 arid <1.7
>1.7 and <2.3
>2.3 and <3
B-16
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Table B-13. Scoring Example for In Vitro Study with Some Metrics Not Rated/Not Applicable
Metric Weighting
Weighted
Domain
Metric
Metric Score
Factor
Score
1. Test Substance Identity
1
2
2
Test Substance
2. Test Substance Source
2
1
2
3. Test Substance Purity
2
1
2
4. Negative Controls
1
2
2
Test Design
5. Positive Controls
6. Assay Procedures
1
2
2
1
2
2
7. Standards for Test
3
1
3
8. Preparation and Storage of Test Substance
NR
9. Consistency of Exposure Administration
2
1
2
Exposure Characterization
10. Reporting of Doses/Concentrations
11. Exposure Duration
1
1
2
2
2
2
12. Number of Exposure Groups and Dose Spacing
1
1
1
13. Metabolic Activation
NR
Test Model
14. Test Model
15. Number per Group
2
3
2
1
4
3
16. Outcome Assessment Methodology
3
2
6
Outcome Assessment
17. Consistency of Outcome Assessment
18. Sampling Adequacy
19. Blinding of Assessors
2
1
NR
1
2
2
2
Confounding/Variable Control
20. Confounding Variables in Test Design and Procedures
21. Outcomes Unrelated to Exposure
3
2
2
1
6
2
22. Data Analysis
1
1
1
Data Presentation and Analysis
23. Data Interpretation
24. Cytotoxicity Data
2
NR
2
4
25. Reporting of Data
3
2
6
NR= not rated/not applicable
Sum
32
58
Overall Study Score
1.8
= Medium
Overall Score = Sum of Weighted Scores/Sum of Metric Weighting Factor
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-17
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
B.1.8. Data Quality Criteria
5.1.8.1. Animal Toxicity Studies
Detailed tables showing quality criteria for the metrics are provided in Tables B-14 through B-17
for animal toxicity and in vitro toxicity studies, including a table that summarizes the serious
flaws that would make the data unacceptable for use in the environmental hazard assessment.
Table B-14. Serious Flaws that Would Make Animal Toxicity Studies Unacceptable
Domain
Metric
Description of Serious Flaw(s) in Data Source
Test Substance Identity
The test substance identity and form (the latter if applicable) cannot
be determined from the information provided (e.g., nomenclature
was unclear and Chemical Abstracts Service Registry Number
(CASRN) or structure were not reported)
OR
for mixtures, the components and ratios were not characterized.
Test Substance
The test substance was not obtained from a manufacturer
Test Substance Source
OR
if synthesized or extracted, analytical verification of the test
substance was not conducted.
Test Substance Purity
The nature and quantity of reported impurities were such that study
results were likely to be due to one or more of the impurities.
Negative and Vehicle
Controls
A concurrent negative control group was not included or reported
OR
the reported negative control group was not appropriate (e.g.,
age/weight of animals differed between control and treated groups).
Test Design
Positive Controls
For study types that require a concurrent positive control group:
When applicable, an appropriate concurrent positive control (i.e.,
inducing a positive response) was not used and its omission is a
serious flaw that makes the study unusable.
Randomized Allocation
of Animals
The study reported using a biased method to allocate animals to
study groups (e.g., judgement of investigator).
Information on preparation and storage was not reported
OR
serious flaws reported with test substance preparation and/or storage
conditions will have critical impacts on dose/concentration estimates
and make the study unusable (e.g., instability of test substance in
exposure medium was reported, or there was heterogeneous
distribution of test substance in exposure matrix (e.g., aerosol
deposition in exposure chamber, insufficient mixing of dietary
matrix)). For inhalation studies, there was no mention of the method
and equipment used to generate the test substance, or the method
used is atypical and inappropriate.
Exposure
Characterization
Preparation and Storage
of Test Substance
Consistency of Exposure
Administration
Critical exposure details (e.g., methods for generating atmosphere in
inhalation studies) were not reported
OR
reported information indicated that exposures were not administered
consistently across study groups (e.g., differing particle size),
resulting in serious flaws that make the study unusable.
B-18
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Doniiiin
Molik-
Description of Serious l'l;i\\(s) in l)iii;i Source
Reporting of
Doses/Concentrations
The reported exposure levels could not be validated (e.g., lack of
food or water intake data for dietary or water exposures in
conjunction with evidence of palatability differences, lack of body
weight (B W) data in conjunction with qualitative evidence for B W
differences across groups, inconsistencies in reporting, etc.). For
inhalation studies, actual concentrations not reported along with
animal responses (or lack of responses) that indicate exposure
problems due to faulty test substance generation. Animals were
exposed to an aerosol but no particle size data were reported.
Exposure
Characterization
(continued)
Exposure Frequency and
Duration
The exposure frcqucncv or duration of exposure were not reported
OR
the reported exposure frequency and duration were not suited to the
study type and,or oulconie(s) of interest (e.g., study length
inadequate to evaluate tiinmrmeiiicity).
Number of Exposure
Groups and
Dose/Concentration
Spacing
The number of exposure groups and spacing were not reported
OR
dose groups and spaeiuu were noi rele\ am lor the assessment (e.g.,
all doses m a de\ elopmeiital toxicity siud> produced overt maternal
lo\ieil>)
1 lie route or method of e\posure was not reported
OK
au inappropriate route or method (e.g., administration of a volatile
oruauie compound \ la the diet) was used for the test substance
u itlioui takiuu sieps to correct the problem (e.e.. mixine fresh diet,
replaeiuu air m sialic chambers). For inhalation studies, there is no
descripi khi of the inhalation chamber used, or an atypical exposure
method was used, such as allowing a container of test substance to
e\ aporale mi a room.
Exposure Roulc and
Mel hod
The lesi animal species was not reported
OK
the lesi animal (species, strain, sex, life-stage, source) was not
appropriate for the evaluation of the specific outcome(s) of interest
(e u . genetically modified animals, strain was uniquely susceptible
or resistant to one or more outcome of interest).
Tesi \mi m;i 1
( haraclerisiics
Test Organisms
\dcquac> and
( oiisisiciic> of \nimal
HushandiA Conditions
There were significant differences in husbandry conditions between
control and exposed groups (e.g., temperature, humidity, light-dark
cycle)
OR
animal husbandry conditions deviated from customary practices in
ways likely to impact study results (e.g., injuries and stress due to
cage overcrowding).
Number of Animals per
Group
The number of animals per study group was not reported
OR
the number of animals per study group was insufficient to
characterize toxicological effects (e.g., 1-2 animals in each group).
B-19
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Domain
Metric
Description of Serious Flaw(s) in Data Source
The outcome assessment methodology was not reported
OR
the reported outcome assessment methodology was not sensitive for
the outcome(s) of interest (e.g., evaluation of endpoints outside the
critical window of development, a systemic toxicity study that
evaluated only grossly observable endpoints, such as clinical signs
and mortality).
Outcome Assessment
Methodology
Consistency of Outcome
Assessment
There were large inconsistencies in the execution of study protocols
for outcome assessment across study groups
OR
outcome assessments were not adequately reported for meaningful
interpretation of results.
Outcome
Assessment
Sampling Adequacy
Sampling was not adequate for the outcome(s) of interest (e.g.,
histopathology was performed on exposed groups, but not controls).
Blinding of Assessors
Information in the study report did not include whether assessors
were blinded to treatment group for subjective outcomes and
suggested that the assessment of subjective outcomes (e.g.,
functional observational battery, qualitative neurobehavioral
endpoints, histopathological re-evaluations) was performed in a
biased fashion (e.g., assessors of subjective outcomes were aware of
study groups). This is a serious flaw that makes the study unusable.
Negative Control
Responses
The biological responses of the negative control groups were not
reported
OR
there was unacceptable variation in biological responses between
control replicates.
Confounding/
variable control
Confounding Variables
in Test Design and
Procedures
The study reported significant differences among the study groups
with respect to initial BW, decreased drinking water/food intake due
to palatability issues (> 20% difference from control) that could lead
to dehydration and/or malnourishment, or reflex bradypnea that
could lead to decreased oxygenation of the blood.
Health Outcomes
Unrelated to Exposure
One or more study groups experienced serious animal attrition or
health outcomes unrelated to exposure (e.g., infection).
Statistical methods used were not appropriate (e.g., parametric test
for non-normally distributed data)
OR
statistical analysis was not conducted
AND
data were not provided preventing an independent statistical
analysis.
Data
Presentation and
Analysis
Statistical Methods
Reporting of Data
Data presentation was inadequate (e.g., the report does not
differentiate among findings in multiple exposure groups)
OR
major inconsistencies were present in reporting of results.
B-20
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Table B-15. Data Quality Criteria for Animal Toxicity Studies
Confidence Level
(Score)
Description
Selected
Score
Domain 1. Test Substance
Metric 1. Test Substance Identity
Was the test substance identified definitively (i.e., established nomenclature, CASRN, and/or structure reported,
including information on the specific form tested (particle characteristics for solid-state materials, salt or base,
valence state, hydration state, isomer, radiolabel, etc.) for materials that may vary in form)? If test substance is a
mixture, were mixture components and ratios characterized?
High
(score = 1)
The test substance was identified definitively, and the specific form was
characterized (where applicable). For mixtures, the components and ratios
were characterized.
Medium
(score = 2)
The test substance and form (the latter if applicable) were identified, and
components and ratios of mixtures were characterized, but there were minor
uncertainties (e.g., minor characterization details were omitted) that are
unlikely to have a substantial impact on results.
Low
(score = 3)
The test substance and form (the latter if applicable) were identified and
components and ratios of mixtures were characterized, but there were
uncertainties regarding test substance identification or characterization that
are likely to have a substantial impact on results.
Unacceptable
(score = 4)
The test substance identity and form (the latter if applicable) cannot be
determined from the information provided (e.g., nomenclature was unclear
and CASRN or structure were not reported)
OR
for mixtures, the components and ratios were not characterized. These are
serious flaws that make the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 2. Test Substance Source
Was the source of the test substance reported, including manufacturer and batch/lot number for materials that may
vary in composition? If synthesized or extracted, was test substance identity verified by analytical methods?
High
(score = 1)
The source of the test substance was reported, including manufacturer and
batch/lot number for materials that may vary in composition and its identity
was certified by manufacturer and/or verified by analytical methods (melting
point, chemical analysis, etc.).
Medium
(score = 2)
The source of the test substance and/or the analytical verification of a
synthesized test substance was reported incompletely, but the omitted details
are unlikely to have a substantial impact on results.
Low
(score = 3)
Omitted details on the source of the test substance and/or the analytical
verification of a synthesized test substance are likely to have a substantial
impact on results.
Unacceptable
(score = 4)
The test substance was not obtained from a manufacturer
OR
if synthesized or extracted, analytical verification of the test substance was not
conducted. These are serious flaws that makes the study unusable.
Not rated/applicable
B-21
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 3. Test Substance Purity
Was the purity or grade (i.e., analytical, technical) of the test substance reported and adequate to identify its
toxicological effects? Were impurities identified? Were impurities present in quantities that could influence the
results?
High
(score = 1)
The test substance purity and composition were such that any observed effects
were highly likely to be due to the nominal test substance itself (e.g., highly
pure or analytical-grade test substance or a formulation comprising primarily
inert ingredients with small amount of active ingredient).
Medium
(score = 2)
Minor uncertainties or limitations were identified regarding the test substance
purity and composition; however, the purity and composition were such that
observed effects were more likely than not due to the nominal test substance,
and any identified impurities are unlikely to have a substantial impact on
results. Alternately, purity was not reported but given other information purity
was not expected to be of concern.
Low
(score = 3)
Purity and/or grade of test substance were not reported or were low enough to
have a substantial impact on results (i.e., observed effects may not be due to
the nominal test substance).
Unacceptable
(score = 4)
The nature and quantity of reported impurities were such that study results
were likely to be due to one or more of the impurities. This is a serious flaw
that makes the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Domain 2. Test Design
Metric 4. Negative and Vehicle Controls
Was an appropriate concurrent negative control group included? If a vehicle was used, was the control group
exposed to the vehicle? For inhalation and gavage studies, were controls sham-exposed?
High
(score = 1)
Study authors reported using an appropriate concurrent negative control group
(i.e., all conditions equal except chemical exposure). If gavage or inhalation
study, a vehicle and/or sham-treated control group was included.
Medium
(score = 2)
Study authors reported using a concurrent negative control group, but all
conditions were not equal to those of treated groups; however, the identified
differences are considered to be minor limitations that are unlikely to have a
substantial impact on results.
Low
(score = 3)
Study authors acknowledged using a concurrent negative control group, but
details regarding the negative control group were not reported, and the lack of
details is likely to have a substantial impact on results.
Unacceptable
(score = 4)
A concurrent negative control group was not included or reported
OR
the reported negative control group was not appropriate (e.g., age/weight of
animals differed between control and treated groups). This is a serious flaw
that makes the study unusable.
Not rated/applicable
B-22
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 5. Positive Controls
Was an appropriate concurrent positive control group included if necessary based on study type (e.g., certain
neurotoxicity studies)?
This metric is not rated/applicable if positive control was not indicated by study type.
High
(score = 1)
When applicable, a concurrent positive control was used (if necessary for the
study type) and a positive response was observed.
Medium
(score = 2)
When applicable, a concurrent positive control was used, but there were minor
uncertainties (e.g., minor details regarding control exposure or response were
omitted) that are unlikely to have a substantial impact on results.
Low
(score = 3)
When applicable, a concurrent positive control was used, but there were
deficiencies regarding the control exposure or response that are likely to have
a substantial impact on results (e.g., the control response was not described).
Unacceptable
(score = 4)
When applicable, an appropriate concurrent positive control (i.e., inducing a
positive response) was not used and its omission is a serious flaw that makes
the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 6. Randomized Allocation of Animals
Did the study explicitly report randomized allocation of animals to study groups?
High
(score = 1)
The study reported that animals were randomly allocated into study groups
(including the control group).
Medium
(score = 2)
The study reported methods of allocation of animals to study groups, but there
were minor limitations in the allocation method (e.g., method with a
nonrandom component like assigmnent to minimize differences in B W across
groups) that are unlikely to have a substantial impact on results.
Low
(score = 3)
The study did not report how animals were allocated to study groups, or there
were deficiencies regarding the allocation method that are likely to have a
substantial impact on results (e.g., allocation by animal number).
Unacceptable
(score = 4)
The study reported using a biased method to allocate animals to study groups
(e.g., judgement of investigator). This is a serious flaw that makes the study
unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
B-23
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Domain 3. Exposure Characterization
Metric 7. Preparation and Storage of Test Substance
Did the study characterize the test substance preparation and storage conditions (e.g., test substance stability,
homogeneity, mixing temperature, stock concentration stirring methods, centrifugation/filtration)? Were the
frequency of preparation and/or storage conditions appropriate to the test substance stability? For inhalation
studies, was the aerosol/vapor generation method appropriate?
High
(score = 1)
The test substance preparation and storage conditions were reported and
appropriate for the test substance (e.g., test substance well-mixed in diet). For
inhalation studies, the method and equipment used to generate the test
substance as a gas, vapor, or aerosol were reported and appropriate.
Medium
(score = 2)
The test substance preparation and storage conditions were reported, but there
were only minor limitations in the test substance preparation and/or storage
conditions (i.e., diet was not mixed fresh daily). Also, any omission of details
regarding preparation and storage that are unlikely to have a substantial
impact on results. For inhalation studies, the method and equipment used to
generate the test substance were incomplete or confusing but there is no
reason to believe there was an impact on animal exposure.
Low
(score = 3)
Deficiencies in reporting of test substance preparation and/or storage
conditions are likely to have a substantial impact on results (e.g., available
information on physical-chemical properties suggested that stability and/or
solubility of test substance in vehicle may be poor). For inhalation studies,
there is reason to question the validity of the method used for generating the
test substance.
Information on preparation and storage was not reported
OR
serious flaws reported with test substance preparation and/or storage
conditions will have critical impacts on dose/concentration estimates and
make the study unusable (e.g., instability of test substance in exposure
medium was reported, or there was heterogeneous distribution of test
substance in exposure matrix (for instance, aerosol deposition in exposure
chamber, insufficient mixing of dietary matrix)). For inhalation studies, there
was no mention of the method and equipment used to generate the test
substance, or the method used is atypical and inappropriate.
Unacceptable
(score = 4)
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 8. Consistency of Exposure Administration
Were exposures administered consistently across study groups (e.g., same exposure frequency; same time of day;
consistent gavage volumes or diet compositions in oral studies; consistent chamber designs, animals/chamber, and
comparable particle size characteristics in inhalation studies; consistent application methods and volumes in
dermal studies)?
High
(score = 1)
Details of exposure administration were reported and exposures were
administered consistently across study groups in a scientifically sound manner
(e.g., gavage volume was not excessive).
Medium
(score = 2)
Details of exposure administration were reported, but minor limitations in
administration of exposures (e.g., accidental mistakes in dosing) were
identified that are unlikely to have a substantial impact on results.
B-24
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Low
(score = 3)
Details of exposure administration were reported, but deficiencies in
administration of exposures (e.g., exposed at different times of day) are likely
to have a substantial impact on results.
Unacceptable
(score = 4)
Critical exposure details (e.g., methods for generating atmosphere in
inhalation studies) were not reported
OR
reported information indicated that exposures were not administered
consistently across study groups (e.g., differing particle size), resulting in
serious flaws that make the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 9. Reporting of Doses/Concentrations
Were doses/concentrations reported without ambiguity (e.g., point estimate in addition to a range)? In oral studies,
if doses were not reported, was information reported that enabled dose estimation (e.g., test animal dietary intake
and BW monitoring data in dietary studies)? In inhalation studies, was test substance vapor/aerosol concentration
measured analytically along with nominal and target concentrations?
For oral and dermal studies, administered doses/concentrations, or the
information to calculate them were reported without ambiguity.
High
(score = 1)
For inhalation studies, several specific considerations apply: Analytical,
nominal and target chamber concentrations were all reported, with high
confidence in the accuracy of the actual concentrations; the range of
concentrations within a treatment group did not deviate widely (range should
be within ±10% for gases and vapors and within ±20% for liquid and solid
aerosols).
The analytical method (high-performance liquid chromatography, gas
chromatography, infrared spectrophotometry, etc.) used to measure chamber
test substance and vehicle concentration was reported and appropriate. Actual
chamber measurements using gravimetric filters are acceptable when testing
dry aerosols and non-volatile liquid aerosols.
The particle size distribution data, mass median aerodynamic diameter
(MMAD), and geometric standard deviation were reported for all exposed
groups (including vehicle controls, when used).
For oral and dermal studies, minor uncertainties in reporting of administered
doses/concentrations occurred (e.g., dietary or air concentrations were not
measured analytically) but are unlikely to have a substantial impact on results.
Medium
(score = 2)
For inhalation studies, several specific considerations apply:
With gases only, actual concentrations were not reported, but there is high
confidence that the animals were exposed at approximately the reported target
concentrations. (There is no comparable medium result for aerosols and
vapors if analytical concentrations are not reported.)
For inhalation studies (gas, vapor, aerosol), the analytical method used was
less than ideal or subject to interference, but, nevertheless, yielded fairly
reliable measurements of chamber concentrations.
Particle size distribution data were not reported, but MMAD and geometric
standard deviation values were reported for all exposed groups (including
vehicle controls, when used).
B-25
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Low
(score = 3)
For oral and dermal studies, deficiencies in reporting of administered
doses/concentrations occurred (e.g., no information on animal BW or intake
were provided) that are likely to have a substantial impact on results.
For inhalation studies, several considerations apply: Using aerosols and
vapors, a score of low is indicated if actual concentrations are not reported or
the analytical method used, such as sampling tubes (e.g., Draeger tubes)
provided imprecise measurements.
An MMAD is reported but no geometric standard deviation or particle size
distribution data were reported.
Unacceptable
(score = 4)
The reported exposure levels could not be validated (e.g., lack of food or
water intake data for dietary or water exposures in conjunction with evidence
of palatability differences, lack of BW data in conjunction with qualitative
evidence forBW differences across groups, inconsistencies in reporting, etc.).
This is a serious flaw that makes the study unusable.
For inhalation studies, actual concentrations were not reported along with
animal responses (or lack of responses) that indicate exposure problems due to
faulty test substance generation.
Animals were exposed to an aerosol but no MMAD or particle size data were
reported.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 10. Exposure Frequency and Duration
Were the exposure frequency (hours/day and days/week) and duration of exposure reported and appropriate for
this study type and/or outcome(s) of interest?
High
(score = 1)
The exposure frequency and duration of exposure were reported and
appropriate for this study type and/or outcome(s) of interest (e.g., inhalation
exposure 6 hours/day, gavage 5 days/week, 2-year duration for cancer
bioassays).
Medium
(score = 2)
Minor limitations in exposure frequency and duration of exposure were
identified (e.g., inhalation exposure of 4 hours/day instead of 6 hours/day in a
repeated exposure study) but are unlikely to have a substantial impact on
results.
Low
(score = 3)
The duration of exposure and/or exposure frequency differed significantly
from typical study designs (e.g., gavage 1 day/week), and these deficiencies
are likely to have a substantial impact on results.
Unacceptable
(score = 4)
The exposure frequency or duration of exposure were not reported
OR
the reported exposure frequency and duration were not suited to the study type
and/or outcome(s) of interest (e.g., study length inadequate to evaluate
tumorigenicity). These are serious flaws that make the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
B-26
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Metric 11. Number of Exposure Groups and Dose/Concentration Spacing
Were the number of exposure groups and dose/concentration spacing justified by study authors (e.g., based on
range-finding studies) and adequate to address the purpose of the study (e.g., to evaluate dose-response
relationships, identify points of departure, inform MOA/AOP, etc.)?
High
(score = 1)
The number of exposure groups and dose/concentration spacing were justified
by study authors and considered adequate to address the purpose of the study
(e.g., the selected doses produce a range of responses).
Medium
(score = 2)
There were minor limitations regarding the number of exposure groups and/or
dose/concentration spacing (e.g., unclear if lowest dose was low enough or the
highest dose was high enough), but the number of exposure groups and
spacing of exposure levels were adequate to show results relevant to the
outcome of interest (e.g., observation of a dose-response relationship) and the
concerns are unlikely to have a substantial impact on results.
Low
(score = 3)
There were deficiencies regarding the number of exposure groups and/or
dose/concentration spacing (e.g., narrow spacing between doses with similar
responses across groups), and these are likely to have a substantial impact on
results.
Unacceptable
(score = 4)
The number of exposure groups and spacing were not reported
OR
dose groups and spacing were not relevant for the assessment (e.g., all doses
in a developmental toxicity study produced overt maternal toxicity). These are
serious flaws that make the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 12. Exposure Route and Method
Were the route and method of exposure reported and suited to the test substance (e.g., was the test substance non-
volatile in dietary studies)?
High
(score = 1)
The route and method of exposure were reported and were suited to the test
substance.
For inhalation studies, a dynamic chamber was used. While dynamic nose-
only (or head-only) studies are generally preferred, dynamic whole-body
chambers are acceptable for gases and for vapors that do not condense.
Medium
(score = 2)
There were minor limitations regarding the route and method of exposure, but
the researchers took appropriate steps to mitigate the problem (e.g., mixed diet
fresh each day for volatile compounds). These limitations are unlikely to have
a substantial impact on results.
For inhalation studies, a dynamic whole-body chamber was used for vapors
that might condense or for aerosols. 5
5 This results in a medium score because in addition to inhalation exposure to the test substance, there may also be
significant oral exposure due to rodents grooming test substance that adheres to their fur. The combined oral and
inhalation exposure results in a lower POD, which makes a test substance appear more toxic than it really is by the
inhalation route.
B-27
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
There were deficiencies regarding the route and method of exposure that are
likely to have a substantial effect on results. Researchers may have attempted
to correct the problem but the success of the mitigating action was unclear.
Low
(score = 3)
For inhalation studies, there are significant flaws in the design or operation of
the inhalation chamber, such as uneven distribution of test substance in a
whole-body chamber, having less than 15 air changes/hour in a whole-body
chamber, or using a whole-body chamber that is too small for the number and
volume of animals exposed.
The route or method of exposure was not reported
OR
an inappropriate route or method (e.g., administration of a volatile organic
compound via the diet) was used for the test substance without takins steos to
correct the problem (e.g., mixing fresh diet). These are serious flaws that
makes the study unusable.
For inhalation studies, either a static chamber was used, there is no description
of the inhalation chamber, or an atypical exposure method was used, such as
allowing a container of test substance to evaporate in a room.
Unacceptable
(score = 4)
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Domain 4. Test Animals
Metric 13. Test Animal Characteristics
Were the test animal species, strain, sex, health status, age, and starting BW reported? Was the test animal from a
commercial source or in-house colony? Was the test species and strain an appropriate animal model for the
evaluation of the specific outcome(s) of interest (e.g., routinely used for similar study types)?
High
(score = 1)
The test animal species, strain sex, health status, age, and starting BW were
reported, and the test animal was obtained from a commercial source or
laboratory-maintained colony. The test species and strain were an appropriate
animal model for the evaluation of the specific outcome(s) of interest (e.g.,
routinely used for similar study types).
Medium
(score = 2)
Minor uncertainties in the reporting of test animal characteristics (e.g., health
status, age, or starting BW) are unlikely to have a substantial impact on
results. The test animals were obtained from a commercial source or in-house
colony, and the test species/strain/sex was an appropriate animal model for the
evaluation of the specific outcome(s) of interest (e.g., routinely used for
similar study types).
Low
(score = 3)
The source of the test animal was not reported
OR
the test animal strain or sex was not reported. These deficiencies are likely to
have a substantial impact on results.
The test animal species was not reported
OR
the test animal (species, strain, sex, life-stage, source) was not appropriate for
the evaluation of the specific outcome(s) of interest (e.g., genetically modified
animals, strain was uniquely susceptible or resistant to one or more outcome
of interest). These are serious flaws that make the study unusable.
Unacceptable
(score = 4)
Not rated/applicable
B-28
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 14. Adequacy and Consistency of Animal Husbandry Conditions
Were all husbandry conditions (e.g., housing, temperature) adequate and the same for control and exposed
populations, such that the only difference was exposure to the test substance?
High
(score = 1)
All husbandry conditions were reported (e.g., temperature, humidity, light-
dark cycle) and were adequate and the same for control and exposed
populations, such that the only difference was exposure.
Medium
(score = 2)
Most husbandry conditions were reported and were adequate and similar for
all groups. Some differences in conditions were identified among groups, but
these differences were considered minor uncertainties or limitations that are
unlikely to have a substantial impact on results.
Low
(score = 3)
Husbandry conditions were not sufficiently reported to evaluate whether
husbandry was adequate and whether differences occurred between control
and exposed populations. These deficiencies are likely to have a substantial
impact on results.
Unacceptable
(score = 4)
There were significant differences in husbandry conditions between control
and exposed groups (e.g., temperature, humidity, light-dark cycle)
OR
animal husbandry conditions deviated from customary practices in ways likely
to impact study results (e.g., injuries and stress due to cage overcrowding).
These are serious flaws that makes the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 15. Number of Animals per Group
Was the number of animals per study group appropriate for the study type and outcome analysis?
High
(score = 1)
The number of animals per study group was reported, appropriate for the
study type and outcome analysis, and consistent with studies of the same or
similar type (e.g., 50/sex/group for rodent cancer bioassay, 10/sex/group for
rodent subchronic study).
Medium
(score = 2)
The reported number of animals per study group was lower than the typical
number used in studies of the same or similar type (e.g., 30/sex/group for
rodent cancer bioassay, 8/sex/group for rodent subchronic study), but it was
sufficient for statistical analysis and this minor limitation is unlikely to have a
substantial impact on results.
Low
(score = 3)
The reported number of animals per study group was not sufficient for
statistical analysis (e.g., varying numbers per group with some groups
consisting of only one animal), and this deficiency is likely to have a
substantial impact on results.
Unacceptable
(score = 4)
The number of animals per study group was not reported
OR
the number of animals per study group was insufficient to characterize
toxicological effects (e.g., 1-2 animals in each group). These are serious flaws
that makes the study unusable.
Not rated/applicable
B-29
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Domain 5. Outcome Assessment
Metric 16. Outcome Assessment Methodology
Did the outcome assessment methodology address or report the intended outcome(s) of interest? Was the outcome
assessment methodology (including endpoints and timing of assessment) sensitive for the outcome(s) of interest
(e.g., measured endpoints that are able to detect a true health effect or hazard)?
Note: Outcome, as addressed in this domain, refers to health effects measured in an animal study (e.g., organ-
specific toxicity, reproductive and developmental toxicity).
High
(score = 1)
The outcome assessment methodology addressed or reported the intended
outcome(s) of interest and was sensitive for the outcomes(s) of interest.
Medium
(score = 2)
The outcome assessment methodology partially addressed or reported the
intended outcomes(s) of interest (e.g., serum chemistry and organ weight
evaluated in the absence of histology), but minor uncertainties are unlikely to
have a substantial impact on results.
Low
(score = 3)
Significant deficiencies in the reported outcome assessment methodology
were identified
OR
due to incomplete reporting, it was unclear whether methods were sensitive
for the outcome of interest. This is likely to have a substantial impact on
results.
The outcome assessment methodology was not reported
OR
the reported outcome assessment methodology was not sensitive for the
outcome(s) of interest (e.g., evaluation of endpoints outside the critical
window of development, a systemic toxicity study that evaluated only grossly
observable endpoints, such as clinical signs and mortality). These are serious
flaws that make the study unusable.
Unacceptable
(score = 4)
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 17. Consistency of Outcome Assessment
Was the outcome assessment carried out consistently (i.e., using the same protocol) across study groups (e.g.,
assessment at the same time after initial exposure in all study groups)?
High
(score = 1)
Details of the outcome assessment protocol were reported, and outcomes were
assessed consistently across study groups (e.g., at the same time after initial
exposure) using the same protocol in all study groups.
Medium
(score = 2)
There were minor differences in the timing of outcome assessment across
study groups or incomplete reporting of minor details of outcome assessment
protocol execution but these uncertainties or limitations are unlikely to have
substantial impact on results.
Low
(score = 3)
Details regarding the execution of the study protocol for outcome assessment
(e.g., timing of assessment across groups) were not reported, and these
deficiencies are likely to have a substantial impact on results.
B-30
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Unacceptable
(score = 4)
There were large inconsistencies in the execution of study protocols for
outcome assessment across study groups
OR
outcome assessments were not adequately reported for meaningful
interpretation of results. These are serious flaws that make the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 18. Sampling Adequacy
Was sampling adequate for the outcome(s) of interest, including experimental unit (e.g., litter vs. individual animal
weight), number of evaluations per dose group, and endpoint (e.g., number of slides evaluated per organ)?
High
(score = 1)
Details regarding sampling for the outcome(s) of interest were reported and
the study used adequate sampling for the outcome(s) of interest (e.g., litter
data provided for developmental studies; endpoints were evaluated in an
adequate number of animals in each group).
Medium
(score = 2)
Details regarding sampling for the outcome(s) of interest were reported, but
minor limitations were identified in the sampling of the outcome(s) of interest
(e.g., histopathology was performed for high-dose group and controls only,
and treatment-related changes were observed at the high dose) that are
unlikely to have a substantial impact on results.
Low
(score = 3)
Details regarding sampling of outcomes were not reported and this deficiency
is likely to have a substantial impact on results.
Unacceptable
(score = 4)
Sampling was not adequate for the outcome(s) of interest (e.g., histopathology
was performed on exposed groups but not controls). This is a serious flaw that
makes the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 19. Blinding of Assessors
Were investigators assessing subjective outcomes (i.e., those evaluated using human judgment, including
functional observational battery, qualitative neurobehavioral endpoints, histopathological re-evaluations) blinded
to treatment group? If blinding was not applied, were quality control/quality assurance procedures for endpoint
evaluation cited?
Note that blinding is not required for initial histopathology review in accordance with Best Practices recommended
by the Society of Toxicologic Pathology. This should be considered when rating this metric.3
This metric is not rated/applicable for initial histopathology review or if no subjective outcomes were assessed
(i.e., only automated measurements were included and/or human judgment was not applied).
B-31
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
High
(score = 1)
The study explicitly reported that investigators assessing subjective outcomes
(i.e., those evaluated using human judgment, including functional
observational battery, qualitative neurobehavioral endpoints, histopathological
re-evaluations) were blinded to treatment group or that quality control/quality
assurance methods were followed in the absence of blinding.
Medium
(score = 2)
The study reported that blinding was not possible, but steps were taken to
minimize bias (e.g., knowledge of study group was restricted to personnel not
assessing subjective outcome) and this minor uncertainty is unlikely to have a
substantial impact on results. Alternately, blinding was not reported; however,
lack of blinding is not expected to have a substantial impact on results.
Low
(score = 3)
The study did not report whether assessors were blinded to treatment group
for subjective outcomes, and this deficiency is likely to have a substantial
impact on results.
Unacceptable
(score = 4)
Information in the study report did not indicate whether assessors were
blinded to treatment group for subjective outcomes or suggested that the
assessment of subjective outcomes (e.g., functional observational battery,
qualitative neurobehavioral endpoints, histopathological re-evaluations) was
performed in a biased fashion (e.g., assessors of subjective outcomes were
aware of study groups). This is a serious flaw that makes the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 20. Negative Control Response
Were the biological responses (e.g., histopathology, litter size, pup viability) of the negative control group(s)
adequate?
High
(score = 1)
The biological responses of the negative control group(s) were adequate (e.g.,
no/low incidence of histopathological lesions).
Medium
(score = 2)
There were minor uncertainties or limitations regarding the biological
responses of the negative control group(s) (e.g., differences in outcome
between untreated and solvent controls) that are unlikely to have a substantial
impact on results.
Low
(score = 3)
The biological responses of the negative control group(s) were reported, but
there were deficiencies regarding the control responses that are likely to have
a substantial impact on results (e.g., elevated incidence of histopathological
lesions).
Unacceptable
(score = 4)
The biological responses of the negative control groups were not reported
OR
there was unacceptable variation in biological responses between control
replicates. These are serious flaws that make the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
B-32
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Domain 6. Confounding/Variable Control
Metric 21. Confounding Variables in Test Design and Procedures
Were there confounding differences among the study groups in initial BW or test substance palatability that could
influence the outcome assessment (e.g., did palatability issues lead to dehydration and/or malnourishment)? Did
reflex bradypnea (i.e., reduced respiration and reduced test substance exposure) induced by respiratory irritants
influence outcome assessment? Were normal signs of reflex bradypnea misinterpreted as neurologic, behavioral,
or developmental effects (e.g., hypothermia, lethargy, unconsciousness, poor performance in behavioral studies,
delayed pup development)?
High
(score = 1)
There were no reported differences among the study groups in initial BW,
food or water intake, or respiratory rate that could influence the outcome
assessment.
Medium
(score = 2)
The study reported minor differences among the study groups (< 20%
difference from control) with respect to initial B W, drinking water and/or
food consumption due to palatability issues, or respiratory rate due to reflex
bradypnea. These minor uncertainties are unlikely to have a substantial impact
on results. Alternately, the lack of reporting of initial BWs, food/water intake,
and/or respiratory rate is not likely to have a significant impact on results.
Low
(score = 3)
Initial BW, food/water intake, and respiratory rate were not reported. These
deficiencies are likely to have a substantial impact on results.
Unacceptable
(score = 4)
The study reported significant differences among the study groups with
respect to initial BW, decreased drinking water/food intake due to palatability
issues (> 20% difference from control) that could lead to dehydration and/or
malnourishment, or reflex bradypnea that could lead to decreased oxygenation
of the blood. These are serious flaws that makes the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 22. Health Outcomes Unrelated to Exposure
Were there differences among the study groups in animal attrition or health outcomes unrelated to exposure (e.g.,
infection) that could influence the outcome assessment? Professional judgement should be used to determine
whether or not signs of infection would invalidate the study. Criteria for High Medium and Low are used when
the study is still usable.
High
(score = 1)
Details regarding animal attrition and health outcomes unrelated to exposure
(e.g., infection) were reported for each study group, and there were no
differences among groups that could influence the outcome assessment.
Medium
(score = 2)
Authors reported that one or more study groups experienced disproportionate
animal attrition or health outcomes unrelated to exposure (e.g., infection), but
data from the remaining exposure groups were valid and the low incidence of
attrition is unlikely to have a substantial impact on results
OR
data on attrition and/or health outcomes unrelated to exposure for each study
group were not reported because only substantial differences among groups
were noted (as indicated by study authors).
B-33
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Low
(score = 3)
Data on attrition and/or health outcomes unrelated to exposure were not
reported for each study group, and this deficiency is likely to have a
substantial impact on results.
OR
data on attrition and/or health outcomes are reported and could have
substantial impact on results.
Unacceptable
(score = 4)
One or more study groups experienced serious animal attrition or health
outcomes unrelated to exposure (e.g., infection). This is a serious flaw that
makes the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Domain 7. Data Presentation and Analysis
Metric 23. Statistical Methods
Were statistical methods clearly described and appropriate for dataset(s) (e.g., parametric test for normally
distributed data)?
High
(score = 1)
Statistical methods were clearly described and appropriate for dataset(s) (e.g.,
parametric test for normally distributed data).
OR
no statistical analyses, calculation methods, and/or data manipulation were
conducted, but sufficient data were provided to conduct an independent
statistical analysis.
Medium
(score = 2)
Statistical analysis was described with some omissions that would unlikely
have a substantial impact on results.
Low
(score = 3)
Statistical analysis was not described clearly, and this deficiency is likely to
have a substantial impact on results.
Unacceptable
(score = 4)
Statistical methods were not appropriate (e.g., parametric test for non-
normally distributed data)
OR
statistical analysis was not conducted
AND
data were not provided preventing an independent statistical analysis. These
are serious flaws that make the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Metric 24. Reporting of Data
Were the data for all outcomes presented? Were data reported by exposure group and sex (if applicable), with
numbers of animals affected and numbers of animals evaluated (for quantal data) or group means and variance (for
continuous data)? If severity scores were used, was the scoring system clearly articulated?
High
(score = 1)
Data for exposure-related findings were presented for all outcomes by
exposure group and sex (if applicable) with quantal and/or continuous
presentation and description of severity scores if applicable. Negative findings
were reported qualitatively or quantitatively.
B-34
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Medium
(score = 2)
Data for exposure-related findings were reported for most, but not all.
outcomes by exposure group and sex (if applicable) with quantal and/or
continuous presentation and description of severity scores if applicable. The
minor uncertainties in outcome reporting are unlikely to have substantial
impact on results.
Low
(score = 3)
Data for exposure-related findings were not shown for each study group, but
results were described in the text and/or data were only reported for some
outcomes. These deficiencies are likely to have a substantial impact on results.
Unacceptable
(score = 4)
Data presentation was inadequate (e.g., the report does not differentiate
among findings in multiple exposure groups)
OR
major inconsistencies were present in reporting of results. These are serious
flaws that make the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
Domain 8. Other (Apply as Needed)
Metric:
High
(score = 1)
Medium
(score = 2)
Low
(score = 3)
Unacceptable
(score = 4)
Not rated/applicable
Reviewer's comments
Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important elements
such as relevance)
aCrissman et al., 2004
B-35
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
B.l.8.2. In Vitro Toxicity Studies
Table B-16. Serious Flaws that Would Make In Vitro Toxicity Studies Unacceptable
Doniiiin
Md ril-
Di-M-riplion of Serious l'l;i\\(s) in l):il;i Souri-i-'1
Test Substance
Identity
The test substance identity and form (if applicable) could not be
determined from the information provided (e.g., nomenclature was
unclear and CASRN or structure were not reported)
OR
the components and ratios of miMures were not characterized.
Test Substance
Test Substance
Source
The test substance was not obtained from a manufacturer
OR
if synthesized or cMiacted. aiiak lical verification of the test substance
was not conducted.
Test Substance
Purity
The rialiiiv and quantity of reported impurities were such that study
results were likely to be due to one or mure of the impurities.
Negative Controls
A concurrent negative control group was not iucluded or reported
OR
the reported ueuali\ e control Lira up was not appropriate (e.g., different
cell lines used lor controls and lest substance e\posure).
Positive Controls
\ concurrent posiii\ e couirol or proficiency group was not used (when
applicable i
Test Design
\ssa\ I'livcdIIICS
\ssa> methods and procedures* were not reported
OK
assa\ methods and procedures were not appropriate forthe study type
ie u . in viii-" skiu corrosion protocol used for in vitro skin irritation
assay).
Standards fur Tom
Quality control criteria were not reported and/or inadequate data were
provided to demonstrate validity, acceptability, and reliability of the test
w lieu compared with current standards and guidelines.
Information on preparation and storage was not reported
OK
serious flaws reported with test substance preparation and/or storage
conditions will have critical impacts on dose/concentration estimates and
make the study unusable (e.g., instability of test substance in exposure
media, test substance volatilized rapidly from the open containers that
were used as test vessels).
Preparation and
Storaue oi l est
Suhsiance
Exposure
Characterization
( OIlslslCllCV of
Exposure
Administration
Critical exposure details (e.g., amount of test substance used) were not
reported
OR
exposures were not administered consistently across and/or within study
groups (e.g., 75 mg/cm2and 87 mg/cm2 administered to reconstructed
corneas replicate 1 and replicate 2, respectively, in in vitro eye irritation
test) resulting in serious flaws that make the study unusable.
Reporting of
Doses/Co ncentratio
ns
The exposure doses/concentrations or amounts of test substance were not
reported resulting in serious flaws.
No information on exposure durations) was reported
OR
the exposure duration was not appropriate for the study type and/or
outcome of interest (e.g., 5 hours for reconstructed epidermis in skin
irritation test, 24 hours exposure for bacterial reverse mutation test).
Exposure Duration
B-36
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Domain
Mel rfc
Deseriplion of Serious l'law(s) in l):il;i Source'1
Exposure
Characterization
(continued)
Number of
Exposure Groups
and Dose Spacing
The number of exposure groups and dose/concentration spacing were not
reported
OR
the number of exposure groups and dose/concentration spacing were not
relevant for the assessment (e.g., all concentrations used in an in vitro
mammalian cell micronucleus test were cytotoxic).
Metabolic
Activation
No information on the characterization and use of a metabolic
activation system was reported.
The test model and descripti\ e information were not reported
OR
the test model was not appropriate for evaluation of the specific outcome
of interest (e.g.. bacterial re\ eise mutation assay to evaluate chromosome
aberrations)
Test Model
Test Model
Number per Group
The niimlvr of organisms or tissues per siudy group and/or replicates per
sfudv group were not reported
OR
the number of organisms or tissues per slud> moup and/or replicates per
study group were insufficient lo characterize lo\icological effects (e.g.,
one tissue/test concentration one exposure time lor in vitro skin
corrosion lest, one replicate sir;iin of bacteria exposed in bacterial reverse
mutation assay).
The outcome assessment methodology was not reported
OR
the assessment nielliodolous was not appropriate for the outcome(s) of
interest (c u . cells were e\ alualcd for chromosomal aberrations
immediately al ter e\posure to the test substance instead of after post-
e\pnsure incubation period, cytotoxicity not determined prior to
( 1 )X(i ( 1) expression measurement assay, and labeling antibodies were
not tested on proficiency substances in an in vitro skin sensitization test
in li-('l. \ 1 cells).
()u icoi lie
\ssessnieut
\1etliodolou\
(Juiconie
\ssessnieut
( 'oilsls|CUC> I'l
On Iconic
Assess IIIC III
1 here were large inconsistencies in the execution of study protocols for
outcome assessment across study groups
OK
outcome assessments were not adequately reported for meaningful
interpretation of results.
Sanipliuu \dequacy
Reported sampling was not adequate for the outcome(s) of interest and/or
serious uncertainties or limitations were identified in how the study
carried out the sampling of the outcome(s) of interest (e.g., replicates
from control and test concentrations were evaluated at different times).
Blinding of
Assessors
Information in the study report suggested that the assessment of
subjective outcomes was performed in a biased fashion (e.g., assessors of
subjective outcomes were aware of study groups).
B-37
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Domain
Metric
Description of Serious Flaw(s) in Data Source3
Confounding/
Variable Control
Confounding
Variables in Test
Design and
Procedures
There were significant differences among the study groups with respect
to the strain/batch/lot number of organisms or models used per group or
size and/or quality of tissues exposed (e.g., initial number of viable
bacterial cells were different for each replicate (105 cells in replicate 1,
108 cell in replicate 2, and 103 cells in replicate 3); tissues from two
different lots were used for in vitro skin corrosion test, but the control
batch quality for one lot was outside of the acceptability range).
Confounding
Variables in
Outcomes Unrelated
to Exposure
One or more replicates or groups (i.e., negative and positive controls)
experienced disproportionate growth or reduction in growth unrelated to
exposure (e.g., contamination) such that no outcomes could be assessed.
Data Analysis
Statistical methods, calculation methods, or data manipulation were not
appropriate (e.g.. Student's t-test used to compare two groups in a multi-
group study, parametric test for non-normally distributed data)
OR
statistical analysis was not conducted
AND
data enabling an independent statistical analysis were not provided.
Data
Presentation and
Analysis
Data Interpretation
The reported scoring and/or evaluation criteria were inconsistent with
established practices resulting in the interpretation of data results that are
seriously flawed.
Cytotoxicity Data
Cytotoxicity endpoints were not defined, methods were not described,
and it could not be determined that cytotoxicity was accounted for in the
interpretation of study results.
Reporting of Data
Data presentation was inadequate (e.g., the report did not differentiate
among findings in multiple exposure groups, no scores or frequencies
were reported), or major inconsistencies were present in reporting of
results.
Note:
aIf the metric does not apply to the study type, the flaw will not be applied to determine unacceptability.
Table B-17. Data Quality Criteria for In Vitro Toxicity Studies
Confidence Level
(Score)
Description
Selected
Score
Domain 1. Test Substance
Metric 1. Test Substance Identity
Was the test substance identified definitively (i.e., established nomenclature, CASRN, physical nature,
physiochemical properties, and/or structure reported, including information on the specific form tested (e.g., salt or
base, valence state, isomer, if applicable) for materials that may vary in form)? If test substance was a mixture,
were mixture components and ratios characterized?
High
(score = 1)
The test substance was identified definitively (i.e., established nomenclature,
CASRN, physical nature, physiochemical properties, and/or structure
reported, including information on the specific form tested (e.g., salt or base,
valence state, isomer, (if applicable)) for materials that may vary in form. For
mixtures, the components and ratios were characterized.
Medium
(score = 2)
The test substance and form (if applicable) were identified, and components
and ratios of mixtures were characterized, but there were minor uncertainties
(e.g., minor characterization details were omitted) that are unlikely to have a
substantial impact on results.
B-38
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Low
(score = 3)
The test substance and form (if applicable) were identified, and components
and ratios of mixtures were characterized, but there were uncertainties
regarding test substance identification or characterization that are likely to
have a substantial impact on the results.
Unacceptable
(score = 4)
The test substance identity and form (if applicable) could not be determined
from the information provided (e.g., nomenclature was unclear and CASRN
or structure were not reported)
OR
the components and ratios of mixtures were not characterized.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 2. Test Substance Source
Was the source of the test substance reported, including manufacturer and batch/lot number for materials that may
vary in composition? If synthesized or extracted, was test substance identity verified by analytical methods?
High
(score = 1)
The source of the test substance was reported, including manufacturer and
batch/lot number for materials that might vary in composition, and its identity
was certified by manufacturer and/or verified by analytical methods (melting
point, chemical analysis, etc.).
Medium
(score = 2)
The source of the test substance and/or the analytical verification of a
synthesized test substance was reported incompletely, but the omitted details
are unlikely to have a substantial impact on the results.
Low
(score = 3)
Omitted details on the source of the test substance and/or analytical
verification of a synthesized test substance are likely to have a substantial
impact on the results.
Unacceptable
(score = 4)
The test substance was not obtained from a manufacturer
OR
if synthesized or extracted, analytical verification of the test substance was
not conducted.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 3. Test Substance Purity
Was the purity or grade (i.e., analytical, technical) of the test substance reported and adequate to identify its
toxicological effects? Were impurities identified? Were impurities present in quantities that could influence the
results?
High
(score = 1)
The test substance purity and composition were such that any observed
effects were highly likely to be due to the nominal test substance itself (e.g.,
American Chemical Society grade, analytical grade, reagent grade test
substance or a formulation comprising primarily inert ingredients with small
amount of active ingredient). Impurities, if identified, were not present in
quantities that could influence the results.
B-39
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Medium
(score = 2)
Minor uncertainties or limitations were identified regarding the test substance
purity and composition; however, the purity and composition were such that
observed effects were more likely than not to be due to the nominal test
substance and impurities, if identified, were unlikely to have a substantial
impact on the results.
Low
(score = 3)
Purity and/or grade of test substance were not reported
OR
the percentage of the reported purity was such that the observed effects may
not have been due to the nominal test substance.
Unacceptable
(score = 4)
The nature and quantity of reported impurities were such that study results
were likely to be due to one or more of the impurities.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Domain 2. Test Design
Metric 4. Negative Controls
Was a concurrent negative (untreated, sham-treated, and/or vehicle, as necessary) control group included?
High
(score = 1)
Study authors reported using a concurrent negative control group (untreated,
sham-treated, and/or vehicle, as applicable) in which all conditions equal
except exposure to test substance.
Medium
(score = 2)
Study authors reported using a concurrent negative control group, but all
conditions were not equal to those of treated groups; however, the identified
differences are considered to be minor limitations that are unlikely to have
substantial impact on results.
Low
(score = 3)
Study authors acknowledged using a concurrent negative control group, but
details regarding the negative control group were not reported, and the lack of
details is likely to have a substantial impact on the results.
Unacceptable
(score = 4)
A concurrent negative control group was not included or reported
OR
the reported negative control group was not appropriate (e.g., different cell
lines used for controls and test substance exposure).
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 5. Positive Controls
Was a concurrent positive or proficiency control group included, if applicable, based on study type, and was the
response appropriate in this group (e.g., induction of positive effect)?
Note: This metric is applicable to studies that require a concurrent positive control.
High
(score = 1)
A concurrent positive control or proficiency control group, if applicable, was
used and the intended positive response was induced.
Medium
(score = 2)
A concurrent positive control or proficiency control was used, but there were
minor uncertainties (e.g., minor details regarding control exposure or
response were omitted) that are unlikely to have a substantial impact on
results.
B-40
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Low
(score = 3)
A concurrent positive control or proficiency control was used, but there were
uncertainties regarding the control exposure or response that are likely to
have a substantial impact on results (e.g., the control response was not
described).
Unacceptable
(score = 4)
A concurrent positive control or proficiency group was not used.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 6. Assay Procedures
Were assay methods and procedures (e.g., test conditions, cell density culture media and volumes, pre- and post-
incubation temperatures, humidity, reaction mix, washing/rinsing methods, incubation with amino acids, slide
preparation, instrument used and calibration, wavelengths measured) described in detail and applicable to the
study type?
High
(score = 1)
Study authors described the methods and procedures (e.g., test conditions,
cell density culture media and volumes, pre- and post-incubation
temperatures, humidity, reaction mix, washing/rinsing methods, incubation
with amino acids, slide preparation, instrument used and calibration,
wavelengths measured) used for the test in detail, and the methods and
procedures were applicable for the study type (e.g., protocol for in vitro skin
irritation test was reported).
Medium
(score = 2)
Methods and procedures were partially described and/or cited in another
publication(s), but appeared to be appropriate (e.g., reporting that
"calculations were used for enumerating viable and mutant cells" in a
mammalian cell gene mutation test using Hprt and xprt genes instead of
inclusion of the equations) to the study type, so the omission is unlikely to
have a substantial impact on results.
Low
(score = 3)
The methods and procedures were not well described or deviated from
customary practices (e.g., post-incubation time was not stated in a
mammalian cell gene mutation test using Hprt and xprt genes), and this is
likely to have a substantial impact on results.
Unacceptable
(score = 4)
Assay methods and procedures were not reported
OR
assay methods and procedures were not appropriate for the study type (e.g.,
in vitro skin corrosion protocol used for in vitro skin irritation assay).
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
B-41
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Metric 7. Standards for Test
For assays with established criteria, were the test validity, acceptability, reliability, and/or quality control criteria
reported and consistent with current standards and guidelines? Example acceptability and quality control criteria
for an in vitro skin corrosion test usins the EmSkin™ (SM) model: Acceptability criteria: nesative control ootical
density values between > 0.6 and < 1.5, variability of the positive control replicates should be < 20% of negative
control, difference of viability between 2 tissue replicates should not exceed 30% in the range of 20%-100%
viability and for EDs > 0.3; duality control criteria: Only quality control-accepted tissue batches having an IC\(J
range of 1.0-3.0 mg/mL were used.)
Note: This metric is generally applicable to studies using reconstructed human cells and may not be applicable to
other studies.
High
(score = 1)
The test validity, acceptability, reliability, and/or quality control criteria were
reported and consistent with current standards and guidelines,3 if applicable.
Medium
(score = 2)
Not applicable for this metric.
Low
(score = 3)
Not applicable for this metric.
Unacceptable
(score = 4)
Quality control criteria were not reported and/or inadequate data were
provided to demonstrate validity, acceptability, and reliability of the test
when compared with current standards and guidelines.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Domain 3. Exposure Characterization
Metric 8. Preparation and Storage of Test Substance
Did the study characterize preparation of the test substance and storage conditions? Were the frequency of
preparation and/or storage conditions appropriate to the test substance stability and solubility (if applicable)?
High
(score = 1)
The test substance preparation and/or storage conditions (e.g., test substance
stability, homogeneity, mixing temperature, stock concentration, stirring
methods, centrifugation/filtration aerosol/vapor generation method, storage
conditions) were reported and appropriate (e.g., stability in exposure media
confirmed, volatile test substances prepared and stored in sealed containers)
for the test substance.
Medium
(score = 2)
The test substance preparation and storage conditions were reported, but
minor limitations in the test substance preparation and/or storage conditions
were identified (e.g., test substance formulations were stirred instead of
centrifuged for a specific number of rotations per minute) that are unlikely to
have a substantial impact on results.
Low
(score = 3)
Deficiencies in reporting of test substance preparation, and/or storage
conditions are likely to have a substantial impact on results (e.g., available
information on physical-chemical properties suggests that stability and/or
solubility of test substance in vehicle or culture media may be poor).
B-42
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Information on preparation and storage was not reported
OR
serious flaws reported with test substance preparation and/or storage
conditions will have critical impacts on dose/concentration estimates and
make the study unusable (e.g., instability of test substance in exposure media,
test substance volatilized rapidly from the open containers that were used as
test vessels).
Unacceptable
(score = 4)
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 9. Consistency of Exposure Administration
Were exposures administered consistently across study groups (e.g., consistent application methods and volumes,
control for evaporation)?
High
(score = 1)
Details of exposure administration were reported, and exposures were
administered consistently across study groups in a scientifically sound
manner (e.g., consistent application methods and volumes, control for
evaporation).
Medium
(score = 2)
Details of exposure administration were reported or inferred from the text,
but the minor limitations in administration of exposures (e.g., accidental
mistakes in dosing) that were identified are unlikely to have a substantial
impact on results.
Low
(score = 3)
Details of exposure administration were reported, but deficiencies in
administration of exposures (e.g., non-calibrated instrument used to
administer test substance) that were reported or inferred from the text are
likely to have a substantial impact on results.
Unacceptable
(score = 4)
Critical exposure details (e.g., amount of test substance used) were not
reported
OR
exposures were not administered consistently across and/or within study
groups (e.g., 75 mg/cm2and 87 mg/cm2 administered to reconstructed corneas
replicate 1 and replicate 2, respectively, in in vitro eye irritation test)
resulting in serious flaws that make the study unusable.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 10. Reporting of Doses/Concentrations
Were exposure doses/concentrations or amounts of test substance reported without ambiguity (e.g., point estimate
instead of range, analytical instead of nominal)?
High
(score = 1)
The exposure doses/concentrations or amounts of test substance were
reported without ambiguity (e.g., point estimate instead of range, analytical
instead of nominal).
Medium
(score = 2)
Not applicable for this metric.
Low
(score = 3)
Not applicable for this metric.
B-43
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Unacceptable
(score = 4)
The exposure doses/concentrations or amounts of test substance were not
reported resulting in serious flaws.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 11. Exposure Duration
Was the exposure duration (e.g., minutes, hours, days) reported and appropriate for this study type and/or
outcome(s) of interest?
High
(score = 1)
The exposure duration (e.g., minutes, hours, days) was reported and
appropriate for the study type and/or outcome(s) of interest (e.g., 60-minute
exposure for reconstructed epidermis in skin irritation test, 48-72-hour
exposure for bacterial reverse mutation assay).
Medium
(score = 2)
Duration(s) of exposure differed slightly from current standards and
guidelines3 for studies of this type (e.g., 65 minutes for reconstructed
epidermis in skin irritation test), but the differences are unlikely to have a
substantial impact on results.
Low
(score = 3)
Durations) of exposure were not clearly stated (e.g., exposure duration was
described only in qualitative terms) or durations) differed significantly from
studies of the same or similar types. These deficiencies are likely to have a
substantial impact on results.
Unacceptable
(score = 4)
No information on exposure durations) was reported
OR
the exposure duration was not appropriate for the study type and/or outcome
of interest (e.g., 5 hours for reconstructed epidermis in skin irritation test, 24-
hour exposure for bacterial reverse mutation test).
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 12. Number of Exposure Groups and Concentrations Spacing
Were the number of exposure groups and dose/concentration spacing justified by study authors (e.g., based on
study type, range-finding study, and/or cytotoxicity studies) and adequate to address the purpose of the study (e.g.,
to evaluate dose-response relationships, inform MOA/AOP)?
High
(score = 1)
The number of exposure groups and dose/concentration spacing were
justified by study authors (e.g., based on study type, range-finding study,
and/or cytotoxicity studies) and considered adequate to address the purpose
of the study (e.g., to evaluate dose-response relationships, inform
MOA/AOP).
Medium
(score = 2)
There were minor limitations regarding the number of exposure groups
and/or dose/concentration spacing, but the number of exposure groups and
spacing of exposure levels were adequate to show results relevant to the
outcome of interest (e.g., observation of a dose-response relationship) and the
concerns are unlikely to have a substantial impact on results.
Low
(score = 3)
There were deficiencies regarding the number of exposure groups and/or
dose/concentration spacing (e.g., one bacterial strain exposed to two
concentrations of the test substance in bacterial reverse mutation assay), and
these concerns likely had a substantial impact on interpretation of the results.
B-44
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
Selected
(Score)
Description
Score
Unacceptable
(score = 4)
The number of exposure groups and dose/concentration spacing were not
reported
OR
the number of exposure groups and dose/concentration spacing were not
relevant for the assessment (e.g., all concentrations used in an in vitro
mammalian cell micronucleus test were cytotoxic).
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 13. Metabolic Activation (if Applicable)
Were exposures conducted in the presence and absence of a metabolic activation system if applicable, for the
study type? Were the source, method of preparation, concentration or volume in final culture, and quality control
information on the metabolic activation system reported?
High
(score = 1)
Study authors reported that exposures were conducted in the presence of
metabolic activation, and the type and source, method of preparation,
concentration or volume in final culture, and quality control information of
the metabolic activation system were described.
Medium
(score = 2)
The presence of a commonly used metabolic activation system (e.g., aroclor-,
ethanol-, or phenobarbitial/p-naphthoflavone-induced rat, hamster, or mice
liver cells) was reported in the study; however, some details regarding type,
composition mix, concentration, or quality control information were not
described. These omissions are unlikely to have a substantial impact on the
results.
Low
(score = 3)
The presence of a metabolic activation system was reported in the study, but
the system described was not validated (e.g., rigorous testing to ensure that it
suitable for the purpose for which it is used) or comparable to commonly
used systems (e.g., aroclor-, ethanol-, orphenobarbitial/p-naphthoflavone-
induced rat, hamster, or mice liver cells).
Unacceptable
No information on the characterization and use of a metabolic activation
(score = 4)
system was reported.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Domain 4. Test Model
Metric 14. Test Model
Were the test models (e.g., cell types or lines, tissue models) and descriptive information (e.g., tissue origin,
number of passages, karyotype features, doubling times, donor information, biomarkers) reported? Was the test
model from a commercial source or an in-house culture? Was the model routinely used for the outcome of interest
(e.g., Chinese hamster ovary cells for micronucleus formation)?
High
(score = 1)
The test model (e.g., cell types or lines, tissue models) and descriptive
information (e.g., tissue origin, number of passages, karyotype features,
doubling times, donor information, biomarkers) were reported, the test model
was obtained from a commercial source or laboratory-maintained culture, and
the test model was routinely used for the outcome of interest (e.g., Chinese
hamster ovary cells for micronucleus formation).
B-45
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Medium
(score = 2)
The test model was reported along with limited descriptive information. The
test model was routinely used for the outcome of interest. Reporting
limitations are unlikely to have a substantial impact on results.
Low
(score = 3)
The test model was reported but no additional details were reported
AND/OR
the test model was not routinely used for the outcome of interest (e.g., feline
cell line for micronucleus formation). This is likely to have a substantial
impact on results.
Unacceptable
(score = 4)
The test model and descriptive information were not reported
OR
the test model was not appropriate for evaluation of the specific outcome of
interest (e.g., bacterial reverse mutation assay to evaluate chromosome
aberrations).
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 15. Number per Group
Was the number of organisms or tissues per study group and/or replicates per study group reported and appropriate
for the study type and outcome analysis?
High
(score = 1)
The number of organisms or tissues per study group and/or number of
replicates per study group were reported and were appropriate3 for the study
type and outcome analysis and consistent with studies of the same or similar
type (e.g., at least two replicates/test substance or three different exposure
times for in vitro skin corrosion test; three replicates/strain of bacteria in
bacterial reverse mutation assay).
Medium
(score = 2)
The number of organisms or tissues per study group and/or replicates per
study group were reported but were lower than the typical number used in
studies of the same or similar type (e.g., three replicates/strain of bacteria in
bacterial reverse mutation assay), but they were sufficient for analysis and
unlikely to have a substantial impact on results.
Low
(score = 3)
The number of organisms or tissues per study group and/or replicates per
study group were reported but were less than recommended by current
standards and guidelines3 (e.g., one tissue/test concentration/exposure time
for in vitro skin corrosion test). This is likely to have a substantial impact on
results.
Unacceptable
(score = 4)
The number of organisms or tissues per study group and/or replicates per
study group were not reported
OR
the number of organisms or tissues per study group and/or replicates per
study group were insufficient to characterize toxicological effects (e.g., one
tissue/test concentration/one exposure time for in vitro skin corrosion test,
one replicate/strain of bacteria exposed in bacterial reverse mutation assay).
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
B-46
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Domain 5. Outcome Assessment
Metric 16. Outcome Assessment Methodology
Did the outcome assessment methodology address or report the intended outcome(s) of interest? Was the outcome
assessment methodology (including endpoints and timing of assessment) sensitive for the outcome(s) of interest
(e.g., measured endpoints that are able to detect a true effect)?
High
(score = 1)
The outcome assessment methodology addressed or reported the intended
outcome(s) of interest and was sensitive for the outcome(s) of interest.
Medium
(score = 2)
The outcome assessment methodology used only partially addressed or
reported the intended outcomes(s) of interest (e.g., mutation frequency
evaluated in the absence of cytotoxicity in a gene mutation test), but minor
uncertainties are unlikely to have a substantial impact on results.
Low
(score = 3)
Significant deficiencies in the reported outcome assessment methodology
were identified (e.g., optimum time for expression of chromosomal
aberrations after exposure to test compound was not determined)
OR
due to incomplete reporting, it was unclear whether methods were sensitive
for the outcome of interest. This is likely to have a substantial impact on
results.
The outcome assessment methodology was not reported
OR
the assessment methodology was not appropriate for the outcome(s) of
interest (e.g., cells were evaluated for chromosomal aberrations immediately
after exposure to the test substance instead of after post-exposure incubation
period).
Unacceptable
(score = 4)
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 17. Consistency of Outcome Assessment
Was the outcome assessment carried out consistently (i.e., using the same protocol) across study groups (e.g.,
assessment at the same time after initial exposure in all study groups)?
High
(score = 1)
Details of the outcome assessment protocol were reported, and outcomes
were assessed consistently across study groups (e.g., at the same time after
initial exposure) using the same protocol in all study groups.
Medium
(score = 2)
There were minor differences in the timing of outcome assessment across
study groups, or incomplete reporting of minor details of outcome assessment
protocol execution but these uncertainties or limitations are unlikely to have
substantial impact on results.
Low
(score = 3)
Details regarding the execution of the study protocol for outcome assessment
(e.g., timing of assessment across groups) were not reported, and these
deficiencies are likely to have a substantial impact on results.
Unacceptable
(score = 4)
There were large inconsistencies in the execution of study protocols for
outcome assessment across study groups
OR
outcome assessments were not adequately reported for meaningful
interpretation of results.
Not rated/applicable
B-47
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 18. Sampling Adequacy
Was the reported sampling adequate for the outcome(s) of interest, including number of evaluations per exposure
group and endpoint (e.g., number of replicates/slides/cells/metaphases evaluated per test concentration)?
High
(score = 1)
The study reported adequate sampling for the outcome(s) of interest,
including number of evaluations per exposure group and endpoint (e.g.,
number of replicates/slides/cells/metaphases (at least 300 well-spread
metaphases scored/concentration in a chromosome aberration test)).
Medium
(score = 2)
Details regarding sampling for the outcome(s) of interest were reported, but
minor limitations were identified in the reported sampling of the outcome(s)
of interest, but those are unlikely to have a substantial impact on results.
Low
(score = 3)
Details regarding sampling of outcomes were not fully reported, and the
omissions are likely to have a substantial impact on results.
Unacceptable
(score = 4)
Reported sampling was not adequate for the outcome(s) of interest, and/or
serious uncertainties or limitations were identified in how the study carried
out the sampling of the outcome(s) of interest (e.g., replicates from control
and test concentrations were evaluated at different times).
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 19. Blinding of Assessors
Were investigators assessing subjective outcomes (i.e., those evaluated using human judgment) blinded to
treatment group?
This metric is not rated/applicable if no subjective outcomes were assessed (i.e., only automated measurements
were included and human judgment was not applied).
High
(score = 1)
The study explicitly reported that investigators assessing subjective outcomes
(i.e., those evaluated using human judgment) were blinded to treatment group
or that quality control/quality assurance methods were followed in the
absence of blinding.
Medium
(score = 2)
The study reported that blinding was not possible, but steps were taken to
minimize bias (e.g., knowledge of study group was restricted to personnel not
assessing subjective outcome), and this minor uncertainty is unlikely to have
a substantial impact on results.
Low
(score = 3)
The study did not report whether assessors were blinded to treatment group
for subjective outcomes, and this deficiency is likely to have a substantial
impact on results.
Unacceptable
(score = 4)
Information in the study report suggested that the assessment of subjective
outcomes was performed in a biased fashion (e.g., assessors of subjective
outcomes were aware of study groups).
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
B-48
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Domain 6. Confounding/Variable Control
Metric 20. Confounding Variables in Test Design and Procedures
Were there confounding differences among the study groups (in the strain/batch/lot number of organisms); models
used (per group, size, and/or quality of tissues exposed); or lot of test substance used that could influence the
outcome assessment?
High
(score = 1)
There were no differences reported among study group parameters (e.g., test
substance lot or batch, strain/batch/lot number of organisms or models used
per group or size, and/or quality of tissues exposed) that could influence the
outcome assessment.
Medium
(score = 2)
Minor differences were reported in initial conditions that are unlikely to have
a substantial impact on results (e.g., tissues from two different lots were used
for in vitro skin corrosion test, and quality control data were similar for both
lots).
Low
(score = 3)
Initial strain/batch/lot number of organisms or models used per group, size,
and/or quality of tissues exposed was not reported. These deficiencies are
likely to have a substantial impact on results.
Unacceptable
(score = 4)
There were significant differences among the study groups with respect to the
strain/batch/lot number of organisms or models used per group or size and/or
quality of tissues exposed (e.g., initial number of viable bacterial cells were
different for each replicate (105 cells in replicate 1, 108 cell in replicate 2, and
103 cells in replicate 3), tissues from two different lots were used for in vitro
skin corrosion test, but the control batch quality for one lot was outside of the
acceptability range).
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 21. Confounding Variables in Outcomes Unrelated to Exposure
Were there differences among the study groups unrelated to exposure to test substance (e.g., contamination) that
could influence the outcome assessment? Did the test material interfere in the assay (e.g., altering fluorescence or
absorbance, signal quenching by heavy metals, altering pH, solubility or stability issues)?
High
(score = 1)
There were no reported differences among the study replicates or groups in
test model unrelated to exposure (e.g., contamination), and the test substance
did not interfere with the assay (e.g., signal quenching by heavy metals).
Medium
(score = 2)
Authors reported that one or more replicates or groups experienced
disproportionate outcomes unrelated to exposure (e.g., contamination), but
data from the remaining exposure replicates or groups were valid and
unlikely to have a substantial impact on results
OR
data on experienced disproportionate outcomes unrelated to exposure were
not reported because only substantial differences among groups were noted
(as indicated by study authors).
OR
the test material interfered in the assay, but the interference did not cause
substantial differences among the groups.
Low
(score = 3)
Data on outcome differences unrelated to exposure were not reported for each
study replicate or group. Assay interference was present or inferred, resulting
in large variabilities among the groups. The absence of this information is
likely to have a substantial impact on results.
B-49
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Unacceptable
(score = 4)
One or more replicates or groups (i.e., negative and positive controls
experienced disproportionate growth or reduction in growth unrelated to
exposure (e.g., contamination), or assay interference occurred such that no
outcomes could be assessed.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Domain 7. Data Presentation and Analysis
Metric 22. Data Analysis
Were statistical methods, calculations methods, and/or data manipulation clearly described and appropriate for
dataset(s)?
High
(score = 1)
Statistical methods, calculation methods, and/or data manipulation were
clearly described and presented for dataset(s) (e.g., frequencies of
chromosomal aberrations were statistically analyzed across groups, trend test
used to determine dose relationships, or results compared to historical
negative control data).
OR
no statistical analyses, calculation methods, and/or data manipulation were
conducted, but sufficient data were provided to conduct an independent
statistical analysis.
Medium
(score = 2)
Statistical analysis was described with some omissions that would unlikely
have a substantial impact on results.
Low
(score = 3)
Statistical analysis was not described clearly, and this deficiency is likely to
have a substantial impact on results.
Unacceptable
(score = 4)
Statistical methods were not appropriate (e.g.. Student's t-test used to
compare two groups in a multi-group study, parametric test for non-normally
distributed data)
OR
statistical analysis was not conducted
AND
data were not provided, preventing an independent statistical analysis.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 23. Data Interpretation
Were the scoring and/or evaluation criteria reported and consistent with standards and guidelines?
High
(score = 1)
Study authors reported the scoring and/or evaluation criteria (e.g., for
determining negative, positive, and equivocal outcomes) for the test and these
were consistent with established practices.3
Medium
(score = 2)
Scoring and/or evaluation criteria were partially reported (e.g., evaluation
criteria were reported following 3- and 60-minute exposures, but not for 240-
minute exposure in in vitro skin corrosion test), but the omissions are
unlikely to have a substantial impact on results.
Low
(score = 3)
Scoring and/or evaluation criteria were not reported, and the omissions are
likely to have a substantial impact on interpretation of the results.
B-50
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Unacceptable
(score = 4)
The reported scoring and/or evaluation criteria were inconsistent with
established practices, resulting in the interpretation of data results that are
seriously flawed.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 24. Cytotoxicity Data
Were cytotoxicity endpoints defined, if necessitated by study type, and were methods for measuring cytotoxicity
described and commonly used for assessment3?
High
(score = 1)
Study authors defined cytotoxicity endpoints (e.g., cell integrity, apoptosis,
necrosis, color induction, cell viability, mitotic index), and the methods for
measuring cytotoxicity were clearly described and commonly used for
assessment.
Medium
(score = 2)
Cytotoxicity endpoints were defined and methods of measurement were
partially reported, but the omissions are unlikely to have substantial impact
on study results.
Low
(score = 3)
Cytotoxicity endpoints were defined, but the methods of measurements were
not fully described or reported, and the omissions are likely to have a
substantial impact on the study results.
Unacceptable
(score = 4)
Cytotoxicity endpoints were not defined, methods were not described, and it
could not be determined that cytotoxicity was accounted for in the
interpretation of study results.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Metric 25. Reporting of Data
Were the data for all outcomes presented? Were data reported by exposure group?
High
(score = 1)
Data for exposure-related findings were presented for all outcomes by
exposure group. Negative findings were reported qualitatively or
quantitatively.
Medium
(score = 2)
Data for exposure-related findings were reported for most, but not all,
outcomes by exposure group (e.g., sensitization percentages reported in the
absence of incidence data). The minor uncertainties in outcome reporting are
unlikely to have substantial impact on results.
Low
(score = 3)
Data for exposure-related findings were not shown for each study group, but
results were described in the text and/or data were only reported for some
outcomes. These deficiencies are likely to have a substantial impact on
results.
Unacceptable
(score = 4)
Data presentation was inadequate (e.g., the report did not differentiate among
findings in multiple exposure groups, no scores or frequencies were
reported), or major inconsistencies were present in reporting of results.
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
B-51
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confidence Level
(Score)
Description
Selected
Score
Domain 8. Other (Apply as Needed)
Metric:
High
(score = 1)
Medium
(score = 2)
Low
(score = 3)
Unacceptable
(score = 4)
Not rated/applicable
Reviewer's comments
(Document concerns, uncertainties, limitations, and deficiencies and any
additional comments that may highlight study strengths or important
elements such as relevance)
Note:
Tor comparison purposes, current standards and guidelines may be reviewed at http://www.oecd-ilibrarv.org/environment/oecd-
guidehnes-for-the-testing-of-chemicals-section-4-health-effects 20745788; https://www.epa.gov/test-guidelines-pesticides-and-
toxic-substances;
https://www.fda.gov/Food/GuidanceRegulation/GuidanceDocumentsRegulatorvInfonnation/IngredientsAdditivesGRASPackagi
ng/ucm2006826.htm#TQC.
References
Cooper, G.S., R.M. Lunn, M. Agerstrand, B.S. Glenn, A.D. Kraft, A.M. Luke, J.M., Ratcliffe.
2016. Study sensitivity: Evaluating the ability to detect effects in systematic reviews of
chemical exposures. Environment International 92-93:605-610.
http://dx.doi.Org/10.1016/i.envint.2016.03.017.
Crissman, J.W., D.G. Goodman, P.K. Hildebrandt, R.R. Maronpot, D.A. Prater, J.H. Riley, W.J.
Seaman, and D.C. Thake. (2004). Best practices guideline: Toxicologic histopathology.
Toxicologic Pathology 32:126-131. http://dx.doi.org/10.1080/0192623049Q268756.
EC (European Commission). 2018. ToxRTool - Toxicological data Reliability assessment Tool.
https://eurl-ecvam.irc.ec.europa.eu/about-ecvam/archive-publications/toxrtool.
ECHA (European Chemicals Agency). 2011. Guidance on Information Requirements and
Chemical Safety Assessment. (ECHA-2011-G-13-EN). European Chemicals Agency.
Hartling, L., M. Hamm, A. Milne, B. Vandermeer, P.L. Santaguida, M. Ansari, A. Tsertsvadze,
S. Hempel, P. Shekelle, andD.M. Dryden. 2012. Validity and Inter-rater Reliability
Testing of Quality Assessment Instruments. Publication No. 12-EHC039-EF. Agency for
Healthcare Research Quality Rockville, MD.
Hooijmans, C., R. de Vries, M. Leenaars, and M. Ritskes-Hoitinga. 2010. The Gold Standard
Publication Checklist (GSPC) for improved design, reporting and scientific quality of
animal studies GSPC versus ARRIVE guidelines. Laboratory Animals 45(1):61.
http://dx.doi.org/10.1258/la.2010.01013Q.
B-52
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Hooijmans, C.R., M.M. Rovers, R.B.M. de Vries, M. Leenaars, M. Ritskes-Hoitinga, and M.W.
Langendam. 2014. SYRCLE's risk of bias tool for animal studies. BMC Medical
Research Methodology 14( 1): 43. http://dx. doi. org/10.1186/147 1 -2288. X. L~J i •
Koustas, E., J. Lam, P. Sutton, P.I. Johnson, D.S. Atchley, S. Sen, K.A. Robinson, D.A. Axelrad,
and T.J. Woodruff. 2014. The Navigation Guide—Evidence-based medicine meets
environmental health: Systematic review of nonhuman evidence for PFOA effects on
fetal growth (Review). Environmental Health Perspectives 122(10): 1015-1027.
http://dx.doi.ore/10.12.89/ehp.1307177.
Kushman, M.E., A.D. Kraft, K.Z. Guyton, W.A. Chin. S I. Malcris, and I. Rusyn. 2013. A
systematic approach for identifying and presenting mechanistic evidence in human health
assessments. Regulatory Toxicology andI'/iarmaco/ogy (->7(2):266-277.
http://dx.doi.orp hMto i.yrtph.f ' '
Lynch, H.N., J.E. Goodman, J.A. Tabony. and I. R Rhomberg. 20lo Systematic comparison of
study quality criteria. Regulatory I oxicology and Pharmacology 76 IS 7-198.
https://doi.Org/l '/j.vrtph.2015.
Moermond, C.T.A., R. Kase, M korkaric. and M. Ageisliand. 2016. CRED: Criteria for
reporting and evaluating ecotoxicitv data Environmental Toxicology and Chemistry
35(5): 1297-1309. http " '
NTP (National Toxicology Program) 2d I 5a Ilamlbook for ('onducting a Literature-based
Health Assessment I sing ()ll. 1 /. \ppvoach for Systematic Review and Evidence
Integration I S Department of I leal ih and Human Services, National Toxicology
Program t/pubs/handbooki atL 08.pdf.
NTP (National Toxicology Program) 2< >1 5h ()/1AT Risk of Bias Rating Toolfor Human and
Animal Similes. I .S. Pepartmeni of / lealth and Human Services, National Toxicology
Trogram, Office of Health. issessment and Translation (OHAT).
M.".- -/ntp.meti; ubs/riskofbiastool 508.pdf.
Samuel, G (). S Hoffmann. R A. Wright, M.M. Lalu, G. Patlewicz, R.A. Becker, G.L.
DeGeorge. D. Fergnsson, T. Hartung, R.J. Lewis, M.L. Stephens. 2016. Guidance on
assessing the methodological and reporting quality of toxicologically relevant studies: A
scoping re\ iew Environment International 92-93:630-646.
¦2016.03.010.
USEPA. (U.S. Environmental Protection Agency). 2006. Approaches for the Application of
Physiologically Based Pharmacokinetic (PBPK) Models and Supporting Data in Risk
Assessment (FinalReport). EPA/600/R-05/043F. U.S. Environmental Protection Agency,
Office of Research and Development, National Center for Environmental Assessment,
Washington, DC. http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=157668.
USEPA. (U.S. Environmental Protection Agency). 2018. Strategic Plan to Promote the
Development and Implementation of Alternative Test Methods Within the TSCA Program.
U.S. Environmental Protection Agency, Office of Chemical Safety and Pollution
B-53
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Prevention, Washington, DC. https://www.epa.gov/sites/production/files/2018-
06/docum ents/epa alt strat plan 6-20-18 clean final.pdf.
Data Evaluation and Scoring Tables
The data evaluation and scoring tables are provided in the subsequent pages for the studies
supporting the development of toxicity values for HFPO dimer acid and dimer acid ammonium
salt (CASRN 13252-13-6 and CASRN 62037-80-3).
B-54
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
DuPont-17751-1579 RV1: E.I. du Pont de Nemours and Company (2009). Cross-Species Comparison of FRD-
902 Plasma Pharmacokinetics in the Rat and Primate Following Intravenous Dosing. Test Guideline Not
Identified. Study conducted by E.I. du Pont de Nemours and Company (Original Report Completed: December
8, 2008; Report Revision 1 Completed: February 2, 2009), Newark, Delaware.
Note:
This study is evaluated in 2 sheets; the table below is for the monkey study. The primate (monkey) portion had
additional information on the study protocol attached as Appendix A. The rat portion did not have the
protocol information attached.
Number of Hours
for Review:
1
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
HIGH
1
2
2
2. Test substance
source
HIGH
Attached a certificate of analysis
from DuPont Labs.
1
1
1
3. Test substance
purity
MEDIUM
82.60%
2
1
2
Test Setup
4. Negative controls
Not rated
Not typically used in TK studies.
0
N/A
N/A
5. Negative control
responses
Not rated
Not typically used in TK studies.
0
N/A
N/A
6. Positive controls
Not rated
Not typical in TK studies.
0
N/A
N/A
7. Randomized
allocation
MEDIUM
The protocol in Appendix A stated
"no randomization necessary" for
primate study. The animals were
within +/- 20% of mean weight of
the group, as per OECD TG 417, and
all described as healthy so do not
expect this has substantial impact
on results.
2
1
2
Exposure
Characterization
8. Preparation and
storage of test
substance
HIGH
Prepared before administration of
single dose.
1
1
1
9. Consistency of
exposure
administration
HIGH
1
1
1
10. Reporting of
doses/
concentrations
HIGH
Administered as i.v. dose.
1
2
2
11. Exposure
frequency and
duration
Not rated
A single dose.
0
N/A
N/A
B-55
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Exposure
Characterization
(continued)
12. Number of
exposure groups
and dose spacing
HIGH
Used a single dose group for
primates; typical in a TK study.
1
1
1
13. Exposure Route
and Method
HIGH
i.v. is relevant for this study to look
at ADME.
1
1
1
Test Organisms
14. Test Animal
Characteristics
HIGH
Adequate for primates.
1
2
2
15. Consistency of
Animal Conditions
HIGH
Adequate reporting for primates in
the protocol (Appendix A).
1
1
1
16. Number per
Group
HIGH
3 animals per sex were used. As per
OCED 417: "the use of both sexes
(four males and four females) is
strongly recommended", however 3
animals are typical and sufficient
for the statistical analysis used,
therefore unlikely to have a
substantial impact on results.
1
1
1
Outcome
Assessment
17. Outcome
Assessment
Methodology
HIGH
The measurement of parent
chemical in blood was described
(metabolite measurement not
needed based on earlier in vitro
study). Analytical method high-
performance liquid
chromatography adequately
described. Level of quantification
(LOQ) documented.
1
2
2
18. Consistency of
outcome
assessment
HIGH
1
1
1
19. Sampling
adequacy
HIGH
Time course data for blood were
presented. Sampling continued
sufficiently beyond the time blood
concentration was below LOQ.
1
1
1
20. Blinding of
assessors
Not rated
The outcome (blood concentration)
is not subjective. Blinding is not
typical for a TK study.
0
N/A
N/A
B-56
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confounding/
Variable Control
21. Confounding
variables in test
setup and
procedures
HIGH
BWs were within +/- 20%, i.v.
administration so no palatability
issues. No reported differences in
respiratory rate. Note
concentrations were detected >
LOQ at time zero, suggesting some
other exposure prior to dosing, or
possibly contamination in the
method.
1
2
2
22. Health
outcomes unrelated
to exposure
HIGH
Health outcomes for primates were
reported and these effects are not
expected to impact the results of
the TK study.
1
1
1
Data Presentation
and Analysis
23. Statistical
methods
MEDIUM
Results of minimal statistical
analysis were provided, but the
method was not described. The raw
data are provided.
2
1
2
24. Reporting of
data
HIGH
All raw data provided.
1
2
2
Sum of scores:
25
28
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
1.1
Overall Quality Level:
HIGH
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-57
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
DuPont-17751-1579 RV1: E.I. du Pont de Nemours and Company (2009). Cross-Species Comparison of FRD- 902
Plasma Pharmacokinetics in the Rat and Primate Following Intravenous Dosing. Test Guideline Not Identified.
Study conducted by E.I. du Pont de Nemours and Company (Original Report Completed: December 8, 2008;
Report Revision 1 Completed: February 2, 2009), Newark, Delaware.
Note:
This study is evaluated in 2 sheets; the table below is for the rat study. The primate (monkey) portion had
additional information on the study protocol attached as Appendix A. The rat portion did not have the protocol
information attached.
Number of Hours
for Review:
1
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
HIGH
1
2
2
2. Test substance
source
HIGH
Attached a certificate of analysis
from DuPont Labs.
1
1
1
3. Test substance purity
MEDIUM
82.6%
2
1
2
Test Setup
4. Negative controls
Not rated
Not typically used in TK studies.
0
N/A
N/A
5. Negative control
responses
Not rated
Not typically used in TK studies.
0
N/A
N/A
6. Positive controls
Not rated
Not typical in TK studies.
0
N/A
N/A
7. Randomized
allocation
MEDIUM
Not reported. Rat BWs were not
given. Expect only healthy animals
of similar age were used, so not
expecting this has substantial
impact on results.
2
1
2
Exposure
8. Preparation and
storage of test
substance
HIGH
Prepared before administration of
single dose.
1
1
1
9. Consistency of
exposure
administration
HIGH
1
1
1
10. Reporting of doses/
concentrations
HIGH
Administered as i.v. dose.
1
2
2
11. Exposure frequency
and duration
Not rated
A single dose.
0
N/A
N/A
12. Number of
exposure groups and
dose spacing
HIGH
Used two dose groups, reasonable
dose levels.
1
1
1
13. Exposure route and
method
HIGH
i.v. is relevant for this study to look
at ADME
1
1
1
B-58
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Test Organisms
14. Test animal
characteristics
MEDIUM
The sex and strain are appropriate
and from the in-house colony.
Health status, age, BW not
reported. Do not expect this will
have substantial impact on the
outcomes.
2
2
4
15. Consistency of
animal conditions
MEDIUM
Not reported, but expect they were
similar to conditions reported for
primates and not likely to impact
outcomes.
2
1
2
16. Number per group
HIGH
3 animals per sex were used. As per
OCED 417: "the use of both sexes
(four males and four females) is
strongly recommended", however 3
animals are typical and sufficient
for the statistical analysis used,
therefore unlikely to have a
substantial impact on results.
1
1
1
Outcome
Assessment
17. Outcome
assessment
methodology
HIGH
The measurement of parent
chemical in blood was described
(metabolite measurement not
needed based on earlier in vitro
study). Analytical method high-
performance liquid
chromatography adequately
described. Level of quantification
(LOQ) documented.
1
2
2
18. Consistency of
outcome assessment
HIGH
1
1
1
19. Sampling adequacy
HIGH
Time course data for blood were
presented. Sampling continued
sufficiently beyond the time blood
concentration was below LOQ.
1
1
1
20. Blinding of
assessors
Not rated
The outcome (blood concentration)
is not subjective. Blinding is
probably not typical for a TK study.
0
N/A
N/A
Confounding/
Variable Control
21. Confounding
variables in test setup
and procedures
MEDIUM
BWs, health status, respiratory
rates not reported; i.v.
administration so no palatability
issues. Note concentrations were
detected > LOQ at time zero,
suggesting some other exposure
prior to dosing, or possibly
contamination in the method.
2
2
4
B-59
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
22. Health outcomes
unrelated to exposure
MEDIUM
Not reported. Based on dose levels
and known effects from acute
single dose tox studies health
effects are not expected to occur
that would impact the results of the
TK study.
2
1
2
Data Presentation
and Analysis
23. Statistical methods
MEDIUM
Results of minimal statistical
analysis were provided, but the
method was not described. The raw
data are provided.
2
1
2
24. Reporting of data
HIGH
All raw data provided.
1
2
2
Sum of scores:
25
34
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
1.4
Overall Quality Level:
HIGH
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-60
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
DuPont-18405-1017 RV1: E.I. du Pont de Nemours and Companv (2011). H-28548: Absorption, Distribution,
Metabolism, and Elimination in the Rat. US EPA OPPTS 870.7485. Study conducted by E.I. du Pont de Nemours
and Company (Original Report Completed: November 3, 2010; Report Revision 1
Completed: April 21, 2011), Newark, Delaware and Wilmington, Delaware.
Number of Hours
for Review:
1
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable, or
Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
HIGH
Addressed on page 42 / 61.
1
2
2
2. Test substance
source
HIGH
Addressed on page 11 / 61.
1
1
1
3. Test substance
purity
MEDIUM
Addressed on page 42 / 61. (84%)
2
1
2
Test Setup
4. Negative controls
Not Rated
Not applicable to this study design.
0
N/A
N/A
5. Positive controls
Not Rated
Not applicable to this study design.
0
N/A
N/A
6. Assay procedures
HIGH
Addressed on page 13 / 61.
1
1
1
7. Standards for test
HIGH
Addressed on page 15 / 61.
1
1
1
Exposure
Characterization
8. Preparation and
storage of test
substance
HIGH
Addressed on page 15 / 61.
1
1
1
9. Consistency of
exposure
administration
Not Rated
Not applicable to this study design.
0
N/A
N/A
10. Reporting of
doses/
concentrations
HIGH
Addressed on page 12 / 61
1
2
2
11. Exposure
duration
Not Rated
Not applicable to this study design.
0
N/A
N/A
12. Number of
exposure groups
and dose spacing
Not Rated
Not applicable to this study design.
0
N/A
N/A
13. Metabolic
activation
Not Rated
Not applicable to this study design.
0
N/A
N/A
Test Model
14. Test model
Not Rated
Not applicable to this study design.
0
N/A
N/A
15. Number per
group
HIGH
Addressed on page 9 / 61.
1
1
1
B-61
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Outcome
Assessment
16. Outcome
assessment
methodology
HIGH
Addressed on page 14 / 61.
1
2
2
17. Consistency of
outcome
assessment
HIGH
Addressed on page 14 / 61.
1
1
1
18. Sampling
adequacy
HIGH
Addressed on page 14 / 61.
1
2
2
19. Blinding of
assessors
Not Rated
Not applicable to this study design.
0
N/A
N/A
Confounding/
Variable Control
20. Confounding
variables in test
setup and
procedures
HIGH
Addressed on page 11 / 61.
1
2
2
21. Outcomes
unrelated to
exposure
Not Rated
Not applicable to this study design.
0
N/A
N/A
Data Presentation
and Analysis
22. Data analysis
HIGH
Addressed on page 21 / 61.
1
1
1
23. Data
interpretation
HIGH
Addressed on page 22 / 61.
1
2
2
24. Cytotoxicity data
Not Rated
Not applicable to this study design.
0
N/A
N/A
25. Reporting of
data
HIGH
Individual data provided for all test
animals.
1
2
2
Sum of scores:
22
23
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
1.0
Overall Quality Level:
HIGH
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-62
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
DuPont-18647-1017 RV1: E.I. du Pont de Nemours and Company (2011). H-28548: Absorption, Distribution,
Metabolism, and Elimination in the Mouse. US EPA OPPTS 870.7485. Study conducted by E.I. du Pont de
Nemours and Company (Original Report Completed: November 3, 2010; Report Revision 1 Completed: April 21,
2011), Newark, Delaware and Wilmington, Delaware.
Number of Hours
for Review:
2
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
HIGH
1
2
2
2. Test substance source
HIGH
1
1
1
3. Test substance purity
HIGH
84% adjusted for purity and CoA
attached.
1
1
1
Test Setup
4. Negative controls
Not rated
Not typically used in TK studies.
Note concentrations were
detected > level of quantification
(LOQ) at time zero, suggesting
some other exposure prior to
dosing, or possibly contamination
in the method (although very low
concentrations)?
0
N/A
N/A
5. Negative control
responses
Not rated
Not typically used in TK studies.
0
N/A
N/A
6. Positive controls
Not rated
Not typical in TK studies.
0
N/A
N/A
7. Randomized
allocation
HIGH
Adequately described.
1
1
1
Exposure
Characterization
8. Preparation and
storage of test substance
HIGH
1
1
1
9. Consistency of
exposure administration
HIGH
1
1
1
10. Reporting of doses/
concentrations
HIGH
1
2
2
11. Exposure frequency
and duration
HIGH
Used a single dose; appropriate for
study type.
1
1
1
12. Number of exposure
groups and dose spacing
HIGH
Used a single dose group; typical in
a TK study.
1
1
1
13. Exposure route and
method
HIGH
Oral gavage is relevant and typical
forTK studies.
1
1
1
B-63
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
14. Test animal
characteristics
HIGH
1
2
2
Test Organisms
15. Consistency of
animal conditions
HIGH
1
1
1
16. Number per group
HIGH
5 per group were used; this is
adequate as described in OECD
TG417.
1
1
1
17. Outcome assessment
methodology
HIGH
The measurement of parent
chemical in blood was described
(metabolite measurement not
needed based on earlier in vitro
study indicating no metabolism).
Analytical method high-
performance liquid
chromatography adequately
described. LOQ documented.
1
2
2
Outcome
Assessment
18. Consistency of
outcome assessment
HIGH
1
1
1
19. Sampling adequacy
HIGH
Time course data for blood were
analyzed and sampling continued
sufficiently beyond the time blood
concentration was below LOQ.
1
1
1
20. Blinding of assessors
Not rated
The outcome is a blood
concentration; it is not subjective.
Blinding is probably not typical for
a TK study.
0
N/A
N/A
21. Confounding
variables in test setup
and procedures
HIGH
BWs were within +/- 20%, oral
gavage so no palatability issues. No
reported differences in respiratory
rate.
1
2
2
Confounding/
Variable Control
22. Health outcomes
unrelated to exposure
MEDIUM
Health outcomes weren't reported.
However, we don't expect health
effects from a single dose of this
amount (based on prior acute
studies) so differences in health of
the animals are not expected to
impact the results of the TK study.
2
1
2
Data Presentation
and Analysis
23. Statistical methods
HIGH
Results of statistical analysis and
method of calculation were
provided. The raw data are also
provided for re- analysis or further
analysis if needed.
1
1
1
24. Reporting of data
HIGH
All raw data were provided.
1
2
2
B-64
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Sum of scores:
26
27
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
1.0
Overall Quality Level:
HIGH
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-65
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
DuPont-18405-849 RV1. E.I. du Pont de Nemours and Company (2011). H-28548: Toxicokinetic Study in
Pregnant Rats. Test Guideline Not Identified. Study conducted by E.I. du Pont de Nemours and Company
(Original Report Completed: March 29, 2011; Report Revision 1 Completed: April 11, 2011), Newark,
Delaware.
Number of Hours
for Review:
0.75
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
HIGH
Identity of the substance was
identified definitively. See page 8
of the report.
1
2
2
2. Test substance
source
HIGH
Test substance was supplied by
sponsor. Batch number is H-28548.
1
1
1
3. Test substance purity
MEDIUM
Purity of the substance was 84%.
There is a certificate of analysis
that reports other components
(water, 12.7%) and
perfluorooctanoic acid (150 ppm).
2
1
2
Test Setup
4. Negative controls
MEDIUM
Used deionized water as the
negative control. Also they tested
the water samples for presence of
contaminants (e.g., total bacterial
counts, coliforms, lead).
2
2
4
5. Negative control
responses
HIGH
1
1
1
6. Positive controls
Not rated
The study did not include a
positive control. It was not
necessary.
0
N/A
N/A
7. Randomized
allocation
HIGH
The study report indicated that the
investigators used a computerized
randomization procedure to
produce homogeneous
distribution of BWs across groups
within each breeding lot.
1
1
1
B-66
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Exposure
Characterization
8. Preparation and
storage of test
substance
HIGH
Study characterized test
preparation and stability after
preparation; said this
demonstrated that the test
substance was stable at room
temperature for up to 12 days.
1
1
1
9. Consistency of
exposure
administration
HIGH
Exposure consistently
administered. Administered same
volume based on BW.
1
1
1
10. Reporting of doses/
concentrations
HIGH
Reported doses (control plus 5
doses).
1
2
2
11. Exposure frequency
and duration
HIGH
Rats exposed daily by oral gavage
on GD6 to 20.
1
1
1
12. Number of
exposure groups and
dose spacing
HIGH
Investigators included justification
for selection of dose levels.
1
1
1
13. Exposure route and
method
HIGH
Investigators included justification
for selection of route of
administration.
1
1
1
Test Organisms
14. Test animal
characteristics
HIGH
Provided information about test
model (rat strain, sex, gestational
day at arrival, age at arrival, age at
start of study, weight at arrival).
Study also reported justification
for animal model.
1
2
2
15. Consistency of
animal conditions
HIGH
Animal husbandry conditions were
reported (section F of report).
1
1
1
16. Number per group
HIGH
5 animals per group; above OECD
TG 417 recommended minimum of
4 animals per group.
1
1
1
Outcome
Assessment
17. Outcome
assessment
methodology
HIGH
Study report describes the
methodology used to conduct the
in- life and post-mortem
observations.
1
2
2
18. Consistency of
outcome assessment
HIGH
In life and post-mortem
observations were carried out
consistently.
1
1
1
19. Sampling adequacy
HIGH
Sampling frequency was reported
for mortality/moribundity, clinical
observations, BWs, food
consumption, blood collection.
1
1
1
20. Blinding of
assessors
MEDIUM
Clinical observations fall within the
definition of subjective outcomes.
However, the study did not discuss
blinding prior to recording clinical
observations.
2
1
2
B-67
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confounding/
Variable Control
21. Confounding
variables in test setup
and procedures
HIGH
Based on study summary and data
tables.
1
2
2
22. Health outcomes
unrelated to exposure
HIGH
Based on study summary and data
tables.
1
1
1
Data Presentation
and Analysis
23. Statistical methods
MEDIUM
Study reported using descriptive
statistics for the endpoints, but the
description was very brief.
2
1
2
24. Reporting of data
HIGH
Very well documented study. Data
reported in appendices.
1
2
2
Sum of scores:
30
35
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
1.2
Overall Quality Level:
HIGH
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-68
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
Gannon, SA; Fasano, WJ; Mawn, MP; Nabb, DL; Buck, RC; Buxton, LW; Jepson, GW; Frame, SR. (2016).
Absorption, distribution, metabolism, excretion, and kinetics of 2,3,3,3-tetrafluoro-2-
(heptafluoropropoxy)propanoic acid ammonium salt following a single dose in rat, mouse, and cynomolgus
monkev. Toxicology 340: 1-9. http://dx.doi.org/10.1016/i.tox.2015.12.006
Note:
This study was reviewed in three parts. The table below pertains to the rat and monkey i.v. dosing section of
the study.
Number of Hours
for Review:
1
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
HIGH
Definitive; Chemical Name and
CASRN.
1
2
2
2. Test substance
source
HIGH
Reported; from DuPont Chemicals
and Fluoro products.
1
1
1
3. Test substance
purity
HIGH
Purity of the substance was 99.4%.
1
1
1
Test Setup
4. Negative controls
Not rated
PK studies - measuring ADME of
SUBSTANCE.
0
N/A
N/A
5. Negative control
responses
Not rated
PK studies - measuring ADME of
SUBSTANCE.
0
N/A
N/A
6. Positive controls
Not rated
PK studies - measuring ADME of
SUBSTANCE.
0
N/A
N/A
7. Randomized
allocation
MEDIUM
Not indicated.
2
1
2
Exposure
Characterization
8. Preparation and
storage of test
substance
LOW
Not reported in journal article; EPA
confirmed it is reported in full Study
Summary.
3
1
3
9. Consistency of
exposure
administration
HIGH
Exposure consistently administered
across dose groups and species based
on BW.
1
1
1
10. Reporting of
doses/ concentrations
HIGH
Reported doses (1 dose).
1
2
2
11. Exposure
frequency and
duration
HIGH
Rats & monkeys exposed ONCE; PK
Study.
1
1
1
12. Number of
exposure groups and
dose spacing
HIGH
Reported; 6 rats: 3 of each sex per
dose; 6 monkeys: 3 of each sex per
dose group.
1
1
1
13. Exposure route
and method
MEDIUM
Reported; IV; via tail vein in rat and
via peripheral vein in monkey.
2
1
2
B-69
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Test Organisms
14. Test animal
characteristics
HIGH
Provided species, strain, sex.
1
2
2
15. Consistency of
animal conditions
HIGH
Consistent; cited DuPont Haskell
Global Center principles, and AAALAC
accreditation.
1
1
1
16. Number per group
HIGH
3 males + 3 females in rat study per
dose (1); 3 males + 3 females in
monkey study per dose (1).
1
1
1
Outcome
Assessment
17. Outcome
assessment
methodology
HIGH
Study describes in detail the
methodology (sampling
times/frequency) and analytical (high-
performance liquid chromatography-
mass spectrometry) methods.
1
2
2
18. Consistency of
outcome assessment
HIGH
Consistent across sexes within species
and across species in sampling
times/methods.
1
1
1
19. Sampling
adequacy
HIGH
Adequate; serum concentrations
leveled off at last several time points.
1
1
1
20. Blinding of
assessors
Not rated
Observations are analytically
determined; not subjective.
0
N/A
N/A
Confounding/
Variable Control
21. Confounding
variables in test setup
and procedures
MEDIUM
Limit of detection in IV test is lower
(better) than in oral test (see other
review); hence, care should be taken
in comparing rat oral vs either species
IV results.
2
2
4
22. Health outcomes
unrelated to exposure
HIGH
None.
1
1
1
Data Presentation
and Analysis
23. Statistical methods
MEDIUM
Described; data tables and graphical
representations show inputs and
results, respectively. Goodness of fit
results not provided in paper.
2
1
2
24. Reporting of data
MEDIUM
Typical of journal article; results of
calculations and graphs provided in
article, but underlying data not
available for recalculation or
modeling. Although paper indicates
supplementary data are available on-
line, it is simply 3 additional graphs;
NOT DATA.
2
2
4
Sum of scores:
26
35
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
1.4
Overall Quality Level:
HIGH
B-70
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
High Medium Low
>1 and <1.7 >1.7 and <2.3 >2.3 an
B-71
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
Gannon, SA; Fasano, WJ; Mawn, MP; Nabb, DL; Buck, RC; Buxton, LW; Jepson, GW; Frame, SR. (2016).
Absorption, distribution, metabolism, excretion, and kinetics of 2,3,3,3-tetrafluoro-2-
(heptafluoropropoxy)propanoic acid ammonium salt following a single dose in rat, mouse, and cynomolgus
monkev. Toxicoloav 340:1-9. http://dx.doi.org/10.1016/i.tox.2015.12.006
Note:
This study was reviewed in three parts. The table below pertains to the rat and mouse PK section of the study.
Number of Hours
for Review:
1
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
HIGH
Definitive; Chemical Name and CASRN.
1
2
2
2. Test substance
source
HIGH
Reported; from DuPont Chemicals and
Fluoro products.
1
1
1
3. Test substance
purity
HIGH
Purity of the substance was 99.4%.
1
1
1
Test Setup
4. Negative controls
Not rated
PK studies - measuring ADME of
SUBSTANCE.
0
N/A
N/A
5. Negative control
responses
Not rated
PK studies - measuring ADME of
SUBSTANCE.
0
N/A
N/A
6. Positive controls
Not rated
PK studies - measuring ADME of
SUBSTANCE.
0
N/A
N/A
7. Randomized
allocation
MEDIUM
Not indicated.
2
1
2
Exposure
Characterization
8. Preparation and
storage of test
substance
LOW
Not reported in journal article; EPA
confirmed it is reported in full Study
Summary.
3
1
3
9. Consistency of
exposure
administration
HIGH
Exposure consistently administered
across dose groups and species based on
BW.
1
1
1
10. Reporting of
doses/
concentrations
HIGH
Reported doses (2 doses).
1
2
2
11. Exposure
frequency and
duration
HIGH
Rats & mice exposed ONCE; PK Study.
1
1
1
12. Number of
exposure groups
and dose spacing
HIGH
Reported; 3 of each sex per dose; MICE:
45 of each sex per dose group.
1
1
1
13. Exposure route
and method
MEDIUM
Reported; oral; method not specified
(gavage is usual).
2
1
2
B-72
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Test Organisms
14. Test animal
characteristics
HIGH
Provided species, strain, sex.
1
2
2
15. Consistency of
animal conditions
HIGH
Consistent; cited DuPont Haskell Global
Center principles, and AAALAC
accreditation.
1
1
1
16. Number per
group
HIGH
3 males + 3 females in rat study per dose
(2); 45 males + 45 females in mouse
study per dose (2).
1
1
1
Outcome
Assessment
17. Outcome
assessment
methodology
HIGH
Study describes in detail the
methodology (sampling times /
frequency) and analytical (high-
performance liquid chromatography-
mass spectrometry) methods.
1
2
2
18. Consistency of
outcome
assessment
HIGH
Consistent across sexes within species
and across species in sampling
times/methods.
1
1
1
19. Sampling
adequacy
HIGH
Adequate; serum concentrations leveled
off at last 2 time points.
1
1
1
20. Blinding of
assessors
Not rated
Observations are analytically
determined; not subjective.
0
N/A
N/A
Confounding/
Variable Control
21. Confounding
variables in test
setup and
procedures
MEDIUM
Limit of detection in oral test different
than in IV test (see other review); hence,
care should be taken in comparing rat
oral vs IV; however, absorption and
alpha elimination phases appears to be
well above LOD; caution comparing beta
elimination values...differences could be
LOD related.
2
2
4
22. Health
outcomes unrelated
to exposure
HIGH
None.
1
1
1
Data Presentation
and Analysis
23. Statistical
methods
MEDIUM
Described; data tables and graphical
representations show inputs and results,
respectively. Goodness of fit results not
provided in paper.
2
1
2
24. Reporting of
data
MEDIUM
Typical of journal article; results of
calculations and graphs provided in
article, but underlying data not available
for recalculation or modeling. Although
paper indicates supplementary data are
available on- line, it is simply 3
additional graphs; NOT DATA. EPA able
to review full study summaries.
2
2
4
B-73
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Sum of scores:
26
35
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
1.3
Overall Quality Level:
HIGH
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-74
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
Gannon, SA; Fasano, WJ; Mawn, MP; Nabb, DL; Buck, RC; Buxton, LW; Jepson, GW; Frame, SR. (2016).
Absorption, distribution, metabolism, excretion, and kinetics of 2,3,3,3-tetrafluoro-2-
(heptafluoropropoxy)propanoic acid ammonium salt following a single dose in rat, mouse, and cynomolgus
monkev. Toxicoloav 340:1-9. http://dx.doi.org/10.1016/i.tox.2015.12.006
Note:
This study was reviewed in three parts. The table below pertains to the in vitro section of the study.
Number of Hours
for Review:
1
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
HIGH
Definitive; Chemical Name and
CASRN.
1
2
2
2. Test substance
source
HIGH
Reported; from DuPont Chemicals
and Fluoroproducts.
1
1
1
3. Test substance purity
HIGH
Purity of the substance was 99.4%.
1
1
1
Test Setup
4. Negative controls
HIGH
Heat-inactivated hepatocytes from
same prep as those used for
metabolism study.
1
2
2
5. Positive controls
HIGH
Included solvent only control (zero
dose).
1
2
2
6. Assay procedures
HIGH
Non-standard/guideline, but
straight forward test.
1
1
1
7. Standards for test
N/A
Non-standard/guideline test.
0
N/A
N/A
Exposure
Characterization
8. Preparation and
storage of test
substance
LOW
Not reported in journal article; EPA
confirmed it is reported in full Study
Summary.
3
1
3
9. Consistency of
exposure
administration
HIGH
Consistent; all dose with same stock
soln; different dilutions.
1
1
1
10. Reporting of
concentrations
HIGH
1 cone for clearance expt; 1 cone for
metabolite ID.
1
2
2
11. Exposure duration
HIGH
Reported; 6 sample points at
different durations.
1
2
2
12. Number of
exposure groups and
dose spacing
MEDIUM
1 exposure group is fine given this is
a screen for in vivo expts; no
rationale for concentrations used
was provided.
2
1
2
13. Metabolic
activation
HIGH
This was the point of the
experiment.
1
1
1
Test Model
14. Test model
HIGH
Primary isolated hepatocytes.
1
2
2
15. Number per group
HIGH
Multiple time points is the 'group1;
used 6.
1
1
1
B-75
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Outcome
Assessment
16. Outcome
assessment
methodology
HIGH
Study describes in detail the
methodology (sampling
times/frequency) and analytical
(high-performance liquid
chromatography-mass
spectrometry) methods.
1
2
2
17. Consistency of
outcome assessment
HIGH
There was no metabolism or
metabolites in any treatment; Very
consistent.
1
1
1
18. Sampling adequacy
HIGH
6 timepoints.
1
2
2
19. Blinding of
assessors
N/A
Not applicable.
0
N/A
N/A
Confounding/
Variable Control
20. Confounding
variables in test setup
and procedures
none
None.
0
N/A
N/A
21. Outcomes
unrelated to exposure
HIGH
No aberrant measures at any
timepoint reported.
1
1
1
Data Presentation
and Analysis
22. Data analysis
N/A
Does not appear to have needed
any 'analysis'; the results were
straight negative.
0
N/A
N/A
23. Data interpretation
HIGH
1
2
2
24. Cytotoxicity data
N/A
0
N/A
N/A
25. Reporting of data
LOW
NEGATIVE RESULTS: Neither
clearance rates nor any
chromatograms from metabolite
were shown/provided.
3
2
6
Sum of scores:
30
37
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
1.2
Overall Quality Level:
HIGH
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-76
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
Sheng, N; Cuir, R; Wang, J; Guo, Y; Wang, J; Dai, J. 2018. Cytotoxicity of novel fluorinated alternatives to long-
chain perfluoroalkyl substances to human liver cell line and their binding capacity to human liver fatty acid-
binding protein. Arch Toxicol 92:359-369. http://dx.doi.org/10.1007/s00204-017-2055-l
Number of Hours
for Review:
5 minutes
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
Unacceptable
Not from manufacturer and no
analysis; no reference to a
supplemental file. This designation
as "unacceptable" means the
remainder of the domains/metrics
will not be reviewed/scored.
4
N/A
N/A
2. Test substance
source
3. Test substance purity
4. Negative controls
Test Setup
5. Negative control
responses
6. Positive controls
7. Randomized
allocation
8. Preparation and
storage of test
substance
9. Consistency of
exposure
administration
Exposure
Characterization
10. Reporting of doses/
concentrations
11. Exposure frequency
and duration
12. Number of
exposure groups and
dose spacing
13. Exposure route and
method
B-77
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Test Organisms
14. Test animal
characteristics
15. Consistency of
animal conditions
16. Number per group
Outcome
Assessment
17. Outcome
assessment
methodology
18. Consistency of
outcome assessment
19. Sampling adequacy
20. Blinding of
assessors
Confounding/
Variable Control
21. Confounding
variables in test setup
and procedures
22. Health outcomes
unrelated to exposure
Data Presentation
and Analysis
23. Statistical methods
24. Reporting of data
Sum of scores:
N/A N/A
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
N/A
Overall Quality Level:
UNACCEPTABLE
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-78
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
Wang, J; Wang, X; Sheng, N; Zhou, X, Cui, R; Zhang, H; Dai, J. 2016. RNA-sequencing analysis reveals the
hepatotoxic mechanism of perfluoroalkyl alternatives, HFP02 and HFP04, following exposure in mice. Journal
of Applied Toxicoloav 37:436-444. https://doi. org/10.1002/iat.3376
Number of Hours
for Review:
0.25
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
HIGH
CAS RN
1
2
2
2. Test substance
source
UNACCEPTABLE
Test materials were synthesized in
the author's lab. However, no
information on the synthetic
process, certificate of analysis,
etc., were included to verify the
identity of the test materials. The
designation of this metric as
"unacceptable" means the
remaining domains/metrics are
not reviewed/scored.
4
N/A
N/A
3. Test substance purity
Test Setup
4. Negative controls
5. Positive controls
6. Assay procedures
7. Standards for test
Exposure
Characterization
8. Preparation and
storage of test
substance
9. Consistency of
exposure
administration
10. Reporting of
concentrations
11. Exposure duration
12. Number of
exposure groups and
dose spacing
13. Metabolic
activation
Test Model
14. Test model
15. Number per group
B-79
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Outcome
Assessment
16. Outcome
assessment
methodology
17. Consistency of
outcome assessment
18. Sampling adequacy
19. Blinding of
assessors
Confounding/
Variable Control
20. Confounding
variables in test setup
and procedures
21. Outcomes
unrelated to exposure
Data Presentation
and Analysis
22. Data analysis
23. Data interpretation
24. Cytotoxicity data
25. Reporting of data
Sum of scores:
N/A N/A
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
4
Overall Quality Level:
UNACCEPTABLE
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-80
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
Rushing, B; Hu, Q; Franklin, J; McMahen, R; Dagnio, Sonia; Higgins, Christopher; Strynar, M; DeWitt, J. 2017.
Evaluation of the Immunomodulatory Effects of 2,3,3,3-Tetrafluoro-2-(Heptafluoropropoxy)- Propanoate in
C57BL/6 Mice. Toxicoloaical Sciences 156:179-189. https://doi.org/10.1093/toxsci/kfw251
Number of Hours
for Review:
1.5
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighte
d Score
Test Substance
1. Test substance
identity
MEDIUM
CAS RN provided. Batch/lot
number not provided; certificate
of analysis by manufacturer not
indicated.
2
2
4
2. Test substance
source
HIGH
Manufacturer reported.
1
1
1
3. Test substance purity
LOW
Purity and/or grade not reported.
3
1
3
Test Setup
4. Negative controls
HIGH
Addressed in Materials and
Methods section.
1
2
2
5. Negative control
responses
HIGH
Addressed in Results section.
1
1
1
6. Positive controls
Not rated
Not relevant to this study type.
0
N/A
N/A
7. Randomized
allocation
HIGH
Addressed in Materials and
Methods section.
1
1
1
Exposure
Characterization
8. Preparation and
storage of test
substance
MEDIUM
Preparation specified but not
storage.
2
1
2
9. Consistency of
exposure
administration
HIGH
1
1
1
10. Reporting of doses/
concentrations
HIGH
1
2
2
11. Exposure frequency
and duration
HIGH
1
1
1
12. Number of
exposure groups and
dose spacing
HIGH
Only limited information on dose
selection provided in report.
1
1
1
13. Exposure route and
method
HIGH
1
1
1
Test Organisms
14. Test animal
characteristics
HIGH
1
2
2
15. Consistency of
animal conditions
HIGH
1
1
1
16. Number per group
HIGH
1
1
1
B-81
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Outcome
Assessment
17. Outcome
assessment
methodology
HIGH
1
2
2
18. Consistency of
outcome assessment
HIGH
1
1
1
19. Sampling adequacy
HIGH
1
1
1
20. Blinding of
assessors
Not rated
0
N/A
N/A
Confounding/
Variable Control
21. Confounding
variables in test setup
and procedures
HIGH
1
2
2
22. Health outcomes
unrelated to exposure
LOW
Not reported.
3
1
3
Data Presentation
and Analysis
23. Statistical methods
HIGH
1
1
1
24. Reporting or data
HIGH
1
2
2
Sum of scores:
29
36
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
1.2
Overall Quality Level:
HIGH
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-82
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
DuPont-24447 (4 Volumes contained in 4 separate PDF files): E.I. du Pont de Nemours and Company (2008). A
28-Day Oral (Gavage) Toxicity Study of H-28397 in Rats with a 28-Day Recovery. OECD TG 407. Study conducted
by WIL Research Laboratories, LLC (Study Completion Date: August 22, 2008), Ashland, Ohio.
Number of Hours
for Review:
1.5
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighte
d Score
Test Substance
1. Test substance
HIGH
Definitive-Pg 15; Analyses in
Appendix C (in Volume 4 of 4).
1
2
2
2. Test substance
source
HIGH
Test substance, H-28397, was from
E.I. duPont Nemours and
Company, Newark, NJ (Pg 21).
1
1
1
3. Test substance purity
MEDIUM
Purity of the substance was 88%.
There is a certificate of analysis
that reports other components
(water, 13.3%) and
perfluorooctanoic acid (3.4 ppm).
2
1
2
Test Setup
4. Negative control
HIGH
Used deionized water as the
negative control. Tested the water
samples for presence of biological
contaminants (e.g., total bacterial
counts, coliforms, lead).
1
2
2
5. Negative control
responses
HIGH
1
1
1
6. Positive control
Not rated
The study did not include a
positive control. It was not
necessary.
0
N/A
N/A
7. Randomized
allocation
HIGH
The study report indicated that the
investigators used a computerized
randomization procedure to
produce homogeneous
distribution of BWs across groups
within each breeding lot.
1
1
1
B-83
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Exposure
Characterization
8. Preparation and
storage of test
substance
HIGH
Study characterized test
preparation and measured
homogeneity and stability after
preparation. Test substance was
stable at room temperature for 5
hours and refrigerated for up to 10
days.
1
1
1
9. Consistency of
exposure
administration
HIGH
Exposure consistently
administered; adjusted based on
BW. Due to lower than nominal
dosing solutions, volume
administered was increased for all
groups from study day 0-24
(males) or study day 0-23
(females). The increased volume
(from 10 uLto 12 nL) did NOT
exceed the 2 mL/100 g limit
designated in TG 407.
1
1
1
10. Reporting of doses/
concentrations
HIGH
Reported doses (control plus 3
doses).
1
2
2
11. Exposure frequency
and duration
HIGH
Rats exposed daily by oral gavage.
1
1
1
12. Number of
exposure groups and
dose spacing
HIGH
Reported; as perTG407; 10X
between doses OK as per TG407.
1
1
1
13. Exposure route and
method
HIGH
Reported; as per TG407.
1
1
1
Test Organisms
14. Test animal
characteristics
HIGH
Provided information about test
model (rat strain, sex, gestational
day at arrival, age at arrival, age at
start of study, weight at arrival).
Study also reported justification
for animal model; as per TG407.
1
2
2
15. Consistency of
animal conditions
HIGH
Consistent; as per TG407 and
described in Sections D, E, F and G
of Final Report (Volume 1 or 4)
and associated appendices.
1
1
1
16. Number per group
HIGH
20 males + 20 females per vehicle
control and high dose groups
(exceeds TG407) and 10 males +5
animals per group.
1
1
1
B-84
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Outcome
Assessment
17. Outcome
assessment
methodology
HIGH
Study report describes the in detail
the methodology used to conduct
the in-life and post-mortem
observations. Outcome
measurements meet or exceed all
requirements of TG 407.
1
2
2
18. Consistency of
outcome assessment
HIGH
Consistent; as per TG 407.
1
1
1
19. Sampling adequacy
HIGH
Adequate; as per TG 407.
1
1
1
20. Blinding of
assessors
MEDIUM
Clinical observations fall within the
definition of subjective outcomes.
The study did not discuss blinding;
however, internationally accepted
TGs do not require (preferable)
this,
2
1
2
Confounding/
Variable Control
21. Confounding
variables in test setup
and procedures
HIGH
No confounding variables; study
followed TG 407 without
exception.
1
2
2
22. Health outcomes
unrelated to exposure
HIGH
None
1
1
1
Data Presentation
and Analysis
23. Statistical methods
HIGH
Described in Section G of Final
Report (Volume 1 of 4). Statistical
tests were selected during the
design of the study (as perTG 407)
in "Protocol" found in Appendix 1
(Volume 4 of 4).
2
1
2
24. Reporting of data
HIGH
Very well documented study.
Summary provide in Final Report
(Volume 1 of 4). Individual data
provided as per TG 407 - found in
Appendices.
1
2
2
Sum of scores:
30
33
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
1.1
Overall Quality Level:
HIGH
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-85
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
DuPont-24459. E.I. du Pont de Nemours and Company (2008). A 28-Day Oral (Gavage) Toxicity Study of H-
28397 in Mice with a 28-Day Recovery. OECD Guideline 407. Study conducted by WIL Research Laboratories,
LLC (Study Completion Date: August 29, 2008), Ashland, Ohio.
Number of Hours
for Review:
1.25
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
HIGH
Identity of the substance was
identified definitively. See page
14 of the report.
1
2
2
2. Test substance
source
HIGH
Test substance was supplied by
sponsor. Batch number is H-
28397.
1
1
1
3. Test substance
purity
MEDIUM
Purity of the substance was 88%.
Test sample had other
components (water, 13.3%) and
perfluorooctanoic acid (3.4 ppm).
Certificate of analysis was
provided.
2
1
2
Test Setup
4. Negative controls
MEDIUM
Used deionized water as the
negative control. The certificate
of analysis indicates that the
formulation contains 3.4 ppm of
perfluorooctanoic acid as "other
components".
2
2
4
5. Negative control
responses
HIGH
Within normal ranges.
1
1
1
6. Positive controls
Not rated
The study did not include a
positive control. It was not
necessary.
0
N/A
N/A
7. Randomized
allocation
HIGH
The study report indicated that
the investigators used a
computerized randomization
procedure. (See Section H.)
1
1
1
B-86
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
8. Preparation and
storage of test
substance
HIGH
Study characterized test
preparation. Appendix C has
details about the analyses of
dosing formulations.
1
1
1
9. Consistency of
exposure
administration
HIGH
Exposure consistently
administered. Administered same
volume based on BW.
1
1
1
Exposure
Characterization
10. Reporting of doses/
concentrations
HIGH
Reported doses (control plus 3
doses) as recommended by OECD
TG 407.
1
2
2
11. Exposure
frequency and
duration
HIGH
Daily by oral gavage for 28 days.
1
1
1
12. Number of
exposure groups and
dose spacing
HIGH
Investigators included
justification for selection of dose
levels.
1
1
1
13. Exposure route and
method
HIGH
Investigators included
justification for selection of route
of administration.
1
1
1
14. Test animal
characteristics
HIGH
Information provided about test
model (Crl:CD-l mice). Study also
reported justification for animal
model.
1
2
2
Test Organisms
15. Consistency of
animal conditions
HIGH
Animal husbandry conditions
were reported (see Sections D, E,
F,G).
1
1
1
16. Number per group
HIGH
Consistent with OECD Guidelines,
there were a minimum often
animals per sex per group. Some
groups had 20 animals per sex
per group.
1
1
1
17. Outcome
assessment
methodology
HIGH
Study report describes the
methodology used to conduct the
endpoint assessments.
1
2
2
18. Consistency of
outcome assessment
HIGH
Endpoint assessments were
carried out consistently.
1
1
1
Outcome
Assessment
19. Sampling adequacy
HIGH
Sampling frequency was reported
for the various endpoints
measured.
1
1
1
20. Blinding of
assessors
MEDIUM
Included clinical observations
which fall within the definition of
subjective outcomes. However,
blinding prior to recording clinical
observations was not addressed.
2
1
2
B-87
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confounding/
Variable Control
21. Confounding
variables in test setup
and procedures
HIGH
Minor differences that would not
result in substantial impact to
results (e.g., food consumption-
see page 16 of Part 1).
1
2
2
22. Health outcomes
unrelated to exposure
HIGH
Two unexplained deaths
reported. Gross necropsy and
histologic findings did not provide
explanation for the mortality.
1
1
1
Data Presentation
and Analysis
23. Statistical methods
HIGH
They were reported and
explained (See Section H).
1
1
1
24. Reporting of data
HIGH
Very well documented study. All
individual animal data reported.
1
2
2
Sum of scores:
30
34
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
1.1
Overall Quality Level:
HIGH
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-88
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
DuPont-17751-1026: E.I. du Pont de Nemours and Company (2009). A 90-Day Oral (Gavage) Toxicity Study of
H-28548 in Rats with a 28-Day Recovery. OECD TG 408. Study conducted by WIL Research Laboratories, LLC
(Study Completion Date: October 5, 2009), Ashland, Ohio.
Number of Hours
for Review:
1
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
HIGH
Certificate of Analysis on page
1676/2076.
1
2
2
2. Test substance
source
HIGH
Provided by manufacturer; Page
24/2076.
1
1
1
3. Test substance
purity
MEDIUM
Low level of PFOA as a
contaminant (150 ppm;
equivalent to 0.015%).
2
1
2
Test Setup
4. Negative controls
HIGH
Page 24/2076
1
2
2
5. Negative control
responses
HIGH
All individual animal data
provided for all dose groups,
including controls.
1
1
1
6. Positive controls
Not rated
Not applicable for this study
design.
0
N/A
N/A
7. Randomized
allocation
HIGH
Page 27/2076.
1
1
1
Exposure
Characterization
8. Preparation and
storage of test
substance
HIGH
Adequately reported; Page 24 /
2076.
1
1
1
9. Consistency of
exposure
administration
HIGH
Adequately described; Page 25 /
2076.
1
1
1
10. Reporting of
doses/ concentrations
HIGH
Page 25/2076.
1
2
2
11. Exposure
frequency and
duration
HIGH
All exposure metrics followed
OECD TG 408; Page 25 / 2076.
1
1
1
12. Number of
exposure groups and
dose spacing
HIGH
Based on results from prior,
shorter- term studies; page 25 /
2076.
2
1
2
13. Exposure route
and method
HIGH
Page 25/2076.
1
1
1
B-89
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Test Organisms
14. Test animal
characteristics
HIGH
Page 56/2076.
1
2
2
15. Consistency of
animal conditions
HIGH
Adequate; Page 27 / 2076.
1
1
1
16. Number per group
HIGH
Page 27/2076.
1
1
1
Outcome
Assessment
17. Outcome
assessment
methodology
HIGH
Page 28/2076.
1
2
2
18. Consistency of
outcome assessment
HIGH
Page 30/2076.
1
1
1
19. Sampling
adequacy
HIGH
Consistent with OECD TG 408;
Page 28/2076.
1
1
1
20. Blinding of
assessors
HIGH
Page 28/2076.
1
1
1
Confounding/
Variable Control
21. Confounding
variables in test setup
and procedures
HIGH
None identified; Page 26 / 2076.
1
2
2
22. Health outcomes
unrelated to exposure
HIGH
None identified; Page 26 / 2076.
1
1
1
Data Presentation
and Analysis
23. Statistical methods
HIGH
Appropriate for study design;
Page 34/2076.
1
1
1
24. Reporting of data
HIGH
All individual animal data
provided for all dose groups,
including controls.
1
2
2
Notes'. PFOA = perfluorooctaiioic acid.
Sum of scores:
30
32
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
1.1
Overall Quality Level:
HIGH
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-90
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
DuPont-18405-1307: E.I. du Pont de Nemours and Companv (2010). H-28548: Subchronic Toxicity 90-Dav
Gavage Study in Mice. OECD Guideline 408. Study conducted by E.I. du Pont de Nemours and Company (Study
Completion Date: February 19, 2010), Newark, Delaware.
Number of Hours
for Review:
1.5
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
HIGH
Certificate of Analysis provided
on page 98 / 339.
1
2
2
2. Test substance
source
HIGH
Provided by manufacturer; Page
14/339.
1
1
1
3. Test substance
purity
MEDIUM
PFOA contamination noted; Page
9/339 (84%).
2
1
2
Test Setup
4. Negative controls
HIGH
Page 10/339.
1
2
2
5. Negative control
responses
HIGH
Individual animal pathology data
provided for all dose groups,
including controls.
1
1
1
6. Positive controls
Not rated
Not relevant to this study type.
0
N/A
N/A
7. Randomized
allocation
HIGH
Page 16/339.
1
1
1
Exposure
Characterization
8. Preparation and
storage of test
substance
HIGH
Page 16/339.
1
1
1
9. Consistency of
exposure
administration
HIGH
Page 16/339.
1
1
1
10. Reporting of
doses/ concentrations
HIGH
Page 16/339.
1
2
2
11. Exposure
frequency and
duration
HIGH
Page 16/339.
1
1
1
12. Number of
exposure groups and
dose spacing
HIGH
Page 12/339.
2
1
2
13. Exposure route
and method
HIGH
Page 16/339.
1
1
1
Test Organisms
14. Test animal
characteristics
HIGH
Page 14/339.
1
2
2
15. Consistency of
animal conditions
HIGH
Page 14/339.
1
1
1
16. Number per group
HIGH
Page 13/339.
1
1
1
B-91
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
17. Outcome
assessment
methodology
HIGH
Adequately described; Page 17 /
339.
1
2
2
Outcome
Assessment
18. Consistency of
outcome assessment
HIGH
Page 17/339.
1
1
1
19. Sampling
adequacy
HIGH
Page 20/339.
1
1
1
20. Blinding of
assessors
HIGH
Appropriate for study type; Page
18/339.
1
1
1
Confounding/
Variable Control
21. Confounding
variables in test setup
and procedures
HIGH
None; Page 25 / 339.
1
2
2
22. Health outcomes
unrelated to exposure
HIGH
None; Page 24 / 339.
1
1
1
Data Presentation
23. Statistical methods
HIGH
Appropriate to study type; Page
22/339.
1
1
1
and Analysis
24. Reporting of data
HIGH
All data presented for all dose
groups, including controls.
1
2
2
Notes'. PFOA = perfluorooctanoic acid.
Sum of scores:
30
32
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
1.1
Overall Quality Level:
HIGH
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-92
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult Appendix B, Section B.3 for the metrics to assist in filling out this form.
Study Reference:
DuPont-18405-841: E.I. du Pont de Nemours and Company (2010). An Oral (Gavage) Prenatal Developmental
Toxicity Study of H-28548 in Rats. US EPA OPPTS 850.3700; OECD Guideline 414. Study conducted by WIL
Research Laboratories, LLC (Study Completion Date: July 2, 2010), Ashland, Ohio.
Number of Hours
for Review:
1.5
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
HIGH
Certificate of analysis provided on
page 215/388.
1
2
2
2. Test substance
source
HIGH
Provided by manufacturer/study
sponsor (page 17 / 388).
1
1
1
3. Test substance
purity
MEDIUM
Low level of PFOA as a
contaminant (150 ppm;
equivalent to 0.015%).
2
1
2
Test Setup
4. Negative controls
HIGH
Page 19 / 388.
1
2
2
5. Negative control
responses
HIGH
All individual animal data
provided for all dose groups,
including controls.
1
1
1
6. Positive controls
Not rated
Not applicable to this study
design.
0
N/A
N/A
7. Randomized
allocation
HIGH
Page 20/388.
1
1
1
Exposure
Characterization
8. Preparation and
storage of test
substance
HIGH
Page 17/388.
1
1
1
9. Consistency of
exposure
administration
HIGH
Adequate (page 18 / 388).
1
1
1
10. Reporting of
doses/ concentrations
HIGH
Page 18/388.
1
2
2
11. Exposure
frequency and
duration
HIGH
Page 19 / 388.
1
1
1
12. Number of
exposure groups and
dose spacing
MEDIUM
Only limited information on dose
selection provided in report.
2
1
2
13. Exposure route
and method
HIGH
Adequately reported (page 19 /
388).
1
1
1
B-93
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Test Organisms
14. Test animal
characteristics
HIGH
Page 19 / 388.
1
2
2
15. Consistency of
animal conditions
HIGH
Page 20/388.
1
1
1
16. Number per group
HIGH
Page 19 / 388.
1
1
1
Outcome
Assessment
17. Outcome
assessment
methodology
HIGH
Page 22/388.
1
2
2
18. Consistency of
outcome assessment
HIGH
Page 22/388.
1
1
1
19. Sampling adequacy
HIGH
Page 22/388.
1
1
1
20. Blinding of
assessors
HIGH
Adequately reported (page 22 /
388).
1
1
1
Confounding/
Variable Control
21. Confounding
variables in test setup
and procedures
HIGH
None reported.
1
2
2
22. Health outcomes
unrelated to exposure
HIGH
None reported.
1
1
1
Data Presentation
and Analysis
23. Statistical Methods
HIGH
Appropriate to study
methodology (page 25 / 388).
1
1
1
24. Reporting of data
HIGH
All individual animal data
provided for all dose groups,
including controls.
1
2
2
Notes'. PFOA = perfluorooctanoic acid.
Sum of scores:
30
32
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
1.1
Overall Quality Level:
HIGH
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-94
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult the word document for the metrics to assist in filling out this form
Study Reference:
DuPont-18405-1037: E.I. du Pont de Nemours and Company (2010). An Oral (Gavage)
Reproduction/Developmental Toxicity Screening Study of H-28548 in Mice. US EPA OPPTS 870.3550; OECD
Guideline 421. Study conducted by WIL Research Laboratories, LLC (Study Completion Date: December 29,
2010), Ashland, Ohio.
Number of Hours
for Review:
2
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
HIGH
ID'ed by manufacturer.
1
2
2
2. Test substance source
HIGH
ID'ed by manufacturer.
1
1
1
3. Test substance purity
MEDIUM
84%; 12.7% water; 150 ppm
PFOA (= 0.015%); PFOA content
led to decreased confidence
but the amount of PFOA is not
expected to impact results.
2
1
2
Test Setup
4. Negative controls
HIGH
1
2
2
5. Negative control
responses
HIGH
1
1
1
6. Positive controls
Not rated
0
N/A
N/A
7. Randomized
allocation
HIGH
The report and protocol
described the randomization
process; BWs at day 0 were
within 20% of mean, as
presented in results summary
tables.
1
1
1
Exposure
Characterization
8. Preparation and
storage of test
substance
HIGH
Test substance was identified
as being stable.
1
1
1
9. Consistency of
exposure administration
MEDIUM
pH increase/dose - highest
dose especially; range probably
still okay.
2
1
2
10. Reporting of doses/
concentrations
HIGH
1
2
2
11. Exposure frequency
and duration
HIGH
1
1
1
12. Number of exposure
groups and dose spacing
HIGH
1
1
1
13. Exposure route and
method
HIGH
1
1
1
B-95
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Test Organisms
14. Test animal
characteristics
HIGH
1
2
2
15. Consistency of
animal conditions
HIGH
1
1
1
16. Number per group
HIGH
1
1
1
Outcome
Assessment
17. Outcome
assessment
methodology
HIGH
Went beyond OECD TG 421.
1
2
2
18. Consistency of
outcome assessment
MEDIUM
Consistency re: time of day for
measurements among groups
not always stated.
2
1
2
19. Sampling adequacy
HIGH
1
1
1
20. Blinding of assessors
MEDIUM
The report and protocol did not
mention blinding of assessors.
Substantial impact from this
lack of information is not
expected.
2
1
2
Confounding/
Variable Control
21. Confounding
variables in test setup
and procedures
HIGH
No confounders identified.
1
2
2
22. Health outcomes
unrelated to exposure
MEDIUM
10 mice died early;
undetermined causes; more
died at the lowest dose (in
females this was > 10%) but
not expected to have
disproportional effect on study.
2
1
2
Data Presentation
and Analysis
23. Statistical methods
HIGH
1
1
1
24. Reporting of data
HIGH
1
2
2
Notes: PFOA = perfluorooctanoic acid.
Sum of scores:
30
35
High
Medium
Low
Overall Score = Sum of
Weighted Scores/Sum of
Metric Weighting Factors:
1.2
and <1.7
>1.7 and <2.3
>2.3 and <3
Overall Quality Level:
HIGH
B-96
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Instructions - Please consult the word document for the metrics to assist in filling out this form
Study Reference:
DuPont-18405-1238 (13 Volumes contained in 10 separate PDF files): E.I. du Pont de Nemours and Company
(2010). H-28548: Combined Chronic Toxicity/Oncogenicity Study 2-Year Oral Gavage Study in Rats. US EPA
OPPTS 870.4300; OECD Guideline 453. Study conducted by MPI Research, Inc. (Study Completion Date: March
28, 2013), Mattawan, Michigan.
Number of Hours
for Review:
2
Domain
Metric
Qualitative
Determination
(i.e., High,
Medium, Low,
Unacceptable,
or Not rated)
Comments
Metric
Score
Metric
Weighting
Factor
Weighted
Score
Test Substance
1. Test substance
identity
HIGH
ID'ed by manufacturer.
1
2
2
2. Test substance
source
HIGH
ID'ed by manufacturer.
1
1
1
3. Test substance
purity
MEDIUM
84%; 12.7% water; 150 ppm PFOA
(= 0.015%); PFOA content led to
decreased confidence but the
amount of PFOA is not expected
to impact results.
2
1
2
Test Setup
4. Negative controls
HIGH
Study report indicates negative
controls and treatment groups
had same test conditions; no
major anomalies reported.
1
2
2
5. Negative control
responses
HIGH
Results appear appropriate for
negative controls.
1
1
1
6. Positive controls
Not rated
Positive controls not necessary
according to OECD test guideline
453.
0
N/A
N/A
7. Randomized
allocation
HIGH
Method of randomization fully
described/acceptable.
1
1
1
B-97
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
8. Preparation and
storage of test
substance
HIGH
Well described; substance
prepared weekly and was
identified as being stable so
storage at room temp seemed
appropriate.
1
1
1
9. Consistency of
exposure
administration
HIGH
Gavage volume was appropriate
and controls received vehicle and
were treated in same manner as
treatment groups.
1
1
1
Exposure
Characterization
10. Reporting of
doses/
concentrations
HIGH
Detailed information regarding
measured test concentrations
was reported in appendix.
1
2
2
11. Exposure
frequency and
duration
HIGH
Daily doses given as specified by
OECD TG 453; 2-yr study duration
appropriate as per OECD TG 453.
1
1
1
12. Number of
exposure groups and
dose spacing
HIGH
Dose groups chosen based on
previous 90-day study; choices
are appropriate; range of effects
and results suggest the same.
1
1
1
13. Exposure route
and method
MEDIUM
OECD TG 453 suggests route for
environmental chemicals to be via
diet or drinking water; however
study authors justify their choice
by noting gavage is most efficient
way to get an accurate dose.
2
1
2
14. Test animal
characteristics
HIGH
BWs at start were within 20% of
mean BWs; source of animals
adequate; species also adequate
as per OECD TG 453.
1
2
2
Test Organisms
15. Consistency of
animal conditions
HIGH
Housing conditions for all animals
were appropriate, as specified
according to OECD TG 453.
1
1
1
16. Number per
group
HIGH
Numbers (80/sex/group) exceed
OECD TG 453 recommendation
(50/sex/group).
1
1
1
17. Outcome
assessment
methodology
HIGH
Standard outcomes were
reported, as specified by OECD TG
453.
1
2
2
Outcome
Assessment
18. Consistency of
outcome assessment
HIGH
Outcomes assessed consistently
at same times among test groups
and controls.
1
1
1
19. Sampling
adequacy
HIGH
Sampling well described; and is
appropriate for study.
1
1
1
20. Blinding of
assessors
MEDIUM
Blinding not mentioned in study;
don't expect it to influence results
significantly.
2
1
2
B-98
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Confounding/
Variable Control
21. Confounding
variables in test
setup and procedures
MEDIUM
Statistically significant differences
at week (-1) in food consumption.
2
2
4
22. Health outcomes
unrelated to
exposure
HIGH
Disease/infection evaluation well
described and negative results
reported.
1
1
1
Data Presentation
and Analysis
23. Statistical
methods
MEDIUM
Statistical methods and use of
statistics clearly stated/identified;
tumor incidence evaluated both
with survival adjusted and
unadjusted tests. Statistics were
also appropriately re-run for male
rats at request of EPA to account
for rats that died early and had no
tumors. However, no mention of
trend tests, which can have
greater power (according to OECD
guidance document 116).
2
1
2
24. Reporting of data
HIGH
The study reported both
individual animal data (in
appendices) and summary tables
with statistics (in results of the
main report).
1
2
2
Notes'. PFOA = perfluorooctanoic acid.
Sum of scores:
30
36
Overall Score = Sum of Weighted
Scores/Sum of Metric Weighting
Factors:
1.2
Overall Quality Level:
HIGH
High
Medium
Low
>1 and <1.7
>1.7 and <2.3
>2.3 and <3
B-99
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Appendix C: Acute and 7-Day Study Summaries
This appendix summarizes studies evaluating acute exposure to HFPO dimer acid or HFPO
dimer acid ammonium salt by the oral, dermal, and inhalation routes of exposure and
investigating dermal and eye irritation.
Oral Toxicity. In a study of the HFPO dimer acid ammonium salt (no technical guideline (TG)
cited), a single dose of 1.5, 12, 130, 1,000, 2,250, 3,400, 5,000, 7,500, 11,000, 12,963, or 17,000
mg/kg HFPO dimer acid ammonium salt was administered by stomach tube to young male rats.
The approximate lethal dose (ALD) was determined to be 7,500 mg/kg. Discomfort, gasping,
and tonic convulsions were observed before death at lethal doses (7,500 mg/kg and higher).
Discomfort, increased water intake, inactivity, polyuria, and initial weight loss were observed in
rats at the three highest sublethal doses (2,250 mg/kg, 3,400 mg kg. and 5,000 mg/kg). Slightly
enlarged livers with enlarged hepatocytes and pronounced cell membranes were also observed in
rats at the three highest sublethal doses. Slight-to-moderate degenerali\ e changes in the pancreas
were also observed in doses at 2,250 mg/kg and higher No effects were observed at doses of <
1,000 mg/kg (DuPont-2-63, 1963).
In another study evaluating toxicity of HFPO dimer acid ammonium salt by the oral route of
exposure (no TG identified), a single dose of 670, 2.3<)(). 3.400, 5,000, 7,500. or I I .<)00 mg/kg
HFPO dimer acid ammonium salt (purity > 99%) was administered to 7-week-old male rats
(1/dose group). Rats were evaluated for clinical signs of toxici ty o\ er a 14-day observation
period. No clinical signs of toxicity were obser\ ed in the rat dosed at 670 mg/kg. Rats dosed at
2,300 and 3,400 mg/kg exhibited weight loss (17% and 14%, respectively); ruffled fur; and a
wet, yellow-stained perineum at 1 day post-exposure. The rats dosed at 2,300 and 3,400 mg/kg
no longer exhibited these effects at 2 days and 4 days post-exposure, respectively. Rats dosed
with > 5,000 mg/kg died by I day after dosing The rat dosed with 1 1,000 mg/kg exhibited
lethargy, low carriage, and low posture before its death The ALD was determined to be 5,000
mg/kg (DuPont-770-95, 1996 )
A single dose of I II PO dimer acid ammonium salt (N2 (•>".. purity) was administered by oral
gavageto l
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Oral toxicity of HFPO dimer acid ammonium salt was also evaluated in male rats in a study
conducted according to OECD TG 425 (Up-and-Down Procedure). A single dose of HFPO
dimer acid ammonium salt (86% purity) was administered by oral gavage to 9- to 11-week-old
male rats at a dose of 175, 550, 1,750, or 5,000 mg/kg (3 rats). Rats were then evaluated for
clinical signs of toxicity over a 14-day observation period. All rats exhibited clinical signs of
toxicity such as lethargy, wet fur, stained fur/skin, decreased muscle tone, low posture, or lung
noise. One rat dosed at 1,750 mg/kg and all rats (3) dosed at 5,000 mg/kg died either the day
dosed or by the day after dosing. Grossly observable evidence of organ or tissue damage in rats
dosed at 5,000 mg/kg included expanded lungs and discolored stomach; discoloration and
cloudiness of eyes and stained skin. With the exception of mis dosed at 5,000 mg/kg, increases in
BW were observed in all rats over the course of the study The oral LD50 was determined to be
1,750 mg/kg for male rats (DuPont-25438 RV1, 2<)i)X)
Another study evaluated oral toxicity of] II PO dimer acid in both male and female rats in a
study conducted according to OECD TG 425 (I p-and-Down Procedure). A single dose of HFPO
dimer acid (98% purity) was administered to to 11-week-old rats Males were dosed at 175,
550, 1,750, or 5,000 mg/kg (2-6 rats/group). I'enuile rats were also dosed at 175, 550, 1,750, or
5,000 mg/kg (1-4 rats/group). Clinical signs were not obser\ ed in rats dosed at 175 mg/kg or in
one male rat dosed at 550 mg/kg The rest of the rats in this study exhibited clinical signs of
toxicity. Clinical signs of toxicity in male rats were obser\ ed up to 5 days after dosing, included
lung noise, absent feces, lethargy, not eating, stained fur skin, wet fur, labored breathing,
decreased muscle tone, prostrate posture, tremors, clear oral discharge, diarrhea, ataxia, and/or
high posture. Clinical signs in female rats were obser\ ed lor up to 3 days after dosing and
included wet fur, stained fur skin, ataxia, labored breathing, cold to touch, clear ocular or oral
discharge, lethargy, lung noise, absent feces, not eating, and/or rubbing face on the bottom of the
cage (DuPont-25875, 2008).
All rats dosed at 5,000 mg/kg died by the day after dosing. Among rats dosed at 1,750 mg/kg,
two males and three females died by the day after dosing. One male rat dosed at 550 mg/kg (rat
#274) was sacrificed in extremis on the fourth day after dosing following a 23% reduction in
BW. Gross lindings were detected in three male rats dosed at 5,000 mg/kg, in four rats dosed at
1,750 mg kg. and in one rat dosed at 550 mg/kg. Small testes and epididymis were observed in
rat #274 A discolored, glandular stomach was observed in two of the male rats dosed at
1,750 mg/kg Gross lindings for male rats dosed at 5,000 mg/kg included a glandular stomach; a
glandular, discolored stomach (rats #640, #796, and #821); and discolored skin (rat #796). Gross
findings for female rats dosed at 1,750 mg/kg included a glandular, discolored stomach (rats
#478, #527, #626); discolored lymph nodes (rat #527); and discolored skin (#527). The female
rat dosed at 5,000 mg/kg exhibited wet skin; a discolored esophagus with foamy fluid; and a
thick, discolored stomach. Increases in BW were observed in animals that survived until the end
of the study. The oral LD50 was estimated to be 1,730 mg/kg for male rats and 1,750 mg/kg for
female rats (DuPont-25875, 2008).
Another study conducted according to OECD TG 425 (Up-and-Down Procedure) evaluated
toxicity of HFPO by the oral route of exposure in female mice. A single dose of HFPO dimer
acid ammonium salt (86% purity) was administered to 8- to 9-week-old female mice at a dose of
175, 550, or 1,750 mg/kg (1-3 mice). No clinical signs of toxicity were observed in mice dosed
at 175 mg/kg or in two mice dosed at 550 mg/kg. One mouse dosed at 550 mg/kg, however,
C-2
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
exhibited wet fur on the day of dosing. All three mice dosed at 1,750 mg/kg died on the day of
dosing. Discoloration of lungs and an ovarian cyst were observed in a mouse dosed at 550
mg/kg. Skin stain was also observed in two mice dosed at 1,750 mg/kg. These observations were
considered by study authors to be nonspecific and not indicative of test substance-related. With
the exception of mice dosed at 1,750 mg/kg, increases in BW were observed in all mice over the
course of the study. The oral LD50 was estimated to be 1,030 mg/kg for female mice (DuPont-
24126, 2007).
Dermal Toxicity. In a study evaluating toxicity through dermal absorption (no TG identified),
5,000 mg/kg HFPO dimer acid ammonium salt (purity ..) was applied directly onto the
shaved, intact skin of two young adult male New Zen kind white rabbits for a period of 24 hours.
One rabbit exhibited necrosis from days 2-6 post-appl i ciiii on in a small area of treated skin. The
necrotic area sloughed off by day 7, and alopecia was then obser\ cd in this area until the study
was completed. Moderate erythema was obser\ cd in both rabbits at I day post-application and
was still observed up to 3 days post-application Isrythema persisted until 13 days post-
application, with the degree of severity decreasing overtime. Both rabbits exhibited scaling and
sloughing of skin 6-13 days after application Increases in BW were obser\ ed for both rabbits at
the conclusion (day 14) of the studv. The ALD was determined to be higher than 5,000 mg/kg
(DuPont-839-95, 1996).
The dermal toxicity of HFPO dimer acid ammonium salt (S(->% purity) was also evaluated in rats
in a study conducted according to OIX'I) IG 4<>2 (OPPIS N7t) 1200). A single dose of
5,000 mg/kg (fi\ e males and fi\ e females) was applied directly onto the shaved, intact skin for
24 hours. Rats ^ere then obser\ ed daily for 14 days post-treatment. All female rats exhibited
mild erythema on the test site I day post-application. Erythema was no longer detectable by the
second day after application. Erythema was not observed in male rats. Hyperkeratosis was
observed in four male and lour female rats I Iceration was observed in one male and two female
rats All dermal effects cleared up by I 3 days post-treatment. Increases in BW were observed for
male and female rats by the conclusion (day 14) of the study. The LD50 of the compound was
determined to be higher than 5.000 mg/kg (DuPont-24113, 2007).
Inhalation Toxicity. The toxicity of IIFPO dimer acid ammonium salt by the inhalation route of
exposure was e\ aluated in S-week-old male and female rats (no TG identified) (DuPont-17751-
723, 2009). One group of li\ e male and five female rats were exposed to an aerosol atmosphere
containing 5,20<) mu 111' of I ll-'PO (84% purity) to determine the inhalation median lethal
concentration (LC5..) Two other groups of three male and three female rats were exposed to
HFPO at concentrations of n. 13, and 100 mg/m3 in air to evaluate respiratory tract pathology.
All rats were exposed nose-only for a single 4-hour period. Rats exposed to 0, 13, and
100 mg/m3 HFPO in air were evaluated for clinical signs of toxicity for 2 days following
exposure and rats exposed to 5,200 mg/m3 of HFPO were evaluated for a period of 14 days
following exposure. Respiratory tract tissues (lung, larynx/pharynx, trachea, and nose) of the 0-,
13-, and 100-mg/m3 exposure groups were also evaluated microscopically. According to study
authors, no clinical signs of toxicity were observed for any animals at any exposure in this study.
However, following the 100 mg/m3 exposure, all rats displayed a red nasal discharge
immediately after exposure. Rat exposed to 5,200 mg/m3 exhibited red discharge from eyes,
nose, and mouth as well as red stains on skin/fur immediately after exposure. Red discharge and
staining were absent within 1 or 2 days after exposure. Rats in the 5,200-mg/m3 exposure group
C-3
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
lost 2.5% to 6.8% of their original BW for 1 or 2 days after exposure, but exhibited normal
weight gain for the remainder of the experiment. The LCso was determined to be greater than
5,200 mg/m3 (DuPont-17751-723, 2009).
Dermal Irritation. The dermal irritation of HFPO dimer acid ammonium salt (86% purity) was
evaluated in three male New Zealand white rabbits in a study conducted according to OECD TG
404 (OPPTS 870.2500). A 0.5-mL aliquot of the compound was applied to an area of shaved
skin for a period of 4 hours. Very slight erythema was observed in one rabbit following removal
of the compound. At 60 minutes post-application, \ cry slight erythema was observed in one
rabbit and well-defined erythema was observed in the oilier two rabbits. Erythema had cleared by
24 hours post-exposure (DuPont-24030, 2007).
Eye Irritation. In an OECD TG 405 (OPPTS 870 24<)()j study c\ aluating eye irritation of HFPO
dimer acid ammonium salt (86% purity), a < > I -m L aliquot of com pound was administered to one
eye of a young adult male New Zealand w hi ic rahbit. Necrosis, characterized by brown and
white discoloration of the conjunctival membrane of the treated eye. was observed at 1, 24, and
28 hours after application. Corneal opacity, iritis, conjunctival chemosis. and discharge were also
observed. Fluorescein stain examination of the treated eye indicated corneal injury (DuPont-
24114,2007).
Seven-day Toxicity Studies. Four 7-day studies arc a\ ai lahle for the HFPO dimer acid or
ammonium salt in rats or mice. The toxicity of I ll7PO dimer acid ammonium salt (86.6% purity)
by the oral route of exposure was evaluated in (¦•-week-old male and female rats (DuPont-24009,
2008). Five rats of each sc\ were exposed to <>. 3d. 3<)o. or 1,000 mg/kg HFPO by oral gavage
for 7 days. No clinical signs of toxicity were ohser\ cd in either sex at any dose level tested. A
significant decrease in BW was observed on test day 7 in males exposed to 1,000 mg/kg versus
control. Significant decreases in red blood cells (RIJCs), hemoglobin, and hematocrit were
obsei \ cd in male rats at 3<)i) mg kg day and in both male and female rats at 1,000 mg/kg/day. A
significant increase in red cell distribution width, reticulocytes, and neutrophils was also
ohscr\ cd in female rats exposed to 1,000 mg/l
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
urea nitrogen and decreased bilirubin, creatinine, total protein, globulin, and calcium were
observed at 30 and/or 300 mg/kg/day. Increased liver weight was observed in males at all doses
and in females at 300 mg/kg/day. Microscopic examination of livers detected hepatocellular
hypertrophy in all treated males and females. Lesions observed in males and females were mild
and minimal, respectively. A statistically significant increase in P-oxidation was detected in
females exposed to 300 mg/kg/day versus control.
A 7-day study was conducted in 6-week-old male mice to evaluate toxicity of HFPO dimer acid
ammonium salt (86.6% purity) by the oral route of exposure (DuPont-24010, 2008). Doses of
0 or 30 mg/kg/day were administered over a period of 7 days liy test day 7, BWs were
significantly higher in exposed males versus controls A twofold increase in liver weight relative
to control was detected in exposed males. No grossly ohser\ uMe lesions in the liver were
observed. Microscopic changes in the liver observed at 30 mu kg day included minimal single-
cell necrosis of hepatocytes, moderate hepatocellular hypertrophy, and moderate increases in
mitotic figures. Minimal vacuolation of hepatoex tes was also obser\ ed in one treated mouse.
Another 7-day gavage study was conducted in (-"-week-old male mice to e\ aluate toxicity of
HFPO dimer acid (99% purity) by the oral route of exposure (DuPont-252S 1. 2008). Doses of
0 or 30 mg/kg/day were administered o\ er a period of 7 days By test day 7. liWs were
significantly higher in exposed males \ ersus controls A twofold increase in liver weight was
detected in exposed males versus control Microscopic changes to the liver of exposed animals
included minimal single-cell necrosis of hepatocytes. moderate hepatocellular hypertrophy, and
moderate increases in mitotic figures Minimal \ acuolixalion was also observed in 2/5 treated
mice.
C-5
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
NOVEMBER 2018
Appendix D: Genotoxicity Study Summary
Table D-l provides a summary of the available genotoxicity data for III PO dimer acid and/or ammonium salt.
Table D-l. Genotoxicity Study Summary
Study
Assay
Strain/Species
Dosing
Activation
Results
DuPont-
19713
RV1
(2008)
In vitro
Bacterial
Reverse
Mutation Test
(OECD
Guideline 471)
Salmonella
typhimurium
(strains TA98,
TA100, TA1535,
and TA1537) and
Escherichia coli
(strain WP2//irA)
HFPO dimer acid ammonium salt (85% purity)
33.3,66.7, 100, 3 -.<.<•". 1. V-land 5,000
Hg/plate forprcliminai'> lo\icil> lesi
333, (>(>". 1 .<>()(). 3."" \ and 5,00(> uu plate for
toxicil> -muialion lesi
Negative coniiol (sierilc ualci) and pnsiu\e control
(hcii/o[a]pyreiic. 2-nii milium no. 2-aniiikiauihracene,
sodium azide, acridino niuiaucu l( k-ll>l. oi'4-
iiisii\H|Minoline-]N-iiMdc) also included in sludv
With S9
Negative.
Willioul S9
Negative.
DuPont-
22620
RV1
(2009)
In vitro
Mammalian
Chromosome
Aberration Tesi
(OECD
Guideline 4~' i
Chinese haniskT
ovary cells
(CIIO-K line)
1 ll 'IM) dimcr acid ammonium salt (83%purity)
4l>. 'JS. 244. 4Xl). K) . 1954. and 3391 ng/mL for
pivllnilIKIIA IO\ICII> k'sl*
') . 1 ''54. and ^ 1 uu/mL for the 4-hour
ikiuacli\alcd and aclivated test conditions*
4SlJ.') . and ll>54 ug/mLforthe 20-hour
iKniacli\alcd lesi condition*
Vualiv c control (sici ile water) and positive control
(niiiomycin C or cyclophosphamide) also included in
s|lld\
With S9
Positive at 3,391 (ig/mL* in 4-
hour activated test conditions.
Without S9
Negative.
* Doses have been corrected to account for 83% HFPO dimer acid ammonium salt purity.
D-l
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
NOVEMBER 2018
Study
Assay
Strain/Species
Dosing
Activation
Results
DuPont-
23219
(2007)
In vivo
Unscheduled
DNA Synthesis
Test in
Mammalian
Cells (OECD
Guideline 486)
Primary hepatocytes
harvested from male
rats (5/dose group)
HFPO dimer acid ammonium sail s>;i\
5(1(1. "50. 1.0(1(1. 1.5(1(1. and ug/niL for
uoiiacti\ ated cultures \x illi a 4-hour exposure
I5(i. 25(i. 500. (•(>(>. and ~5(> uu mL for S9-activated
cultures u itli a 4-Ikmii" e\posure
25(i. 5(i(i. (»(hi. ~5(i. and 1.(>(>(> ug/mLfor
iioiiacliN ated cultures with a 24-hour exposure
\euali\ e control (sterile, distilled water) and positive
control (iiicthv 1 niethanesulfonate or 7,12-
dinielhvlheii/(a)anthracene) also included in study
Willi S'J
Negative.
Without S9
Negative.
D-2
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
NOVEMBER 2018
Study
Assay
Strain/Species
Dosing
Activation
Results
Dupont-
19714
RV1
(2008)
In vitro
Mammalian
Chromosome
Aberration Test
(OECD
Guideline 473)
Chinese hamster
ovary cells
(CHO-Ki line)
HFPO dimer acid ammonium sail iX5".. puriiy)
0.3, 1, 3, 10, 30, 100, 300, 1 and V4" 1 ug/mL
for preliminary toxicity tesl
100, 500, 1,000,25(1(1. and \4~l (ig/mL for ilie
chromosome aberration assay for the 4-hour
nonactivated, 4-hour S'J-aclivated, and 20-houi'
nonactivated test conditions
Cytogenetic evaluations were conducted al 1,000,
2,500. and v47l \ieJmL for ilic 4-linur iiniactivated
and 4-hniir S'J-aclivated tesi conditions and at 100,
500, and 1.(>(>(> uu nil. for ilie 2(>-liour uouacli\ated
test condition
Xeuali\e control (sterile water) and positi\e control
(nnioni\ciu-( or csdnplkisphaniidc) also included
in siuds
With S9
The percentage of cells with
structural aberrations in the test
substance-treated groups was
not increased above that of the
vehicle control at any
concentration.
The percentage of cells with
numerical chromosome
aberrations at 2,500 and 3,471
(ig/mL in the 4-hour S9-
activated test conditions was
increased in a dose-dependent
manner above that of the vehicle
control. The change was outside
the historical control range and
considered biologically relevant.
Without S9
In the 20-hour nonactivated test
condition, substantial toxicity
was observed at 3,471 (ig/mL
and a substantial reduction in
mitotic index relative to vehicle
control was observed in the
mitotic index relative to vehicle
control.
The percentage of cells with
structural aberrations in the test
substance-treated groups was
not increased above that of the
vehicle control at any
concentration.
An increase in the percentage of
cells with numerical
chromosome aberrations was
observed at 3,471 (ig/mL in the
4-hour nonactivated condition
relative to vehicle control.
D-3
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE
NOVEMBER 2018
Study
Assay
Strain/Species
Dosing
Activation
Results
DuPont-
In vitro
Salmonella
HFPO dimer acid ammonium sail
325, 652, 977, 325(>. and 4S85 jig/plate lor ilie
mutagenicity test*
Without S9
Negative.
*Doses have been coi'icclcd lo account for 82.0% 11I IJ() dimer acid ammonium salt purity.
DuPont-
23220
(2007)
In vivo
Micronucleus
and
Chromosome
Aberration
Assay (OECD
Guidelines 474
and 475)
Primary bone
marrow cells
harvested from male
and female ICR
mice
(2 males or 5 of
each sex/dose lor
preliminary io\icil>
study)
(5 of each sc\ dose
for toxicity s>iud\)
HFPO dimer acid ammonium sail 5. I5(>u. and l."55nmkuh\ oralgavage
for to\ial\ siud\
''I7. 634, and l.2<>8 mu ku h\ oral uavaue lor
Micronucleus and ( 'hioniosiinie Xhei ialiou \ssay*
\cgalivc-No statistically significant increases in the
incidence of micronucleated polychromatic
erythrocytes or structural or numerical chromosomal
aberrations in bone marrow of male and female ICR
mice at doses up to and including the maximum
tolerated dose (1,268 mg/kg*).
l'nsiii\c control (coLhiciuci and negathc control
(sicnlc waieri also included mi ilie study
(5 of each sc\ dose
lor Microiiiiclcus
and C hromosome
\hciTaliou \ssa\ i
1 )oscs ha\ c hccu corrcclcd lo account for 82.6% HFPO dimer acid ammonium salt purity.
D-4
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Appendix E: Benchmark Dose Modeling
28 Day Oral (Gavage) Toxicity Study in Mice (DuPont-24459 2008)
Increased incidence of single-cell necrosis of hepatocytes and correlative increases in liver
enzymes were observed with liver identified as the primary target organ of toxicity.
Dichotomous models were used to fit of the incidence of single-cell necrosis in the liver. A
benchmark response (BMR) of 10% extra risk was chosen per the EPA's Benchmark Dose
Technical Guidance (USEPA, 2012). The data used for the modeling are in Table E-l below.
Table E-l. Single-Cell Necrosis in the Liver Selected lor Dose-Response Modeling
Dose (mg/kg bw/day)
Number of Animals
(Males)
Incidence ol' Single-Cell
Necrosis
0
10
0
0.1
10
0
3
10
4
30
II)
10
Note: mg/kg bw/day = milligrams per kilogram body weight per day.
The BMD modeling results for single-cell necrosis are summarized in Table E-2 and Figures E-l
and E-2. All the models had adequatep-values ( n I) The liMI) liMDL ratios were less than 5
for Gamma, Wei hull. I .ogProbit, Dichotomous-I Mil and I .ogl .ouislic so these models were not
further considered The remaining models ha\ e benchmark dose lower limits (BMDLs) that are
sufficiently close so the models with the lowest AICs were considered (these are the Multistage 2
and the Quantal-I.inear). Both the Multistage 2 and the Quantal-Linear are selected because the
BMT)l.|iiS of hotli these models rounded to I significant figure are the same 0.3 mg/kg bw/day.
Table K-2. Summary ol" HMD Modeling Results for Single-Cell Necrosis in Male Mice
Model"
(ioodness ol' I'il
BMDioPct
(mg/kg/day)
BMDLioPct
(mg/kg/day)
BMDioPct to
BMDLioPct
ratio
Basis for model selection
|)-\ aluc
AIC
Quantal-Linear
0.972
15 'J 1S
0.603
0.305
1.97
EPA OPPT selected both the
Quantal-Linear and Multistage
2 models. All the models had
adequate p-values (>0.1). The
BMD:BMDL ratios were > 5
for Gamma, Weibull,
LogProbit, Dichotomous-Hill
and LogLogistic so these
Multistage 2°
I ()()()
15.472
1.36
0.323
4.23
Logistic
I ODD
17.460
2.72
1.16
2.34
Probit
1.000
17.460
2.45
1.04
2.37
Multistage 3°
0.998
17.469
1.45
0.323
4.48
Gamma
1.000
17.460
1.88
0.323
5.80
Weibull
1.000
17.460
1.95
0.323
6.04
models were not further
considered. The remaining
models have BMDLs<4-fold
difference and the Quantal-
Linear and Multistage 2
models had the lowest AICs.
LogProbit
1.000
17.460
2.01
0.299
6.72
Dichoto mo us-Hill
LogLogistic
1.000
17.460
2.42
0.343
7.06
Notes: OPPT = Office of Pollution Prevention and Toxics.
E-l
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
a Selected model in bold; scaled residuals for selected model for doses 0, 0.1, 3, and 30 mg/kg/day were 0, -0.42, -0.05, 0.23,
respectively.
Quantal Linear Model, with BMR of 10% Extra Risk for the BMD and 0.95 Lower Confidence Limit for the BMDL
18 02/07 2018
Mice Rate by Dose with Fitted Cur
Single-Cell Necrosis in Male Mice;
Multistage Model, with BMR of 10% Extra Risk for the BMD and 0.95 Lower Confidence
Multistage
BMDL
Figure E-l. Plot of Incidence
Quantal-Linear Model for
Curve for the Selected Model
Dose Shown in mg/kg/day
Limit for the BMDL
Figure E-2. Plot of Incidence Rate by Dose with Fitted Curve for the Selected Model
Multistage 2° Model for Single-Cell Necrosis in Male Mice; Dose Shown in mg/kg/day
Modified Oral Reproductive/Developmental Toxicity Study in Mice (DuPont-18405-1037
2010)
Increased incidence of single-cell necrosis in the liver was observed in the parental males.
Dichotomous models were used to fit dose-response data (DuPont-18405-1037, 2010). A BMR
E-2
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
of 10% extra risk was chosen per the EPA's Benchmark Dose Technical Guidance (USEPA,
2012). The doses and response data used for the modeling are listed in Table E-3.
Table E-3. Single-Cell Necrosis in the Liver Selected for Dose-Response Modeling
Dose (mg/kg/day)
Number of Animals
(Males)
Incidence of Single-Cell
Necrosis
0
25
1
0.1
24
1
0.5
24
5
5
24
24
The benchmark dose (BMD) modeling results for single-cell necrosis are summarized in Table
E-4 and Figure E-3. The best fitting model was the Multistage 2 model based on adequate
^-values (>0.1), the BMDLs are sufficiently close, and the Multistage 2 model had the lowest
Akaike information criterion (AIC). The Multistage 2. T.ogistic, and Probit models had the
lowest AICs and the AICs were very close to each other Of these models the Multistage 2 model
has the lowest scaled residuals for the dose groups near the liMD and the BMDL and the scaled
residuals for the Multistage 2 model are show n in TaMe I>4 The BMDLio for Multistage 2
model is 0.15 mg/kg/day.
Table E-4. Summary ol' HMD Modeling Results lor Single-Cell Necrosis in Male Mice
Model"
Goodness of l-'ii
Scaled Residual for:
BMDioPct
(mg/kg/day)
BMDLioPct
(mg/kg/day)
Basis for
Model
Selection
/'-value
AIC
Dose Group
near KM 1)
Dose Group
near BMDL
Multistage 2°
0.995
45.285
0.007
-0.078
0.368
0.151
EPA OPPT
selected the
Multistage 2
model. All of
the models had
adequate p-
values (> 0.1),
the BMDLs
are sufficiently
close, and the
Multistage 2
model had the
lowest AIC.
Lo.uimic
I) ')(¦')
45.337
0.019
-0.181
0.362
0.253
Probil
0.960
45.358
0.029
-0.212
0.349
0.236
Weibull
1.000
47.275
0
0
0.407
0.166
Multistage 3°
I ODD
47.2 "5
0
0
0.408
0.145
Gamma
0 <><>2.
47.275
0
0.006
0.399
0.172
Dichotomous-
Hill
LogLogistic
0.977
47.275
0
0.021
0.464
0.253
LogProbit
0.977
47.275
0
0.021
0.443
0.248
Quantal-Linear
0.261
48.991
-0.756
-0.756
0.162
0.106
Notes'. OPPT = Office of Pollution Prevention and Toxics.
a Selected model in bold.
E-3
-------
This document is a draft for review purposes only and does not constitute Agency policy.
DRAFT FOR PUBLIC COMMENT - DO NOT CITE OR QUOTE NOVEMBER 2018
Multistage Model, with BMR of 10% Extra Risk for the BMD and 0.95 Lower Confidence Limit for the BMDL
Multistage
1
0.8
0.6
0.4
0.2
0
BiyiDL
0
1
2
3
4
5
dose
11:33 02/07 2018
Figure E-3. Plot of Incidence Rate by Dose with Fitted Curve for the Selected Multistage 2°
Model for Single-Cell Necrosis in Male Mice; Dose Shown in mg/kg/day
Table E-5 provides a summary of goodness of fit of the BMD modeling results for single-cell
necrosis in male mice.
Table E-5. Goodness of Fit Table of BMD Modeling Results for Single-Cell Necrosis in
Male Mice
Dose
Est. Prob.
Expected
Observed
Size
Scaled Residual
0
0.0375
0.937
1
25
0.07
0.1
0.0449
1.079
1
24
-0.08
0.5
0.2077
4.985
5
24
0.01
5
1
24
24
24
0
E-4
------- |